METHOD AND SYSTEM FOR MONITORING OPERATIONS OF A MINING SHOVEL

Abstract

A method and system for monitoring a mining shovel having a boom supported by a plurality of suspension cables is disclosed. The method involves receiving accelerometer signals from a plurality of accelerometers, each accelerometer being mounted on one of the plurality of suspension cables. The method also involves processing the accelerometer signals to extract a fundamental frequency associated with vibration of each suspension cable, the fundamental frequency being proportional to a tension in the suspension cable. The method further involves determining changes in the fundamental frequency as a function of time, the changes being indicative of an operating state of the mining shovel.

Description

RELATED APPLICATIONS

This application is claims priority benefits to U.S. Provisional Patent Application No. 63/401,840, filed Aug. 29, 2022, entitled “METHOD AND SYSTEM FOR MONITORING OPERATIONS OF A MINING SHOVEL.” This application is incorporated by reference herein in its entirety and made a part hereof.

BACKGROUND

Field of the Invention

This disclosure relates generally to heavy equipment monitoring and more particularly to monitoring forces in components of a mining shovel via a determination of a frequency of vibration in suspension cables of the shovel.

Description of Related Art

Large electrical mining shovels are commonly used in mining operations to load excavated ore onto haul trucks. The shovel or bucket is a critical element in the productivity of mining operations and failure or inefficient operation of the shovel can have a profound effect on ore throughput to downstream processes. Examples of mining shovels include rope shovels and dragline shovels each of which typically include a boom supported by suspension cables that support the weight of the boom and other loads. There remains a need for monitoring methods and systems that provide information on aspects such as productivity, operating health, and potential failure of mining shovels.

SUMMARY

In accordance with one disclosed aspect there is provided a method for monitoring a mining shovel having a boom supported by a plurality of suspension cables. The method involves receiving accelerometer signals from a plurality of accelerometers, each accelerometer being mounted on one of the plurality of suspension cables. The method also involves processing the accelerometer signals to extract a fundamental frequency associated with vibrations in each suspension cable, the fundamental frequency being proportional to a tension in the suspension cable. The method further involves determining changes in the fundamental frequency as a function of time, the changes facilitating a determination of an operating state of the mining shovel.

Receiving the accelerometer signals may involve receiving accelerometer signals from an accelerometer mounted on the suspension cable at a distance of at least one-third of the length from an end of the suspension cable.

Determining changes in the fundamental frequency may involve detecting changes in fundamental frequency that are indicative of a boom jacking event associated with an excavation being performed by the rope shovel.

Determining changes in the fundamental frequency may involve determining changes in fundamental frequency that are indicative of a potential failure of one of the plurality of suspension cables.

The method may involve determining a tension in each of the plurality of suspension cables based on the respective fundamental frequencies relating the fundamental frequency to tension.

The method may involve determining a tension in each of the plurality of suspension cables based on the respective fundamental frequencies and calibration data relating the fundamental frequency to tension.

The method may involve estimating forces on components of the mining shovel by receiving orientation signals from one or more attitude sensors associated with the components, the orientation signals defining an orientation of the components, determining a kinematic condition defining the position and orientation of the components based on the orientation signals and kinematic calibration data, and determining forces acting on the components of the mining shovel based on the kinematic condition, the orientation signals, the fundamental frequency tension in each of the plurality of suspension cables, and/or dynamic calibration data.

Estimating the forces on components of the mining shovel may involve estimating a weight of a payload in a payload container component of the mining shovel.

The method may involve performing a kinematic calibration to establish the kinematic calibration data by controlling the mining shovel to cause one of the components of the shovel to be successively located in each of a plurality of known positions and orientations with respect to the mining shovel, and processing the orientation signals based on the plurality of known positions and orientations to determine the kinematic calibration data for the mining shovel.

The method may involve performing a dynamic calibration to establish the dynamic calibration data by causing an unloaded payload container component of the mining shovel to maneuver through a trajectory while receiving the orientation signals and determining the changes in the fundamental frequency, and determining forces on the components of the mining shovel under unloaded conditions based on the changes in fundamental frequency.

The trajectory may be selected to emulate a digging operation of the mining shovel.

In accordance with another disclosed aspect there is provided a method for monitoring a mining shovel having a boom supported by a plurality of suspension cables. The method involves receiving accelerometer signals from a plurality of accelerometers, each accelerometer being mounted on one of the plurality of suspension cables at a distance of at least one-third of the length from an end of the suspension cable. The method also involves processing the accelerometer signals to extract a fundamental frequency associated with vibrations in each suspension cable, the fundamental frequency being proportional to a tension in the suspension cable. The method further involves determining changes in the fundamental frequency as a function of time, the changes facilitating a determination of an operating state of the mining shovel.

Other aspects and features will become apparent to those ordinarily skilled in the art upon review of the following description of specific disclosed embodiments in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

In drawings which illustrate disclosed embodiments,

FIG. 1A is a side view of a rope shovel in accordance with one disclosed embodiment;

FIG. 1B is a perspective view of an accelerometer and a portion of a suspension cable of the rope shovel shown in FIG. 1A;

FIG. 2 is block diagram of an embedded processor of the rope shovel shown in FIG. 1A;

FIG. 3 is a flowchart depicting blocks of code for directing the embedded processor circuit of FIG. 2 to process the accelerometer signals generated by the accelerometer shown in FIG. 1B;

FIG. 4A is a graphical depiction of examples of unprocessed accelerometer signals for three different placements of accelerometer along the suspension cable of the rope shovel shown in FIG. 1A;

FIG. 4B is a graphical depiction of examples of a filtered, windowed, and up-sampled accelerometer signal;

FIG. 4C is a graphical depiction of inverse autocorrelation results for a zero-crossing binary signal shown in FIG. 4B;

FIG. 4D is a graphical depiction of examples of f₀² values for a suspension cable subjected to an increasing and then reducing tension;

FIG. 5 is a schematic view of a kinematic model of the rope shovel shown in FIG. 1A;

FIG. 6 is process flowchart example of a kinematic calibration process executed by the embedded processor circuit shown in FIG. 3;

FIG. 7 is process flowchart example of a dynamic calibration process executed by the embedded processor circuit shown in FIG. 3;

FIG. 8 is a schematic representation of the rope shovel of FIG. 1A executing a simulated digging cycle; and

FIG. 9 is process flowchart example of a payload weight estimation process executed by the embedded processor circuit shown in FIG. 3.

