Abstract: A method and system for estimating wax or hydrate deposits is desirable for the oil industry and important for assuring flow conditions and production, avoiding downtime, and reducing or preventing costly interventions. The method and system disclosed herein use artificial intelligence and machine learning techniques combined with oil well historical operational sensor data and historical operational event records (such as diesel hot flush, slick line, coil tubing, etc.) to build an oil well model. The method and system enable oil well practitioners to test and validate the built model and deploy the model online to estimate and/or detect wax or hydrate deposition status. By using one or more such models in operating an oil well, users can monitor and/or detect the status of wax of hydrate deposits in an oil well and can optimize production, maintenance, and planning for oil wells.

Abstract: Systems and methods for building predictive and prescriptive analytics of wind turbines generate a historical operational dataset by loading historical operational SCADA data of one or more wind turbines. Each sensor measurement is associated with an engineering tag and at least one component of a wind turbine. The system creates one or more performance indicators corresponding to one or more sensor measurements, and applies at least one data clustering algorithm onto the dataset to identify and label normal operation data clusters. The system builds a normal operation model using normal operational data clusters with Efficiency of Wind-To-Power (EWTP) and defines a statistical confidence range around the normal operation model as criterion for monitoring wind turbine performance. As real-time SCADA data is received by the system, the system can detect an anomalous event, and issue an alert notification and prescriptive early-action recommendations to a user, such as a turbine operator, technician or manager.

Abstract: Systems and methods for building predictive and prescriptive analytics of wind turbines generate a historical operational dataset by loading historical operational SCADA data of one or more wind turbines. Each sensor measurement is associated with an engineering tag and at least one component of a wind turbine. The system creates one or more performance indicators corresponding to one or more sensor measurements, and applies at least one data clustering algorithm onto the dataset to identify and label normal operation data clusters. The system builds a normal operation model using normal operational data clusters with Efficiency of Wind-To-Power (EWTP) and defines a statistical confidence range around the normal operation model as criterion for monitoring wind turbine performance. As real-time SCADA data is received by the system, the system can detect an anomalous event, and issue an alert notification and prescriptive early-action recommendations to a user, such as a turbine operator, technician or manager.

Abstract: A method and system for estimating wax or hydrate deposits is desirable for the oil industry and important for assuring flow conditions and production, avoiding downtime, and reducing or preventing costly interventions. The method and system disclosed herein use artificial intelligence and machine learning techniques combined with oil well historical operational sensor data and historical operational event records (such as diesel hot flush, slick line, coil tubing, etc.) to build an oil well model. The method and system enable oil well practitioners to test and validate the built model and deploy the model online to estimate and/or detect wax or hydrate deposition status. By using one or more such models in operating an oil well, users can monitor and/or detect the status of wax of hydrate deposits in an oil well and can optimize production, maintenance, and planning for oil wells.

Abstract: Embodiments are directed to systems that build and deploy inferential models for generating predictions of a plant process. The systems select input variables and an output variable for the plant process. The systems load continuous measurements for the selected input variables. For the selected output variable, the systems load measurements of type: continuous from the subject plant process, intermittent from an online analyzer, or intermittent from lab data. If continuous or analyzer measurements are loaded, the systems build a FIR model with a subspace ID technique using continuous output measurements. From intermittent analyzer measurements, the systems generate continuous output measurements using interpolation. If lab data is loaded, the systems build a hybrid FIR model with subspace ID and PLS techniques, using continuous measurements of a reference variable correlated to the selected output variable.

Abstract: Embodiments are directed to systems that build and deploy inferential models for generating predictions of a plant process. The systems select input variables and an output variable for the plant process. The systems load continuous measurements for the selected input variables. For the selected output variable, the systems load measurements of type: continuous from the subject plant process, intermittent from an online analyzer, or intermittent from lab data. If continuous or analyzer measurements are loaded, the systems build a FIR model with a subspace ID technique using continuous output measurements. From intermittent analyzer measurements, the systems generate continuous output measurements using interpolation. If lab data is loaded, the systems build a hybrid FIR model with subspace ID and PLS techniques, using continuous measurements of a reference variable correlated to the selected output variable.

Abstract: Computer system and methods for optimally controlling the behavior of an industrial process, in accordance with plant operating goals, without requiring a complicated trial and error process. The system and methods enable configuring optimization preference and optimization priority for key manipulated variables (MVs) of the industrial process. The system and methods translate the configured optimization preference and optimization priority for each key MV into prioritized economic objective functions. The system and methods calculate a set of normalized cost factors for use in a given prioritized economic functions based on a model gain matrix of manipulated variables and controlled variables of the industrial process.

Abstract: A system and method of model predictive control executes a model predictive control (MPC) controller of a subject dynamic process (e.g., processing plant) in a configuration mode, identification mode and model adaptation mode. Users input and specify model structure information in the configuration mode, including constraints. Using the specified model structure information in the identification mode, the MPC controller generates linear dynamic models of the subject process. The generated linear dynamic models collectively form a working master model. In model adaptation mode, the MPC controller uses the specified model structure information in a manner that forces control actions based on the formed working master model to closely match real-world behavior of the subject dynamic process. The MPC controller coordinates execution in identification mode and in model adaptation mode to provide adaptive modeling and preserve structural information of the model during a model update.

