Abstract: A control apparatus which controls a controlled system with a transfer function regarded as a secondary delay, comprises an outer loop which performs negative feedback of an output of the controlled system in order to obtain a deviation between the output and a desired value, an inner loop which performs negative feedback of a signal obtained by multiplying a differential value of the output of the controlled system by a first gain to the deviation, and a nonlinear inner feedback loop which uses the differential value of the output of the controlled system and a product obtained by multiplying an absolute value of the deviation e or n-th (n is an integer) power of the absolute value by the second gain to perform positive feedback to the deviation.

Abstract: Failure prediction for complex processes is performed utilizing one or more nonlinear regression models to relate operational variable values measured at two or more times to predicted process metric values and maintenance variable values.

Abstract: Systems and methods of complex process control utilize driving factor identification based on nonlinear regression models and process step optimization. In one embodiment, the invention provides a method for generating a system model for a complex process comprised of nonlinear regression models for two or more select process steps of the process where process steps are selected for inclusion in the system model based on a sensitivity analysis of an initial nonlinear regression model of the process to evaluate driving factors of the process.

Abstract: A method of controlling a multivariable system includes the step of receiving a plurality of sensor signals indicating current conditions of the system and receiving a plurality of commands. The desired dynamic response of the system is then determined based upon the commands and the sensor signals. The problem of controlling the system to achieve the desired dynamic response without violating numerous actuator and physical constraints is then formulated as a quadratic programming problem. By solving the quadratic programming problem, the effector commands are determined and the physical constraints are enforced.

Abstract: A method for facilitating an optimization of a control space in n dimensions is provided. The control space includes a plurality of points, and the method includes removing at least one hyperplane of the control space, and perturbing the remaining points of the control space to lessen errors relative to the control space prior to the removal of at least one hyperplane.

Abstract: In the NH3 injection rate control method for an NOx removal apparatus, operating area of the apparatus is divided into first area in which molar ratio of NH3 injection rate with respect to NOx flow rate at an inlet of the apparatus is smaller than molar ratio at minimum point where NOx concentration assumes minimum value and second area in which the molar ratio is equal to or larger than the minimum point molar ratio, and virtual NOx concentration with respect to the molar ratio is set according to virtual characteristic line which varies monotonically to stride across desired operating point without rising with increase of the molar ratio from the first area to the second area. Moreover, feedback control is implemented for adjusting the NH3 injection rate with respect to the NOx flow rate in direction that the virtual NOx concentration is brought close to the desired NOx concentration.

Abstract: There are provided a control apparatus capable of controlling a controlled object with reduce dead time in sign inversion between the input to and the output from one of ?? and ?? modulation algorithms, thereby attaining improved convergence of the output from the controlled object to a target value and improved controllability. An ECU of the control apparatus calculates a limiting value DSMVO2L of a reference input DSMVO2, as DSMVO2L=?1 when DSMVO2<?1, DSMVO2L=1 when 1<DSMVO2, and DSMVO2L=DSMVO2 in the other cases. By inputting the limiting value DSMVO2L to a ? ? modulation algorithm, a DSM output is calculated, according to which the air-fuel ratio of a mixture supplied to an internal combustion engine is controlled such that output from an oxygen concentration sensor converges to a target value.

Abstract: A system and method for performing trajectory generation, interpolation, and control for a motion control application, where the trajectory generation, interpolation, and control are performed in parallel with each other. In one embodiment, an FPGA in a motion control device may be configured to perform the trajectory generation, interpolation, and control in parallel. Performing trajectory generation, interpolation, and control in parallel on an FPGA may increase the efficiency of the motion control application.

Abstract: A method for predicting transient response of a closed loop apparatus includes the steps of: (a) providing a first reference tool that relates load-free impedance response with a first design gain-phase variable; (b) providing a second reference tool that relates load-free impedance response with a second design gain-phase variable; (c) determining a combined impedance response as a function of frequency; (d) employing at least one of the first and second reference tool to establish a first design value for one of the phase variable and the design load impedance at a characteristic frequency that occurs at a peak value of the combined impedance response; (e) employing at least one of the first and second reference tool to establish a second design value for the other parameter of the phase variable and the design load impedance at the characteristic frequency; (f) establishing a transient multiplier as a function of frequency associated with the output voltage with the design load impedance for selected value