DETAILED DESCRIPTION

Referring to FIG. 1A, a mining shovel is shown generally at 100. In the embodiment shown the mining shovel 100 is a rope shovel including a dipper 102 that acts as a payload container for loading an excavated payload 103. In other embodiments the mining shovel may be a dragline shovel that includes a payload container in the form of a bucket that is dragged to load the payload. The dipper 102 of the rope shovel 100 includes a plurality of ground engaging teeth 104 for excavating ore. The rope shovel 100 also includes a crawler track 106 and a superstructure 108 mounted for rotation on the crawler track. The superstructure 108 includes an operator's cab 110 that accommodates the shovel controls and the operator of the shovel. A boom 112 is pivotably mounted to the superstructure 108 at a boom pivot 114. A crowd 116 is received within a saddle block 118 mounted on the boom 112 for rotation about a saddle pivot 120. The boom 112 is supported by a plurality of suspension cables 122 connected between an upper end of the boom 112 and an A-frame 124 located on the superstructure 108. In this embodiment, while only two suspension cables are visible due to the elevational view, there would typically be four suspension cables, each suspension cable 122 supporting a portion of the weight of the boom 112.

The dipper 102 is pivotably mounted to a distal end of the crowd 116. A pair of hoist cables 126 run from a winch drum (not shown) within the superstructure 108 over a pulley 128 and are coupled to either side of the dipper 102 via a bail hanger or padlocks 130. The crowd 116 pivots within the saddle block 118 about the saddle pivot 120 when the hoist cables 126 are extended or retracted or the angle of the dipper 102 is changed. The saddle block 118 also permits the crowd 116 to be extended or retracted to force the dipper 102 into a mine face 134 of a bench 136 for excavating ore from the bench.

The suspension cables 122 support the weight of the boom 112, the crowd 116, the saddle block 118 and the dipper 102. The suspension cables 122 are further subjected to load forces when excavating the mine face 134 or payload forces when carrying a payload of ore in the dipper 102. An accelerometer 138 is coupled to each of the suspension cables 122 for generating signals representing vibrational oscillations of the respective suspension cables 122. The suspension cables 122, when under tension T will tend to vibrate at a frequency that is proportional to the square root of the tension and inversely proportional to the length L of the cable extending between the A-frame 124 and the end of the boom 112. The tension in each suspension cable 122 is thus proportional to its fundamental harmonic frequency, which in turn is a function of the length L of the cable. The tension in each of the suspension cables 122 may be estimated by processing the signals generated by the respective accelerometers 138 to extract a value for the fundamental frequency of vibration associated with the cable. In one embodiment, the tension T in each suspension cable 122 may be written in the form:

T=αf₀²+γ;   Equation 1

where α is a function of the linear density and length L of the suspension cable 122, γ reflects the sag-extensibility and bending stiffness properties of the rope, and f₀ is the fundamental frequency of vibration of the cable determined from the accelerometer signal. The tension T is thus proportional to f₀². If it is desired to determine the actual tension T in the suspension cables 122, the constants α and γ may be empirically determined in a calibration process or calculated from characteristics of the cable 122 such as the mass, length, cross-sectional area, Young's (elastic) modulus, and moment of inertia.

Referring to FIG. 1B, the accelerometer 138 and a portion of the suspension cable 122 are shown in enlarged detail. In one embodiment the accelerometer 138 may be a triaxial accelerometer sensor element that measures vibration in three orthogonal axes (X,Y,Z). One example of a suitable accelerometer is the Kistler 8763 integrated electronics piezoelectric triaxial accelerometer available from Kistler Instrument Corp. of Amherst NY, USA. The accelerometer sensor element 138 is mounted on a base 160, which in this embodiment is configured for mounting to various cables of diameters ranging from about 60 mm to 115 mm. The base 160 includes slots 162 and 164 for receiving straps 166 and 168 that secure the base to the suspension cable 122. A cover 170 encloses the accelerometer 138 to prevent ingress of dust and debris into the sensor enclosure. In one embodiment, the accelerometer 138 is connected via a cable 172 that provides power to the accelerometer and carries generated accelerometer signals back to a control room of the rope shovel 100. In other embodiments the respective accelerometers 138 may be powered by a battery and signals may be transmitted wirelessly to the control room of the rope shovel 100.

Referring back to FIG. 1A, in the embodiment shown the rope shovel 100 includes an embedded processor circuit 150 for receiving and processing the accelerometer signals from the accelerometers 138. A block diagram of the embedded processor circuit 150 is shown in FIG. 2. Referring to FIG. 2, the embedded processor circuit 150 includes a microprocessor 200, a memory 202, and an input output port (I/O) 204, all of which are in communication with the microprocessor 200. In other embodiments (not shown), the embedded processor circuit 150 may be partly or fully implemented using a hardware logic circuit including discrete logic circuits and/or an application specific integrated circuit (ASIC), for example.

The embedded processor circuit 150 also includes a mass storage unit 208 such as a hard drive or solid state drive in communication with the microprocessor 200. Program codes for directing the microprocessor 200 to carry out functions related to monitoring conditions associated with the mining shovel 102 may be stored in the memory 202 or the mass storage unit 208.

The I/O 204 includes an interface 220 for receiving the accelerometer signals from the respective accelerometers 138. In this embodiment, each of the accelerometers 138 are connected via a signal conditioner 230 to the respective inputs 222-228. The signal conditioner 230 acts as a power supply for the accelerometers 138 and also receives, conditions and amplifies low signal level analog accelerometer signals produced by the accelerometers to generate a ±10V analog output signal. The interface 220 of the I/O 204 includes an analog to digital converter that is configured to covert the conditioned analog accelerometer signals into digital representations thereof for processing by the microprocessor 200. In embodiments where the accelerometers 138 are configured for wireless transmissions, the I/O 204 would include a wireless radio in place of or in addition to the interface 220 for receiving wirelessly transmitted accelerometer signals.

