593 results about "Predictive controller" patented technology

An MPC adaptation and tuning technique integrates feedback control performance better than methods commonly used today in MPC type controllers, resulting in an MPC adaptation / tuning technique that performs better than traditional MPC techniques in the presence of process model mismatch. The MPC controller performance is enhanced by adding a controller adaptation / tuning unit to an MPC controller, which adaptation / tuning unit implements an optimization routine to determine the best or most optimal set of controller design and / or tuning parameters to use within the MPC controller during on-line process control in the presence of a specific amount of model mismatch or a range of model mismatch. The adaptation / tuning unit determines one or more MPC controller tuning and design parameters, including for example, an MPC form, penalty factors for either or both of an MPC controller and an observer and a controller model for use in the MPC controller, based on a previously determined process model and either a known or an expected process model mismatch or process model mismatch range. A closed loop adaptation cycle may be implemented by performing an autocorrelation analysis on the prediction error or the control error to determine when significant process model mismatch exists or to determine an increase or a decrease in process model mismatch over time.

A methodology for process modeling and control and the software system implementation of this methodology, which includes a rigorous, nonlinear process simulation model, the generation of appropriate linear models derived from the rigorous model, and an adaptive, linear model predictive controller (MPC) that utilizes the derived linear models. A state space, multivariable, model predictive controller (MPC) is the preferred choice for the MPC since the nonlinear simulation model is analytically translated into a set of linear state equations and thus simplifies the translation of the linearized simulation equations to the modeling format required by the controller. Various other MPC modeling forms such as transfer functions, impulse response coefficients, and step response coefficients may also be used. The methodology is very general in that any model predictive controller using one of the above modeling forms can be used as the controller. The methodology also includes various modules that improve reliability and performance. For example, there is a data pretreatment module used to pre-process the plant measurements for gross error detection. A data reconciliation and parameter estimation module is then used to correct for instrumentation errors and to adjust model parameters based on current operating conditions. The full-order state space model can be reduced by the order reduction module to obtain fewer states for the controller model. Automated MPC tuning is also provided to improve control performance.

An MPC adaptation and tuning technique integrates feedback control performance better than methods commonly used today in MPC type controllers, resulting in an MPC adaptation / tuning technique that performs better than traditional MPC techniques in the presence of process model mismatch. The MPC controller performance is enhanced by adding a controller adaptation / tuning unit to an MPC controller, which adaptation / tuning unit implements an optimization routine to determine the best or most optimal set of controller design and / or tuning parameters to use within the MPC controller during on-line process control in the presence of a specific amount of model mismatch or a range of model mismatch. The adaptation / tuning unit determines one or more MPC controller tuning and design parameters, including for example, an MPC form, penalty factors for either or both of an MPC controller and an observer and a controller model for use in the MPC controller, based on a previously determined process model and either a known or an expected process model mismatch or process model mismatch range. A closed loop adaptation cycle may be implemented by performing an autocorrelation analysis on the prediction error or the control error to determine when significant process model mismatch exists or to determine an increase or a decrease in process model mismatch over time.

A control system simultaneously controls a multi-zone process with a self-adaptive model predictive controller (MPC), such as temperature control within a plastic injection molding system. The controller is initialized with basic system information. A pre-identification procedure determines a suggested system sampling rate, delays or “dead times” for each zone and initial system model matrix coefficients necessary for operation of the control predictions. The recursive least squares based system model update, control variable predictions and calculations of the control horizon values are preferably executed in real time by using matrix calculation basic functions implemented and optimized for being used in a S7 environment by a Siemens PLC. The number of predictions and the horizon of the control steps required to achieve the setpoint are significantly high to achieve smooth and robust control. Several matrix calculations, including an inverse matrix procedure performed at each sample pulse and for each individual zone determine the MPC gain matrices needed to bring the system with minimum control effort and variations to the final setpoint. Corrective signals, based on the predictive model and the minimization criteria explained above, are issued to adjust system heating / cooling outputs at the next sample time occurrence, so as to bring the system to the desired set point. The process is repeated continuously at each sample pulse.

An MPC Control system provides a life extending control that includes life-extending goals in the performance index of the MPC controller and limits in the inequality equations. The MPC controller performs the normal functions of a control system for a physical system, but does so in a manner that extends the life or time-to-next maintenance or reduces the number of parts that need to be replaced. If the life extending functions do not degrade other control functions, they can be always enabled, making the system less expensive to maintain. If the life extending functions degrade some other control functions, they can be adjusted in-the-field or on-the-fly to stretch the time-until-maintenance until it is more convenient, but with some impact on performance.

An MPC controller technique integrates feedback control performance better than methods commonly used today in MPC type controllers, resulting in an MPC controller that performs better than traditional MPC techniques in the presence of process model mismatch. In particular, MPC controller performance is enhanced by adding a tunable integration block to the MPC controller that develops an integral component indicative of the prediction or other control error, and adds this component to the output of an MPC controller algorithm to provide for faster or better control in the presence of model mismatch, which is the ultimate reason for the prediction error in the first place. This technique enables the MPC controller to react more quickly and to provide better set point change and load disturbance performance in the presence of model mismatch, without decreasing the robustness of the MPC controller.

An MPC controller technique integrates feedback control performance better than methods commonly used today in MPC type controllers, resulting in an MPC controller that performs better than traditional MPC techniques in the presence of process model mismatch. In particular, MPC controller performance is enhanced by adding a tunable integration block to the MPC controller that develops an integral component indicative of the prediction or other control error, and adds this component to the output of an MPC controller algorithm to provide for faster or better control in the presence of model mismatch, which is the ultimate reason for the prediction error in the first place. This technique enables the MPC controller to react more quickly and to provide better set point change and load disturbance performance in the presence of model mismatch, without decreasing the robustness of the MPC controller.