Abstract: A synchronous bi-directional active power conditioning system (11) suitable for wide variable frequency systems or active loads such as adjustable speed drives which require variable voltage variable frequency power management systems is disclosed. Common power electronics building blocks (100, 200) (both hardware and software modular blocks) are presented which can be used for AC-DC, DC-AC individually or cascaded together for AC-DC-AC power conversion suitable for variable voltage and/or wide variable frequency power management systems. A common control software building block (2, 5) includes a digital control strategy/algorithm and digital phase lock loop method and apparatus which are developed and implemented in a digital environment to provide gating patterns for the switching elements (3, 6) of the common power-pass modular power electronics building blocks (100, 200).

Abstract: A method and a system for controlling a force feedback member able to interact with another member, including a local model for calculating a set point addressed to the force feedback member from a plurality of variables, and a remote model for estimating interactions and variables of the other member, with updating on receiving data from another remote system, and resynchronizer means able to send a resynchronization message to the other system.

Abstract: A process is modeled by a dynamic model, handling time dependent relations between manipulated variables of different process sections (10A-D) and measured process output variables. Suggested input trajectories for manipulated variables for a subsequent time period are obtained by optimizing an objective function over a prediction time period, under constraints imposed by the dynamic process model and/or preferably a production plan for the same period. The objective function comprises relations involving predictions of controlled process output variables as a function of time using the process model, based on the present measurements, preferably by a state estimation procedure. By the use of a prediction horizon, also planned future operational changes can be prepared for, reducing any induced fluctuations. In pulp and paper processes, process output variables associated with chemical additives can be used, adapting the optimization to handle chemical additives aspects.

Abstract: Loads 1, 2, . . . , N are controlled by predicting a total energy requirement for the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K, by allocating the total energy requirement to the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K, by determining a dynamic energy demand requirement for each of the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K based on the allocated energy requirements, and by controlling the loads 1, 2, . . . , N based on the dynamic energy demand requirements.

Abstract: A system and method is provided for predictive modeling of technical and non-technical components in a business infrastructure that implementing one or more business solutions. According to a first aspect of the invention, performance metrics generated from a predictive model of a business infrastructure are translated into enterprise decision or indicators that correspond to the service, performance and financial states of a business enterprise. As a result, non-technical executives can utilize the enterprise decision metrics or indicators to evaluate, support, and monitor the effect of business decisions, for example, with respect to profitability, productivity, growth, and risk of the business. According to a second aspect of the invention, the accuracy of the predictive modeling is improved by mathematically expressing the dynamic characteristics and behavior of each infrastructure component as a result of direct and indirect effects of the infrastructure components impacting one another.

Abstract: Systematic methods to detect, verify, and repair a collinear model are presented. After detecting collinearity in a model or subsets of a model, a directional test is carried out to verify if the collinearity is real. The model can then be adjusted in either direction to making a near collinear model exactly collinear or less collinear, subject to model uncertainty bounds or other linear constraints. When doing the model adjustment, deviations from the original model are minimized and the directionality of the model is kept unchanged.

Abstract: The invention relates to a method for context-dependent user input prediction for at least one device in the environment of at least one user comprising the steps of: detecting the parameters of the actual states of said at least one device, comparing said detected parameters with previously saved parameters of said at least one device, obtaining a most probable next input proposal according to the result of the comparison, and offering said input proposal to at least one user.

Abstract: A controlling device of high follow-up accuracy is provided, the prediction accuracy of which is not degraded by an FF signal when the controlled object is feedforward-controlled. The prediction controlling device 1 has a feedforward signal generation command filter 2 for generating a future command increment which is an increment from one sampling period to the next sampling period of a target command signal from the present to a multiple-sampling future and the feedforward signal from the future target command signal. A prediction controller 3 receives the future command increment, the feedforward signal, and a controlled object output at the past sampling time over zero sampling, and determines the future error prediction value by using a transfer function model from the feedforward signal and the control input to the controlled object output.

Abstract: A system and method for controlling a controlled parameter that affects a target parameter of a target zone is disclosed. The method comprises providing a feedback control loop having a switching controller, a controlled device, and an averaging device. The controlled device comprises a time constant and a specified operational characteristic. The controlled device comprises a first operational state and a second operational state. The method further comprises calculating a time constant for the averaging device based at least on the time constant for the controlled device, and the specified operational characteristic. The specified operational characteristic may comprises a minimum amount of time that the controlled device operates before it can be switched between the first operational state and the second operational state.