In this embodiment, the I/O 204 also includes an interface 240 having an output 242 for producing display signals for driving a display 250. The display 250 may be located in the operator's cab 110 of the rope shovel 100. In one embodiment the display 250 may be implemented as a touchscreen display and the interface 240 may also include a USB port 244 in communication with a touchscreen interface of the display for receiving inputs from an operator. The I/O 204 may alternatively have additional USB ports (not shown) for connecting a keyboard and/or other peripheral interface devices.

Referring to FIG. 3, a flowchart depicting blocks of code for directing the microprocessor 200 to process the accelerometer signals is shown generally at 300. The blocks generally represent codes that may be read from the mass storage unit 208 or memory 202 for directing the microprocessor 200 to perform various signal processing functions. The actual code to implement each block may be written in any suitable program language, such as C, C++, C#, Java, and/or assembly code, for example.

Block 302 directs the microprocessor 200 to receive the analog accelerometer signal from one of the inputs 222-228 of the interface 220. The signals at the inputs 222-228 have been conditioned by the signal conditioner 230 but are otherwise unprocessed raw signals generated by the respective accelerometers 138. For embodiments where the accelerometer 138 is a tri-axis accelerometer, separate signals are generated for each of the X-axis, Y-axis and Z-axis directions (shown in FIG. 1B). For the mounting condition shown in FIG. 1B, the Z-axis is directed generally downwardly with respect to the suspension cable 122. This Z-axis direction is much more closely aligned with the direction of gravitational forces than either the X-Axis or Y-axis directions. It is thus expected that gravitational forces on the suspension cables 122 would provide increased excitation of the harmonic modes of vibration of the respective cables and result in stronger signals. The amplitude of the Z-axis signal will generally be substantially larger than the X-Axis and Y-axis signals. In the embodiments described herein, the Z-axis signals are received as the respective analog accelerometer signals at the inputs 222-228.

Referring back to FIG. 1A, in this embodiment each of the accelerometers 138 is located a distance D from an end of the suspension cable 122 connected to the A-frame 124. Referring to FIG. 4A, examples of unprocessed Z-axis accelerometer signals for three different placements of the accelerometer 138 along the cable 122 are shown generally at 402, 404, and 406. The signal 402 is produced when the accelerometer 138 is placed at D=0.25 L, the signal 404 is for the accelerometer placed at D=0.33 L, and the signal 406 is for the accelerometer placed at D=0.4 L. When the accelerometer 138 is placed at a location where D<0.33 L, the accelerometer signal is substantially attenuated as in the case of the signal 402. The same effect applies if the accelerometer 138 is placed at the other end of the suspension cable 122 proximate the pulley 128. Placement of the accelerometer 138 at 0.33 L≤D<0.5 L was found to effectively negate this attenuation effect as in the case of the signals 404 and 406. In the embodiments described below the signals are generated for the accelerometer placement at D=0.33 L.

Referring back to FIG. 3, block 302 further directs the microprocessor 200 to cause the interface 220 to perform analog to digital conversion on the signal 404. In one embodiment the interface 220 samples the analog signal at a sampling rate of 5 kHz to generate a sampled digital signal version of the analog accelerometer signal, but other variations are possible.

The process 300 then continues at block 304, which directs the microprocessor 200 to filter the digital accelerometer signal to remove higher harmonics. Typical suspension cables 122 have strong first and second harmonics with third and higher harmonics being substantially attenuated. For an example of a typical suspension cable 122 having a length of about 20 meters, the fundamental frequency under tension may be in the region of about 4 Hz and the filter may be configured to have a cutoff frequency at about 12 Hz, which effectively removes the third and higher harmonics. In other embodiments the material suspension cable 122 may result in higher order harmonics that have significant amplitude, in which case the filter block 304 may be configured differently to include these harmonics. In some embodiments the filtering performed at block 304 may be implemented as a band pass filter in which frequencies lower than an expected fundamental frequency are also removed by filtering. In one example, the low cut-off may be at a frequency of about 0.5 Hz to remove DC offset from the signal, but other frequencies may be used.

Block 306 then directs the microprocessor 200 to multiply the digital samples in the signal 404 by a finite-length window to generate a waveform that may be digitally processed to extract the cable tension T. In this embodiment the window has a duration of about 2 seconds, which for a suspension cable 122 that has a fundamental vibration frequency of at least 4 Hz will capture at least two periods of the fundamental harmonic. The window duration is selected to provide a sufficiently long sample of the accelerometer signal to facilitate extraction of the fundamental frequency of the cable vibration. The window length is a parameter that may also be adjusted to improve computational speed.

Block 308 then directs the microprocessor 200 to up-sample the filtered signal to provide a finer time spacing between adjacent discrete sample values. For the above example of a 5 kHz sampled signal 404, the adjacent sample values are separated by 200 microseconds, but other frequencies may be used. By up-sampling the signal 410 to 100 kHz, additional sample values are generated, thus reducing the sample spacing to 10 microseconds. The up-sampling process may involve interpolating between adjacent sample values of the signal to generate the additional samples spaced at 10 microsecond intervals.

Examples of a filtered, windowed, and up-sampled signal from one of the accelerometers 138 is shown graphically in FIG. 4B at 410. Referring to FIG. 4B, the accelerometer signal in this case includes two superimposed sinusoidal vibration harmonics with third and higher harmonics having been substantially removed by the filtering block 304. While the signal 410 is shown as a continuous line, it should be understood that this is a digital signal comprising a plurality of discrete time-sampled values spaced apart by about 10 microseconds. In this case, a windowing function has also been applied and the digital signal extends in time between 0 and 2 seconds.

Block 310 then directs the microprocessor 200 to extract all of the zero-crossings in the signal 410. In this embodiment, the zero-crossing information is generated in the form of a binary signal 412. The binary signal 412 transitions from binary 0 to binary 1 at each zero-crossing from negative to positive. The binary signal 412 also transitions from binary 1 and binary 0 at each zero-crossing from positive to negative. These transitions are indicated by the broken lines in FIG. 4B, shown for the first few cycles of the signal 410. The up-sampled version 410 of the signal 400 facilitates a more accurate zero-crossing determination than would be the case if the 5 kHz sampled signal 404 were to be processed at block 310.