Abstract: A method, apparatus, and computer program product for increasing closed-loop stability in a MPC controller controlling a process where there are significant uncertainties in the model used by the controller. This invention focuses on the improvement of the robustness of the steady-state target calculation. This is achieved through the use of a user defined robustness factor, which is then used to calculate an economic objective function giveaway tolerance and controlled variable constraint violation tolerance. The calculation engine uses these tolerances to find a solution that minimize the target changes between control cycles and prevent weak direction moves caused by near collinearity in the model. If the controller continues to exhibit large variations in the process, it can slow down the manipulated variable movement to stabilize the process.

Abstract: An integrated multivariable predictive controller (MPC) and tester is disclosed. The invention system provides optimal control and step testing of a multivariable dynamic process using a small amplitude step for model identification purposes, without moving too far from optimal control targets. A tunable parameter specifies the trade-off between optimal process operation and minimum movement of process variables, establishing a middle ground between running a MPC on the Minimum Cost setting and the Minimum Move setting. Exploiting this middle ground, embodiments carry out low amplitude step testing near the optimal steady state solution, such that the data is suitable for modeling purposes. The new system decides when the MPC should run in optimization mode and when it can run in constrained step testing mode. The invention system determines when and how big the superimposed step testing signals can be, such that the temporary optimization give-away is constrained to an acceptable range.

Abstract: Computer system and methods for optimally controlling the behavior of an industrial process, in accordance with plant operating goals, without requiring a complicated trial and error process. The system and methods enable configuring optimization preference and optimization priority for key manipulated variables (MVs) of the industrial process. The system and methods translate the configured optimization preference and optimization priority for each key MV into prioritized economic objective functions. The system and methods calculate a set of normalized cost factors for use in a given prioritized economic functions based on a model gain matrix of manipulated variables and controlled variables of the industrial process.

Abstract: A method, apparatus, and computer program product for increasing closed-loop stability in a MPC controller controlling a process where there are significant uncertainties in the model used by the controller. This invention focuses on the improvement of the robustness of the steady-state target calculation. This is achieved through the use of a user defined robustness factor, which is then used to calculate an economic objective function giveaway tolerance and controlled variable constraint violation tolerance. The calculation engine uses these tolerances to find a solution that minimize the target changes between control cycles and prevent weak direction moves caused by near collinearity in the model. If the controller continues to exhibit large variations in the process, it can slow down the manipulated variable movement to stabilize the process.

Abstract: A computer-based apparatus and method for automated data screening and selection in model identification and model adaptation in multivariable process control is disclosed. Data sample status information, PID control loop associations and internally built MISO (Multi-input, Single-output) predictive models are employed to automatically screen individual time-series of data, and based on various criteria bad data is automatically identified and marked for removal. The resulting plant step test/operational data is also repaired by interpolated replacement values substituted for certain removed bad data that satisfy some conditions. Computer implemented data point interconnection and adjustment techniques are provided to guarantee smooth/continuous replacement values.

Abstract: A system and method of model predictive control executes a model predictive control (MPC) controller of a subject dynamic process (e.g., processing plant) in a configuration mode, identification mode and model adaptation mode. Users input and specify model structure information in the configuration mode, including constraints. Using the specified model structure information in the identification mode, the MCP controller generates linear dynamic models of the subject process. The generated linear dynamic models collectively form a working master model. In model adaptation mode, the MPC controller uses the specified model structure information in a manner that forces control actions based on the formed working master model to closely match real-world behavior of the subject dynamic process. The MPC controller coordinates execution in identification mode and in model adaptation mode to provide adaptive modeling and preserve structural information of the model during a model update.

Abstract: A computer-based apparatus and method for automated data screening and selection in model identification and model adaptation in multivariable process control is disclosed. Data sample status information, PID control loop associations and internally built MISO (Multi-input, Single-output) predictive models are employed to automatically screen individual time-series of data, and based on various criteria bad data is automatically identified and marked for removal. The resulting plant step test/operational data is also repaired by interpolated replacement values substituted for certain removed bad data that satisfy some conditions. Computer implemented data point interconnection and adjustment techniques are provided to guarantee smooth/continuous replacement values.

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: A multivariable process controller controls a chemical, polymer or other physical process. Slow tuning and over-conservative controlled variable values are employed during step testing. While all controlled process variables are within safe limits, only one manipulated variable (MV) at a time is step changed. Several manipulated variables are moved when process variables exceed safe limits to ensure that the controlled process variables return to the safe range, such that suitable MV targets for step testing are able to be automatically discovered within a closed loop control environment. Thus, the step test is able to be conducted mostly unsupervised and/or remotely via a telephone or network connection. A new process perturbation approach simultaneously perturbs multiple or all of the process input variables in such a way that the process responses (process outputs) are maximized, while the process variables are maintained inside its predefined operating constraints.

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: A multivariable process controller controls a chemical, polymer or other physical process. Slow tuning and over-conservative controlled variable values are employed during step testing. While all controlled process variables are within safe limits, only one manipulated variable (MV) at a time is step changed. Several manipulated variables are moved when process variables exceed safe limits to ensure that the controlled process variables return to the safe range, such that suitable MV targets for step testing are able to be automatically discovered within a closed loop control environment. Thus, the step test is able to be conducted mostly unsupervised and/or remotely via a telephone or network connection.

Abstract: A multivariable process controller controls a chemical, polymer or other physical process. Slow tuning and over-conservative controlled variable values are employed during step testing. While all controlled process variables are within safe limits, only one manipulated variable (MV) at a time is step changed. Several manipulated variables are moved when process variables exceed safe limits to ensure that the controlled process variables return to the safe range, such that suitable MV targets for step testing are able to be automatically discovered within a closed loop control environment. Thus, the step test is able to be conducted mostly unsupervised and/or remotely via a telephone or network connection.