Abstract: A method and an apparatus for dynamic targeting for a process control system. Inline parameter data relating to a processed workpiece is received. A determination is made whether the inline parameter would result in a value of a device operation parameter within a predetermined range. At least one process operation performed upon the workpiece is adjusted in response to a determination that the inline parameter would not result in a value of the device operation parameter.

Abstract: In a rotor assembly having a rotor supported for rotation by magnetic bearings, a processor controlled by software or firmware controls the generation of force vectors that position the rotor relative to its bearings in a “bounce” mode in which the rotor axis is displaced from the principal axis defined between the bearings and a “tilt” mode in which the rotor axis is tilted or inclined relative to the principal axis. Waveform driven perturbations are introduced to generate force vectors that excite the rotor in either the “bounce” or “tilt” modes.

Abstract: A method and apparatus for controlling a non-linear mill. A linear controller is provided having a linear gain k that is operable to receive inputs representing measured variables of the plant and predict on an output of the linear controller predicted control values for manipulatible variables that control the plant. A non-linear model of the plant is provided for storing a representation of the plant over a trained region of the operating input space and having a steady-state gain K associated therewith. The gain k of the linear model is adjusted with the gain K of the non-linear model in accordance with a predetermined relationship as the measured variables change the operating region of the input space at which the linear controller is predicting the values for the manipulatible variables.

Abstract: An instruction value Duty for bringing a slip rotation speed Nslip to a target rotation speed is feedback-controlled based on a difference e between the slip rotation speed Nslip and the target rotation speed, and a weight &thgr; from a parameter map. Weights &thgr; are assigned individually to a plurality of models, each of which includes a group of parameters and which are used to form a control model that represents a slip control system. Based on a weight that is assigned to one of the plurality of models, a weight that is assigned to at least one model that is other than the one of the plurality of models is specified. Thus, the amount of calculation required can be reduced by estimating weights for the models including the groups of parameters, which contain parameters that are used to construct the control model, instead of directly estimating the control model-constructing parameters.

Abstract: A method for generating a predictor of failure of a manufacturing process. The predictor is generated by generating a candidate solution for the predictor and then determining a fitness of the candidate solution using fitness cases pertaining to the manufacturing process and a set of costs associated with the manufacturing process. The fitness may then be used to formulate a next generation of candidate solutions for a design method modeled on evolution.

Abstract: A method for estimating a state of a system including determining, during a current cycle of the system, an estimated state of the system based on currently measured values of the system, boundary conditions of the system, an accuracy of the currently measured values, and an estimated state from a preceding cycle. The method also determines an accuracy of the estimated state, and determines an uncertainty of the estimated state based on the estimated state and the accuracy of the estimated state.

Abstract: A method for designing high performance products incorporating signal processing and feedback control is disclosed. In one embodiment, a block diagram may be used for a design cycle, for design optimization, or for design estimation. The block diagram contains a set of differential equations or difference equations, and the solution of these sets of equations may be performed by commercially available software tools. In order to utilize the software tools without requiring access to source code or other descriptions of the internal structure of the tools, the system is decomposed using the technique of waveform relaxation. The decomposition using waveform relaxation operates directly to speed up the computations for the block diagram system. The remaining interprocessor communications may be held pending until the end of each iteration's calculations in each block, allowing the software tools to be executed on independent multiple processors.

Abstract: In a method and apparatus for predicting process results of objects being processed in a processing chamber of a processing apparatus, operation data and process result data obtained at the time of processing each of the objects are collected and a multivariate analysis is performed on the basis of the collected operation data and process result data to obtain a first correlation therebetween. Process results are predicted on the basis of the first correlation. Weighting coefficients for the respective operation data are set on the basis of the predicted process results. A second correlation between the weighted operation data and the process result data is obtained by a multivariate analysis. Process results are predicted by using operation data, obtained when objects other than the objects used to obtain the second correlation are processed, on the basis of the second correlation.