Block 312 then directs the microprocessor 200 to perform a bitwise autocorrelation between the zero-crossing binary signal 412 and a delayed copy of the same signal. A delay that produces a highest autocorrelation result yields an optimal estimate for the periodicity of the signal 410. The estimated periodicity may then be used to determine the frequency of the fundamental vibration of the suspension cable 122 and thus the tension T (using equation 1 above).

In one embodiment, the autocorrelation may be efficiently computed by performing a series of bitwise exclusive OR (XOR) operations between the original zero-crossing binary signal 412 and a copy of the signal 412 delayed to each subsequent rising edge of the original signal. The delayed copies of the zero-crossing binary signal 412 may be zero padded to prevent detection of subharmonic frequencies. The closest match between the delayed signal 412 and original signal produces the smallest XOR result. The original zero-crossing binary signal 412 and the delayed copy of the signal may be encoded as a 32-bit or 64-bit integers, which facilitates performing a computationally efficient single XOR operation for each autocorrelation on the microprocessor 200. In contrast, performing an autocorrelation on the accelerometer signal 410 would be significantly more computationally expensive than the simplified binary signal 412. Performing the autocorrelation on the zero-crossing binary signal 412 also has the effect of removing any variability due to changes in amplitude of the signal 410.

Block 314 then directs the microprocessor 200 to select the minimum autocorrelation value as an estimate of the fundamental frequency of vibration f₀ of the suspension cable 122. In one embodiment the bitwise autocorrelation is performed as follows:

${\mathrm{XOR}}_{\mathrm{SUM}}=\sum \mathrm{XOR}\left(\mathrm{original}\mathrm{bitstream},\mathrm{delayed}\mathrm{bitstream}\right),\mathrm{and}$ $\mathrm{Periodicity}=\frac{1}{\max \left({\mathrm{XOR}}_{\mathrm{SUM}},1\right)}$

where the periodicity thus assumes a value between 0 and 1 and where 1 indicates a perfect match. Referring to FIG. 4C, periodicity results for the zero-crossing binary signal 412 are graphically depicted as a function of the delay time at 420. The maximum periodicity value occurs at a point 422 having an associated time of 0.2905 seconds. This corresponds to the minimum autocorrelation value indicative of the closest match between the delayed signal 412 and original signal that provides the periodicity of the signal. Referring back to FIG. 4B, the first rising edge 414 of the zero-crossing binary signal 412 occurs at a time of 0.0698 seconds. The fundamental frequency may thus be calculated from the period provided by the autocorrelation as follows:

$f_o=\frac{1}{\mathrm{Period}}=\frac{1}{0.2905-0.0698}=4.52960\mathrm{Hz}.$

From the value of f₀, the frequency of the fundamental frequency of vibration of the suspension cable 122 above may then be used along with determined values for the α and γαconstants in Equation 1 to yield the tension of the suspension cable 122. The process blocks 302-312 thus yield a single value of f₀ for the example of the 2-second windowed time interval above.

Blocks 302 to 312 may then be continuously repeated for further 2-second windowed samples of the accelerometer signal to yield an ongoing plurality of f₀ estimates. In one embodiment the process 300 may be repeated at a rate of about 3 Hz to 10 Hz.

Referring to FIG. 4D, examples of a f₀² values for a cable subjected to an increasing and then reducing tension are shown generally at 430. A series of f₀² values determined in accordance with the process 300 are shown at 432. The f₀² values 430 include some frequency impulse noise that may be imparted by higher frequency external stimuli. The impulse noise may be substantially eliminated by performing a median filtering of the f₀² values to yield the filtered values shown at 434 in FIG. 4D. Median filtering is effective in removing impulse noise. The filtered values 434 in FIG. 4D show a relatively smooth progression in f₀² and may thus be taken as being proportional to the tension T applied to the cable.

In one embodiment, filtered f₀² values may be monitored to reveal issues associated with the structural integrity of the suspension cables 122 and other structural elements such as the boom 112. Rope shovels are subject to a structural fatigue process known as boom jacking, which may result in fatigue of structural elements leading to cracking or other potential component failures, e.g. the cables 122. Referring back to FIG. 1A, when excavating the mine face 134 of the mining bench 136, the crowd 116 is initially extended to force the ground engaging teeth 104 of the dipper 102 into the mine face and the hoist cable 126 is retracted to raise the bail hanger 130 into the orientation shown. Excavating forces on the crowd 116 or bail hanger 130 may cause the boom 112 to initially stall in the mine face 134 and then rapidly release. This rapid release may cause the boom 112 to snap backwards subjecting the boom and cables 122 to large vibrations that may be detrimental to the shovel 100.

In one embodiment, a potential boom jacking event may be detected by monitoring the f₀² values for stall events (i.e. high f₀² values) and/or release events (i.e. rapidly reducing f₀² values). This pattern may be detected in the foe values and a warning issued to the operator. Alternatively, some rope shovels may be configured with an ability to automatically compensate crowding and bail hanger forces. This automatic compensation may be activated based on a detection of a potential boom jacking event from the f₀² values to prevent a possible boom jacking event from occurring.

The f₀² values also provide information that may be monitored over time to detect issues in the structural integrity of the suspension cables 122. By monitoring f₀² values over successive dig cycles for each of the plurality of suspension cables 122, a change in cable tension occurring in one or more of the cables may be indicative of a potential structural issue. For example, if the tension indicated by the f₀² values for one cable decreases over time while the tension on other cables increases over time. This condition may be caused by a structural issue on the cable having the reducing tension over time, such as wearing or beginning of a yield failure, or from individual strands of the suspension cables 122 fracturing.

In the above embodiments where changes in tension are inferred from changes in f₀² values, the constants α and γ need not be determined since the f₀² values are directly proportional to cable tension. As disclosed above the constants could alternatively be estimated or calculated from characteristics of the cable 122 if it is desired to determine an actual approximate tension in each cable 122. The actual tension may be useful in determining forces on the components of the rope shovel 100 or in other applications such as payload monitoring. However, in practice slight differences in characteristics between the plurality of cables 122, such as the exact length of each cable, may reduce the accuracy of determined tension T. In one embodiment the constants α and γ may be determined in a calibration process that accounts for variations between the plurality of cables.

Referring back to FIG. 1A, in the embodiment shown the rope shovel 100 further includes a plurality of attitude sensors disposed on components of the shovel to measure the orientation of these elements with respect to the Earth. The plurality of attitude sensors include an attitude sensor 140 mounted in the operator's cab 110 to provide the roll, pitch, and yaw (swing) of the operator's cab 110. The plurality of attitude sensors also include a boom attitude sensor 142 on the boom, a saddle attitude sensor 144 on the saddle block 118, and a bail hanger attitude sensor 146 on the bail hanger 130, which provide the orientation signals representing the orientation of these components. The attitude sensors 140-146 may be implemented using Inertial Measurement Unit (IMU) sensors that include 3-axis accelerometers, gyroscopes, and magnetometers, to provide real-time 3D orientation signals that identify the attitude of the respective components to which the sensor is attached. An example of a suitable IMU sensor is the VN-100 IMU/AHRS sensor available from Vectornav of Dallas, TX, USA. The VN-100 IMU/AHRS sensor generates a signal in either ASCII or binary format that identifies the attitude of the sensor.

Referring back to FIG. 2, in the embodiment of the embedded processor circuit 150 shown, the I/O 204 includes data inputs 252-258 for receiving the orientation signals from the respective attitude sensors 140-146. The respective orientation signals received at the data inputs 252-258 facilitate computation by the microprocessor 200 of the rectilinear and angular position, velocity, and acceleration of the above components of the rope shovel 100 in real time based on a kinematic model including the various linkages of the rope shovel 100.

Referring to FIG. 5, a kinematic model representation of the rope shovel 100 is shown generally at 500 with the rope shovel 100 shown for reference in the background. A ground coordinate frame 502 (x₀, y₀, z₀, where y₀ is aligned extending out of the plane of the page) has an origin at G and acts as a global frame of reference for the kinematic model. A shovel coordinate frame 504 (x₁, y₁, z₁) has an origin at O and represents the orientation of the superstructure 108 and operator's cab 110 with respect to the ground coordinate frame 502. The shovel coordinate frame 504 is displaced from the ground coordinate frame 502 by a distance indicated by the line GO. The orientation signal generated by the attitude sensor 140 in the operator's cab 110 provides the real-time orientation of the shovel coordinate frame 504 with respect to the ground coordinate frame 502. In FIG. 5 the shovel coordinate frame 504 is shown in alignment with the ground coordinate frame 502, but in practice the shovel coordinate frame may be pitched at an angle q_P or rolled at an angle q_R with respect to the ground coordinate frame 502 and may also be rotated about the z₁ axis depending on a swing angle q_w about the z₁-axis between the superstructure 108 and the crawler track 106.

The boom 112 is represented in the kinematic model 500 by a line BV extending along the boom from the boom pivot 114 located at the coordinate frame (x₂, y₂, z₂) to the center V of the pulley 128. The boom 112 is disposed at an angle q_B about the z₂-axis which extends out of the plane of the page. The boom angle q_B may be calculated by the microprocessor 200 from the orientation signal generated by the boom attitude sensor 142. The saddle pivot 120 is located at an origin L of a saddle coordinate system (x₃, y₃, z₃) disposed along the line BV. The saddle is oriented at an angle about the z₃-axis which also provides the angle q_S of the crowd 116. A point S on the saddle is displaced by the line LS from the saddle pivot 120 (L) and the crowd 116 extends outwardly through the point S and terminates at a point C at the end of the crowd. The extension of the crowd 116 caused by movement of the crowd through the point S on the saddle block 118 is represented by a variable r_C (i.e. the distance between S and C). The crowd angle q_S may be calculated by the microprocessor 200 based on the orientation signal generated by the saddle attitude sensor 144. In other embodiments the saddle attitude sensor 144 may be placed on the crowd 116 since the saddle 118 and crowd will be disposed at the same angle.

The dipper 102 is pivotably connected to the crowd 116 at the point C. An angle θ_CH of the dipper 102 is defined between the line SC and a line CH extending between the origin C and the connection between the bail hanger 130 and the dipper. The tips of the plurality of ground engaging teeth 104 are located at a point E with respect to the bail hanger connection H.

From the center of the pulley at V, a line VU defines the offset between the pulley center and the hoist cable 126. The hoist cable 126 extends perpendicular to the line VU along a line UH extending between the bail hanger 130. The length of the line UH is represented by the variable r_H corresponding to the length of the hoist cables 126 when played out from the winch drum. The line UH is at an angle β with respect to the vertical. The angle β is provided by the orientation signal generated by the bail hanger attitude sensor 148.

The kinematic model of FIG. 5 may be used in conjunction with the orientation signals received from the attitude sensors 140-146 to determine the location of the various shovel components. The angles q_B, q_S and β are calculable from information contained in the signals generated by the attitude sensors 142-148. The variable crowd extension r_C and the hoist cable extension r_H may be determined from the kinematic model following a calibration procedure detailed below. The distances OB, BL, LS and VU may be measured or determined from a specification of the rope shovel 100. The distance CH may be configured according to a desired rake angle and tooth angle at the operating site. The remaining parameters CH and θ_CH may be determined in a kinematic calibration process. During installation, the sensors 142-146 may be offset or oriented at an angle that differs from the actual q_B, q_S, or β angles of the boom 112, saddle block 118 and crowd 116, and bail hanger 130 and the calibration process may also take these offsets into account.

Referring to FIG. 6, a process flowchart example of a kinematic calibration process executed by the microprocessor 200 that may be implemented for the rope shovel 100 is shown generally at 600. The process begins at 602 when the operator of the rope shovel 100 positions the ground engaging teeth 104 of the dipper 102 in a known location within the ground coordinate frame 502. As an example, the dipper 102 may be lowered and angled as shown in FIG. 5 in broken outline at 506 until the teeth 104 engage the ground at a point P₀ within the x₀-z₀ plane of the ground coordinate frame 502.

Block 604 then directs the microprocessor 200 to receive the orientation signals from the attitude sensors 140-146 which provides values for the angles q_B, q_S and β. Block 606 directs the microprocessor 200 to determine whether data has been generated for all points P₀ to P₄ to facilitate calibration calculations. The distances OB, BL, LS and VU may be measured or determined from a specification of the rope shovel 100. In this embodiment the distance CH and the angle θ_CH, which may vary from the specification based on user configuration, may be calculated along with any sensor offsets and misalignments. In the embodiment shown, a single point P₀ does not provide sufficient data to uniquely calculate the unknown parameters (q_B, q_S, β, CH, θ_CH, r_c and r_H). In this embodiment block 604 is repeated for five different points P₀ to P₄.