Abstract: A method for predicting stability of a closed loop apparatus is disclosed. The closed loop apparatus has an open loop impedance and at least one inherent internal gain. The method comprises the steps of: (a) identifying an impedance scaling factor associated with the closed loop apparatus that may be expressed in terms including the open loop impedance, the at least one inherent internal gain, a gain variable and a phase variable; (b) vectorally establishing a first scaling value for the impedance scaling factor as a function of frequency while maintaining a first variable of the gain variable and the phase variable at a first working value to record the first scaling value for a plurality of frequencies.

Abstract: A method for monitoring a manufacturing system includes defining a plurality of observed states associated with the manufacturing system. State estimates are generated for the observed states. An uncertainty value is generated for each of the state estimates. Measurement data associated with an entity in the manufacturing system is received. The state estimates are updated based on the measurement data and the uncertainty values associated with the state estimates. A system for monitoring a manufacturing system includes a controller configured to define a plurality of observed states associated with the manufacturing system, generate state estimates for the observed states, generate an uncertainty value for each of the state estimates, receive measurement data associated with an entity in the manufacturing system, and update the state estimates based on the measurement data and the uncertainty values associated with the state estimates.

Abstract: A filtering process is adapted for eliminating the need of prediscretizing a continuous-time differential model into a discrete-time difference model. It provides a universal robust solution to the most general formulation, in the sense that the system dynamics are described by nonlinear continuous-time differential equations, and the nonlinear measurements are taken at intermittent discrete times randomly spaced. The filtering process includes the procedures of validating the measurement using fuzzy logic, and incorporating factorized forward filtering and backward smoothing to guarantee numerical stability. It provides users a reliable and convenient solution to extracting internal dynamic system state estimates from noisy measurements, with wider applications, better accuracy, better stability, easier design, and easier implementation.

Abstract: A new method is used to model the class probability from data that is based on a novel multiplicative adjustment of the class probability by a plurality of items of evidence induced from training data. The optimal adjustment factors from each item of evidence can be determined by several techniques, a preferred embodiment thereof being the method of maximum likelihood. The evidence induced from the data can be any function of the feature variables, the simplest of which are the individual feature variables themselves. The adjustment factor of an item of evidence E1 is given by the ratio of the conditional probability P(C|E1) of the class C given E1 to the prior class probability P(C), exponentiated by a parameter a1. The method provides a new and useful way to aggregate probabilistic evidence so that the final model Output exhibits a low error rate for classification, and also gives a superior lift curve when distinguishing between any one class and the remaining classes.

Abstract: A feedforward control method for an absorption chiller includes determining the disturbance transfer function, determining the capacity valve transfer function, measuring the actual disturbance, and implementing the feedforward control function in a feedforward controller. The feedforward control function is represented by the ratio of the disturbance transfer function divided by the capacity valve transfer function. The disturbance transfer function and the capacity valve transfer function are measured by applying a known amplitude input perturbation to the disturbance or capacity valve and recording the resulting perturbation in the output leaving chilled water temp. The disturbance transfer function is then the ratio of the delta leaving chilled water temperature divided by the delta change in the disturbance. The capacity transfer function is the ratio of the delta leaving chilled water temperature divided by the delta change in the capacity valve.

Abstract: A method for providing independent static and dynamic models in a prediction, control and optimization environment utilizes an independent static model (20) and an independent dynamic model (22). The static model (20) is a rigorous predictive model that is trained over a wide range of data, whereas the dynamic model (22) is trained over a narrow range of data. The gain K of the static model (20) is utilized to scale the gain k of the dynamic model (22). The forced dynamic portion of the model (22) referred to as the bi variables are scaled by the ratio of the gains K and k. The bi have a direct effect on the gain of a dynamic model (22). This is facilitated by a coefficient modification block (40). Thereafter, the difference between the new value input to the static model (20) and the prior steady-state value is utilized as an input to the dynamic model (22). The predicted dynamic output is then summed with the previous steady-state value to provide a predicted value Y.

Abstract: A method for scheduling activities in a manufacturing system includes defining a plurality of observed states associated with the manufacturing system. State estimates are generated for the observed states. Uncertainty values for the state estimates are generated. A plurality of candidate schedules for performing activities in the manufacturing system is identified. Changes to the uncertainty values are predicted based on the candidate schedules. One of the candidate schedules is selected based on the predicted changes to the uncertainty values.