If at block 606 there remain further points to be processed, the microprocessor 200 is directed to block 608 where the operator controls the shovel to position the teeth 104 of the dipper 102 at the next point on the ground. Block 604 is then repeated for each subsequent point P₂ to P₄. At block 606, when data for the point P₄ has been generated, the microprocessor 200 is directed to block 610.

Block 610 then directs the microprocessor 200 to calculate the kinematic parameters including parameters CH and θ_CH, the crowd extension r_c, and the hoist extension r_H based on the angles q_B, q_S and β and based on the ground engaging teeth 104 being located within the ground coordinate frame 502 at the respective points P₀ to P₄. Block 610 also directs the microprocessor 200 to store the determined parameters for the kinematic model 500 in the memory 202 for later use. These kinematic parameters permit the location of the components of the rope shovel 100 shown in FIG. 5 to be computed for any combination of the angles q_B, q_S and β received in real time from the attitude sensors 140-148.

Generally for a mining shovel 100, the shovel is propelled via the crawler track 106 to position it with respect to the face 134 of the mining bench 136. The crawler track 106 is generally not further actuated during a digging operation on the mine face 134. The pitch and roll of the superstructure 108 and operator's cab 110 may vary during digging depending on the level and orientation of the ground engaged by the crawler track 106, interactions of the dipper 102 with the bench face, and any changes in the center-of-mass of the dipper and payload due to movement of the shovel components. The rotation of the boom 112, while not directly actuated, is also considered during digging operations to characterize boom jacking. Digging movements of the mining shovel 100 include actuated movements in three axes of the hoist cables 126, the crowd 116, and lateral swinging about the z₁=axis (i.e. the q_w angle). A state vector q may be used to describe the generalized coordinates corresponding to the degrees of freedom of the rope shovel 100:

$q=\left\{\begin{array}{c}{q}_P\\ q_R\\ q_W\\ q_B\\ q_S\\ \beta \end{array}\right\}$

with parameters as follows:

• • q_P for cab pitch, rotation about y₁;
  • q_R for cab roll, rotation about x₁;
  • q_W for swing orientation, rotation about z₁;
  • q_B for boom orientation, rotation about z₂;
  • q_S for crowd orientation, rotation about z₃; and
  • β0 for bail angle, with respect to the vertical.

The displacement of the crowd 116 r_C and extension of the hoist cables 126 r_H, may be determined using the process 600 to resolve the kinematic model of FIG. 5, and are therefore not independent. The equations of motion characterizing the shovel dynamics may be expressed as follows:

F_ext=M({umlaut over (q)})+S(q, {dot over (q)})+g(q)   Equation 2

which represents a system of N nonlinear equations of motion describing the dynamics for N degrees of freedom with kinematic states q. In Equation 2, M includes the inertial terms, S the centripetal and Coriolis terms, and g the gravitational terms. F_ext includes the external forces and torques on the system from the suspension and hoist cables, crowding forces and torques, dig forces on the dipper, and the payload weight. Friction and other non-conservative forces are also modelled in the representation for F_ext in Equation 2. The equations of motion may be derived via classical mechanics using Newton-Euler, Lagrangian mechanics, or other similar considerations.

Equation 1 above expresses the tension T for each suspension cable 122 as a function of its fundamental resonant frequency f₀. For a plurality of m suspension cables each exerting a suspension force T_m=αf₀²+γ on the boom, it is convenient to describe the combined suspension moment τ on the boom as follows:

$\tau =\mathrm{\upsilon \alpha }$ $\tau =\left[\begin{array}{ccccc}{f}_1^2{r}_1& f_2^2{r}_2& \dots & f_m^2{r}_m& 1\end{array}\right]\left[\begin{array}{c}{\alpha }_1\\ {\alpha }_2\\ ⋮\\ {\alpha }_m\\ \delta \end{array}\right]$

where:

• • υ∈^1×m+1r_i=r_i(q) contains the frequency-squared and moment arm for each suspension cable,
  • υ∈^1×m+1r_i=r_i(q);
  • α∈^m+1×1 contain length and linear density properties for each cable α₁-α_m and generally corresponds to the term α in Equation 1; and
  • δ is an aggregate bias term for the sag-extensibility and bending stiffness properties of the plurality of cables 122 and generally corresponds to the y parameter in Equation 1.

The system of equations described in Equation 2 may be manipulated to derive a single nonlinear equation expressed solely as a function of the suspension moment τ and the corresponding shovel kinematics q, {dot over (q)} and {umlaut over (q)} such that unknown inertial and geometric parameters Φ are linearly separable from the nonlinear kinematic terms contained in the n-dimensional regressor Y(q, {dot over (q)}, {umlaut over (q)}):

τ=vα=YΦ  Equation 3

where:

• • Y ∈^1×n represents the regressor containing nonlinear functions of the shovel kinematics; and
  • Φ∈^n×1 contains the unknown inertial and geometric parameters of the shovel.

The parameters α and Φ in Equation 3 above are unknown but may be determined in a calibration process. Referring to FIG. 7, a process flowchart example of a dynamic calibration process executed by the microprocessor 200 that may be implemented for the rope shovel 100 is shown generally at 700. The process begins at block 702 when the operator controls the rope shovel 100 to cause the dipper 102 to execute a maneuvering sequence in which the crowd 116 and hoist cables 126 are extended and retracted to follow a trajectory.

Referring to FIG. 8, the rope shovel 100 is schematically represented and the crowd 116 and hoist cables 126 are controlled to cause the dipper 102 to move along a maneuvering trajectory 800. The calibration process involves collecting p measurement samples of the shovel's kinematics q, {dot over (q)}, {umlaut over (q)} and suspension cable frequencies f₁ . . . f_m while controlling the rope shovel 100 to follow the maneuvering trajectory 800 with the dipper 102 unloaded. The p samples may be represented as follows:

υα=YΦ  Equation 4

Which may be written as:

$\left[\begin{array}{ccccc}{f}_{11}^2{r}_{11}& f_{21}^2{r}_{21}& \dots & f_{m1}^2{r}_{m1}& 1\\ f_{12}^2{r}_{12}& f_{22}^2{r}_{22}& \dots & f_{m2}^2{r}_{m2}& 1\\ ⋮& ⋮& \ddots & ⋮& ⋮\\ f_{1p}^2{r}_{1p}& f_{2p}^2{r}_{2p}& \dots & f_{mp}^2{r}_{mp}& 1\end{array}\right]\left\{\begin{array}{c}{\alpha }_1\\ {\alpha }_2\\ ⋮\\ {\alpha }_4\\ \delta \end{array}\right\}=\left[\begin{array}{cccc}{Y}_{11}& Y_{21}& \dots & Y_{n1}\\ Y_{12}& Y_{22}& \dots & Y_{n2}\\ ⋮& ⋮& \ddots & ⋮\\ Y_{1p}& Y_{2p}& \dots & Y_{\mathrm{np}}\end{array}\right]\left\{\begin{array}{c}{\Phi }_1\\ {\Phi }_1\\ ⋮\\ {\Phi }_n\end{array}\right\}$

with υ∈^p×m+1, +∈^m+1×1, Y∈^p×n, and Φ∈^n×1. To preserve the modelling assumptions, the dynamic calibration maneuver along the trajectory 800 must be performed without the dipper 102 engaging the ground or the introduction of any external forces. The maneuvering trajectory 800 is established to elicit an appropriate velocity and acceleration response in the kinematics to provide sufficient data for computations performed to solve for the unknown parameters. In one embodiment, the maneuvering trajectory 800 is established to emulate a digging operation of the rope shovel 100 and the trajectory includes lateral swing movements of the superstructure 108 (i.e. the angle q_W) with respect to the crawler track 106.

Referring back to FIG. 7, block 704 then directs the microprocessor 200 to receive f₀² values generated at block 314 of the process 300 as described above. Block 706 then directs the microprocessor 200 to receive the orientation signals generated by the attitude sensors 140-146 for the superstructure 108, boom 112, saddle 116, and a bail hanger 130. These values are thus generated for each of a plurality of points along the maneuvering trajectory 800 to generate a dynamic calibration dataset.

Block 708 then directs the microprocessor 200 to use the dynamic calibration dataset to solve Equation 4 for the unknown parameters via least squares:

Y^†υα=Y^†YΦ

Y^†υαα⁺=Y^†YΦα⁺

Y^†υ=Φα⁺

where Y^† represents the left pseudoinverse of Y, such that Y^†Y=I, and is computed as follows:

Y^†=(Y^TY)⁻¹Y^T

and α⁺ represents the right pseudoinverse of a such that αα⁺=I, and is computed as follows:

α⁺=α^T(α^Tα)⁻¹

The unknown parameters may subsequently be lumped together as Π=Φα⁺, where Π∈^n×m+1 and may be estimated as follows:

T^†υ={circumflex over (Π)}  Equation 5

The no-load calibration maneuver along the trajectory 800 establishes a baseline tension of the suspension cables 122 for any extension of the crowd 116 and hoist cables 126 and results in a set of no-load system parameters Π_NL being identified:

υ_NL=YΠ_NL   Equation 6

{circumflex over (Π)}_NL=Y^†υ_NL

In one embodiment the no-load system parameters may be used to determine a weight of payload in the dipper 102 based on a difference between the no-load and loaded conditions. Other digging forces such as a crowd extension force may be similarly determined.

Referring to FIG. 9, a process flowchart example of a payload weight estimation process executed by the microprocessor 200 is shown generally at 900. The process 900 is only executed after the kinematic parameters and dynamic parameters under no load have been determined in the calibration process 600 and 700 in FIGS. 6 and 7 respectively. The process 900 begins at block 902, which directs the microprocessor 200 to receive f₀² values generated at block 314 of the process 300. Block 904 then directs the microprocessor 200 to receive the signals generated by the attitude sensors 140-146 for the orientation of the superstructure 108, the boom 112, the saddle 116, and the bail hanger 130. Block 906 then directs the microprocessor 200 to determine the current kinematic condition of the rope shovel 100 using the orientation signals received at block 904. Following the kinematic calibration process 600, the unknown kinematic parameters will have been determined and may be used to calculate the current kinematic condition of the rope shovel 100 based on the signals generated by the attitude sensors 140-146.

Block 908 then directs the microprocessor 200 to determine whether the dipper 102 is in a position that facilitates estimation of a payload being carried in the dipper based on the kinematic condition of the rope shovel 100. In one embodiment block 908 involves aspects such as whether the payload and payload rate of change meet a threshold value and the lateral swing angular velocity and crowd 116 rectilinear and angular velocities. These determinations ensure that the dipper 102 has disengaged from the mine face 134 such that only the payload exerts an external force on the shovel components, and that the mining shovel 100 is operating within a viscous (linear) friction regime so as to neglect the contribution of nonlinear stiction effects. As an example, during active excavating of the mine face 134 the tension in the suspension cables 122 will be increased by the digging engagement forces, which would not be an appropriate time to perform payload estimation. However, immediately following the excavating operation when the dipper is not moving or moving slowly, forces acting on the dipper will be primarily due to the payload, which facilitates payload estimation. If at block 908, the rope shovel 100 is not in a kinematic condition that facilitates payload estimation, the microprocessor 200 is directed back to block 902 and blocks 902-908 are repeated.