Abstract: Loads 1, 2, . . . , N are controlled by predicting a total energy requirement for the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K, by allocating the total energy requirement to the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K, by determining a dynamic energy demand requirement for each of the loads 1, 2, . . . , N at prediction points k=0, 1, 2, . . . , K based on the allocated energy requirements, and by controlling the loads 1, 2, . . . , N based on the dynamic energy demand requirements.

Abstract: An integrated sensor processing cell device capable of transforming, reshaping, and modulating an original sensed image includes a sensing medium. At least one memory device stores weight bits. Multiplexers are associated with at least one of the memory devices. At least one transconductance amplifier is associated with at least one of the multiplexers. A multiple input logic gate, associated with at least one of the memory devices, is configured to store a signed pixel output derived from the sensing medium output.

Abstract: The present invention includes a method and system for limiting the ramp rate of a variable under control in order to control stress in a process or plant under control. The present invention takes a limit curve provided by the user and combines that limit curve with a scale factor curve obtained from a model of the system to produce an allowable limit curve. A ramp rate limiter then uses the allowable limit curve to control the ramp rate of the variable under control such that the ramp rate of the variable under control is able to achieve the maximum allowable limit but no more.

Abstract: A device characteristic changing apparatus is provided for changing a characteristic of a device. The device characteristic changing apparatus comprises a time scale generator unit for generating a time scale on which a device operates, and a time scale changing unit for changing the time scale, so that a characteristic of a device is changed by changing the time scale on which the device operates.

Abstract: A model predictive controller having a set of inputs adapted to receive process output signals and further having a set of outputs adapted to supply process input signals to process control devices is configured to perform integrated optimal model predictive control. The controller provides integrated optimal control using a feedback path that couples a selected output to an input having a setpoint that is equal to the constraint limit of the selected output. The controller maintains the selected output at its constraint limit until one of the outputs, other than the selected output, is going to reach or exceed the associated constraint limit for that output. When the output is going to reach or exceed the associated constraint limit for that output, the controller relaxes the setpoint until the output is no longer going to reach or exceed the associated constraint limit for that output.

Abstract: The invention relates to a method for pre-calculating the parameters of industrial processes. According to said method, an admissible process input variable vector is determined with definition ranges allocated to each variable and a process output variable vector is determined with the pre-calculable process parameters. Known information on the process is stored in a data bank and ranges of validity for the process input variables are allocated to said information. For each process input vector inputted from an admissible definition range provided with valid information, exactly one process output vector is determined according to the information valid therefor.

Abstract: A method and apparatus for steady-state target calculation that explicitly accounts for model uncertainty is disclosed. In accordance with one aspect of the invention, when model uncertainty is incorporated, the linear program associated with the steady-state target calculation can be recast as a highly structured nonlinear program. In accordance with another aspect of the invention, primal-dual interior point methods can be applied to take advantage of the resulting special structure. For a system having characteristic gain parameters G having a known uncertainty description, the present invention provides a method and apparatus for selecting steady-state targets for said system-manipulated variables such that all system-controlled variables will remain feasible at steady-state for all possible values of the parameters G within the known uncertainty description.

Abstract: A system of case-based reasoning for sensor prediction in a technical process, especially in a cement kiln, method and apparatus therefore, and wherein the system provides accurate predictions of the cement kiln behavior for a limited period into the future. The invention utilizes a method of case-base-reasoning (CBR) for the task of sensor value prediction. An apparatus for implementing the method is characterized by a database (100) in which the relevant time interval data are stored, a test generator (101) and an optimisation unit (102-104) for the selection and optimisation of the time interval data.

Abstract: The present invention discloses the use of ion implant recipe changes to control the effective channel length by compensating for any variation in the gate electrode width. The invention provides a method for controlling the effective channel length in FETs by measuring the gate electrode width, sending the measured gate electrode width to an ion implant controller, calculating a desired ion implant condition which compensates for any deviation in the effective channel length from target, and subsequently selecting or generating an ion implant recipe based on the desired conditions. Such ion implant recipe is then implanted into the FET to control the effective channel length by defining the halo, LDD, source, drain, or any other doped regions of the device which define the effective channel length, thereby resulting in a manufacturing process with higher yields and less scrap.