If at block 908, the rope shovel 100 is in a kinematic condition that facilitates payload estimation, the microprocessor 200 is directed to block 910. Block 910 directs the microprocessor 200 to calculate a payload estimation value for the current payload. For suspension moment measurements with a loaded dipper 102, the system parameters contain payload inertial and geometric parameters encoded within Π_L:

υ_L=YΠ_L

When the dipper 102 is loaded, a set of parameters may be expressed using a moment difference:

υ_L−υ_NL=Y(Π_L−Π_NL)υ_L−YΠ_NL=Yϵ_Δυ=Yϵ

Where the subscript L represents the loaded dipper condition and the subscript NL represents the unloaded dipper condition. The no-load moment may be estimated by Equation 6, and ϵ=Π_L−Π_NL includes only the unknown payload inertial and geometric terms. The friction parameters are assumed to be payload invariant. The term e may thus be estimated as:

{circumflex over (ϵ)}=Y^†Δυ  Equation 7

where Y^† is the left pseudoinverse of Y. In general, ϵ is represented by some function of the payload parameters ρ_p:

${\rho }_p=\left\{\begin{array}{c}{m}_p\\ m_p{x}_p\\ m_p{y}_p\\ I_p+m_p\left(x_p^2+y_p^2\right)\end{array}\right\}$

Where m_p, x_p, y_p, and I_p correspond to the payload weight, center-of-mass positions (x_p, y_p) with respect to the dipper, and inertia for rotation in the x₁-z₁ plane, respectively. The term ϵ maybe expressed as a linear combination of ρ_p such that it is mapped via a payload Jacobian J_ϵ as a function of constant shovel geometric parameters (BL, BV, CH, θ_CH, etc.):

ϵ=J_ϵρ_p

Such that ρ_p may be obtained from an estimated {circumflex over (ϵ)} as follows:

ρ_p=J_ϵ^†{circumflex over (ϵ)}

Where J₆₈ ^† represents the left pseudoinverse of J_ϵ. Combining this with Equation 7 above, ρ_p and therefore the payload weight may be calculated as follows:

{circumflex over (p)}_p=J_ϵ^†Y_l^†Δυ

The determined payload weight may be displayed on the display 250 in the operator's cab 110 on an ongoing basis to provide the operator with a real-time assessment of a loaded payload weight. The payload weight over time may also be recorded for further analysis to determine operator performance and/or digging effectiveness in bucket fill per load and loading truck fill.

Language of degree used herein, such as the terms “approximately,” “about,” “generally,” and “substantially” as used herein represent a value, amount, or characteristic close to the stated value, amount, or characteristic that still performs a desired function or achieves a desired result. For example, the terms “approximately”, “about”, “generally,” and “substantially” may refer to an amount that is within less than 10% of, within less than 5% of, within less than 1% of, within less than 0.1% of, or within less than 0.01% of the stated value.

While specific embodiments have been described and illustrated, such embodiments should be considered illustrative only and not as limiting the disclosed embodiments.

Claims

1. A method for monitoring a mining shovel having a boom supported by a plurality of suspension cables, the method comprising:

receiving accelerometer signals from a plurality of accelerometers, each accelerometer being mounted on one of the plurality of suspension cables;

processing the accelerometer signals to extract a fundamental frequency associated with vibrations in each suspension cable, the fundamental frequency being proportional to a tension in the suspension cable; and

determining changes in the fundamental frequency as a function of time, the changes facilitating a determination of an operating state of the mining shovel.

2. The method of claim 1 wherein receiving the accelerometer signals comprises receiving accelerometer signals from an accelerometer mounted on the suspension cable at a distance of at least one-third of the length from an end of the suspension cable.

3. The method of claim 1 wherein determining changes in the fundamental frequency comprises detecting changes in fundamental frequency that are indicative of a boom jacking event associated with an excavation being performed by the rope shovel.

4. The method of claim 1 wherein determining changes in the fundamental frequency comprises determining changes in fundamental frequency that are indicative of a potential failure of one of the plurality of suspension cables.

5. The method of claim 4 further comprising determining a tension in each of the plurality of suspension cables based on the respective fundamental frequencies relating the fundamental frequency to tension.

6. The method of claim 1 further comprising estimating forces on components of the mining shovel by:

receiving orientation signals from one or more attitude sensors associated with the components, the orientation signals defining an orientation of the components;

determining a kinematic condition defining the position and orientation of the components based on the orientation signals and kinematic calibration data; and

determining forces acting on the components of the mining shovel based on the kinematic condition, the orientation signals, and the fundamental frequency tension in each of the plurality of suspension cables.

7. The method of claim 6 wherein estimating the forces on components of the mining shovel comprises estimating a weight of a payload in a payload container component of the mining shovel.

8. The method of claim 6 further comprising performing a kinematic calibration to establish the kinematic calibration data by:

controlling the mining shovel to cause one of the components of the shovel to be successively located in each of a plurality of known positions and orientations with respect to the mining shovel; and

processing the orientation signals based on the plurality of known positions and orientations to determine the kinematic calibration data for the mining shovel.

9. The method of claim 6, wherein determining forces acting on the components of the mining shovel further includes performing a dynamic calibration to establish dynamic calibration data by:

causing an unloaded payload container component of the mining shovel to maneuver through a trajectory while receiving the orientation signals and determining the changes in the fundamental frequency; and

determining forces on the components of the mining shovel under unloaded conditions based on the changes in fundamental frequency.

10. The method of claim 9 wherein the trajectory is selected to emulate a digging operation of the mining shovel.

11. A monitoring system for monitoring a mining shovel having a boom supported by a plurality of suspension cables, the system comprising:

an accelerometer being mounted on one of the plurality of suspension cables at a distance of at least one-third of the length from an end of the suspension cable; and

a processor in communication with the accelerometer includes programmable logic to process the accelerometer signals to extract a fundamental frequency associated with vibrations in the suspension cable, the fundamental frequency being proportional to a tension in the suspension cable; and determine changes in the fundamental frequency as a function of time, the changes facilitating a determination of an operating state of the mining shovel.

12. The monitoring system of claim 11, wherein the processor includes programmable logic to detect changes in fundamental frequency that are indicative of a boom jacking event associated with an excavation being performed by the rope shovel.

13. The monitoring system of claim 11, wherein the processor includes programmable logic to detect changes in fundamental frequency that are indicative of a potential failure of one of the plurality of suspension cables.

14. The monitoring system of claims 11, further comprising

one or more attitude sensors associated with the boom and other components of the mining shovel, the one or more attitude sensors sending orientation signals defining an orientation of the components;

the processor having programmable logic to determine a kinematic condition defining the position and orientation of the components based on the orientation signals; and

the processor having programmable logic to determine forces acting on the components of the mining shovel based on the kinematic condition, the orientation signals, and the fundamental frequency tension in each of the plurality of suspension cables.

15. The monitoring system of claim 11, wherein estimating the forces on components of the mining shovel comprises estimating a weight of a payload in a payload container component of the mining shovel.