Abstract: A flexible closed-loop controller capable of being quickly and easily configured for a broad range of applications. The controller contains a set of control elements that may be configured or reconfigured for any specific application. Built-in configuration management automates configuration of the control elements dependent on user commands or operating conditions. The controller is autonomous after initial configuration, although it may be used interactively. This flexible closed-loop controller is suitable for switching power supplies, linear amplifiers, AC inverters, battery chargers, electronic loads, temperature and pressure management, motor or actuator drives, as well as industrial automation, data acquisition, and automatic test equipment. The controller reduces development time and cost, reduces parts count and cost, and improves performance and reliability of real-time closed-loop systems.

Abstract: This invention unifies a set of statistical signal processing, neuromorphic systems, and microelectronic implementation techniques for blind separation and recovery of mixed signals. A set of architectures, frameworks, algorithms, and devices for separating, discriminating, and recovering original signal sources by processing a set of received mixtures and functions of said signals are described. The adaptation inherent in the referenced architectures, frameworks, algorithms, and devices is based on processing of the received, measured, recorded or otherwise stored signals or functions thereof. There are multiple criteria that can be used alone or in conjunction with other criteria for achieving the separation and recovery of the original signal content from the signal mixtures. The composition adopts both discrete-time and continuous-time formulations with a view towards implementations in the digital as well as the analog domains of microelectronic circuits.

Abstract: A control system for a throttle valve actuating device is disclosed. The throttle valve actuating device includes a throttle valve of an internal combustion engine and an actuator for actuating the throttle valve. The control system includes a predictor for predicting a future throttle valve opening and controls the throttle actuating device according to the throttle valve opening predicted by the predictor so that the throttle valve opening coincides with a target opening.

Abstract: An adaptive predictive expert control system for controlling single-input single-output, or multi-variable time-variant processes with known or unknown parameters and with or without time delay, is disclosed. The adaptive predictive expert control system of the present invention adds an expert block into the operation of previously known adaptive predictive control systems. This expert block based on rules and in the evolution of the process variables, determines and/or modifies the operation of the driver block, control block and adaptive feedback mechanism of the previous art, in order to improve the performance, robustness and stability of the overall control system.

Abstract: A computer based method of controlling an industrial process, including the steps of measuring the values of at least one process variable (y1, y2), and predicting future deviations of one process variable (y1) with regard to the measured value of said at least one process variable (y1, y2), and providing a control signal (uFB) based on said predictions, and by means of a first control law. The method comprises the further steps of measuring a measurable disturbance (d) in the process, predicting future deviations of said process variable (y1) with regard to said disturbance but without regard to the measured value of said at least one process variable (y1, y2), and providing a control signal (ud) based on said predictions by means of a second control law.

Abstract: A first principles, steady state model of a desired polymer process is applied with a non-linear optimizer to a linear controller. Model process gains and optimal target values for controller variables result. These results are utilized by a multivariable linear controller to achieve nonlinear control of the subject process. Preferably the nonlinear optimizer is DMO/SQP. The steady state model is produced by Polymers Plus and the linear controller is DMCplus, all of Aspen Technology, Inc. in Cambridge Mass.

Abstract: A multivariable analysis tool is provided wherein an orthogonal decomposition method such as a Partial Least Squares algorithm is applied to a disturbance model, which relates the known loop disturbances to the loop model residue. The tool according to the invention first extracts the most dominant correlation to the loop model residue and then uses the residue to search for secondary dominant correlation in an orthogonal space, and repeats this process until no further output variation can be significantly attributed by the next dominant correlation. In this way, the analysis tool of the present invention is able to estimate the performance potential of each control loop under different disturbance conditions and provide a control performance index by comparing the achieved performance to the performance potential in a multi-variable environment. This index indicates whether or not a given loop is under optimal operation and shows variance of the loop from the best achievable loop performance.

Abstract: A kiln thermal and combustion control. A predictive model is provided of the dynamics of selected aspects of the operation of the plant for modeling the dynamics thereof. The model has at least two discrete models associated therewith that model at least two of the selected aspects, the at least two discrete models having different dynamic responses. An optimizer receives desired values for the selected aspects of the operation of the plant modeled by the model and optimizes the inputs to the model to minimize error between the predicted and desired values. A control input device then applies the optimized input values to the plant after optimization thereof.