A model predictive controller (MPC) for predictive control of nonlinear processes utilizing an EKF (Extended Kalman Filter) and a nominal trajectory generator. The nominal trajectory generator includes another instance of EKF, a linear corrector, and a time-varying deviation model. A nominal control trajectory can be predicted and an optimal solution for the time-varying deviation model can be computed based on an approximation of a system inverse known as signal de-convolution. The EKF can be utilized to estimate a current process state by supplying a measured output and to predict a future nominal trajectory by supplying a reference output. A Kalman smoother can also be utilized for the signal de-convolution in order to obtain enhanced trajectory estimates.

The invention relates to a current predictive control method of a permanent magnet synchronous motor, which belongs to the electric control field. The dynamic response speed and the control accuracy for the stator current control of the permanent magnet synchronous motor are improved through stator current prediction and deadbeat control, and the system delaying is compensated, so that the noise and the torque ripple of the motor in operation are reduced. The method comprises the steps as follows: obtaining a three phase stator current signal, and the electrical angle and the electrical angular speed of the motor rotor through the technologies of sensor sampling and a photoelectric coded disc or position sensorless detection; carrying out Clarke transformation and Park transformation on the stator current signal to obtain the stator current in a synchronous revolution dq coordinate system; substituting the obtained stator current signal in the dq coordinate system into a control equation of a current predictive controller of the permanent magnet synchronous motor, carrying out deadbeat control according to the given value of the obtained stator current in the dq coordinate system in an outer ring controller to obtain a stator voltage vector in the dq coordinate system; carrying out Park inverse transformation on the obtained stator voltage vector in the dq coordinate system, to obtain a pulse-width modulation (PWM) control signal of an inverter by a space vector pulse width modulation (SVPWM) method, controlling the stator current through the inverter, and then implementing the current predictive control over the permanent magnet synchronous motor.

An automated tuning method of a large-scale multivariable model predictive controller for multiple array papermaking machine cross-directional (CD) processes can significantly improve the performance of the controller over traditional controllers. Paper machine CD processes are large-scale spatially-distributed dynamical systems. Due to these systems' (almost) spatially invariant nature, the closed-loop transfer functions are approximated by transfer matrices with rectangular circulant matrix blocks, whose input and output singular vectors are the Fourier components of dimension equivalent to either number of actuators or number of measurements. This approximation enables the model predictive controller for these systems to be tuned by a numerical search over optimization weights in order to shape the closed-loop transfer functions in the two-dimensional frequency domain for performance and robustness. A novel scaling method is used for scaling the inputs and outputs of the multivariable system in the spatial frequency domain.

A method for controlling a non-linear process includes the steps of first providing a controller that is operable to receive inputs representing measured variables of the process and predicting on an output of the controller predicted control values for manipulatible variables that control the process. An expert system is provided that models the actions of an operator of the process over an operating region of the process that represents a set of rules for actions to be taken by an operator upon the occurrence of predetermined conditions in the operation of the process. The operation of the controller is modified with the expert system when one of the predetermined conditions exists.

An adaptive multivariable process control system includes a multivariable process controller, such as a model predictive controller, having a multivariable process model characterized as a set of two or more single-input, single-output (SISO) models and an adaptation system which adapts the multivariable process model. The adaptation system detects changes in process inputs sufficient to start an adaptation cycle and, when such changes are detected, collects process input and output data needed to perform model adaptation. The adaptation system next determines a subset of the SISO models within the multivariable process model which are to be adapted, based on, for example, a determination of which process inputs are most correlated with the error between the actual (measured) process output and the process output developed by the multivariable process model. The adaptation system then performs standard or known model switching and parameter interpolation techniques to adapt each of the selected SISO models. After the adaptation of one or more of the SISO models, the resulting multivariable process model is validated by determining if the adapted multivariable process model has lower modeling error than the current multivariable process model. If so, the adapted multivariable process model is used in the multivariable controller.

A method of predictive control for a single input, single output (SISO) system, including modeling the SISO system with model factors, detecting output from the SISO system, estimating a filtered disturbance from the output, determining a steady state target state from the filtered disturbance and a steady state target output, populating a dynamic optimization solution table using the model factors and a main tuning parameter, and determining an optimum input from the dynamic optimization solution table. Determining an optimum input includes determining a time varying parameter, determining a potential optimum input from the time varying parameter, and checking whether the potential optimum input is the optimum input.

A technique for fast performance prediction of multivariable model predictive controllers (MPC) for large-scale spatially-distributed dynamic systems is provided. When operating an MPC for a two-dimensional sheetmaking process, the controller's performance can be separated into spatial performance and dynamic performance. The steady-state spatial performance for different sheet property modes is first predicted, thereafter, the dynamic performance is predicted on the basis of the steady-state actuator profiles. During the steady-state spatial performance prediction, the original MPC's cost function is approximated to be a steady-state cost function. Then the steady-state measurement profiles and the actuator profiles are calculated through the steady-state cost function. The actuator profiles in time series can be efficiently solved by minimizing the difference between the actuator profiles and the steady-state actuator profiles in the face of constraints. Finally, the model can be employed to expeditiously compute the measurement profiles in times series.

Described and illustrated is a system for management of diabetes that includes an infusion pump, glucose sensor and controller with a method programmed in the controller. The infusion pump is configured to deliver insulin. The glucose sensor senses glucose levels in the subject and provide output signals representative of the glucose levels in the subject. The controller is programmed receives signals from at least one of the glucose sensor and the pump and configured to issue signals to the pump to deliver an amount of insulin determined by a feedback controller that utilizes a model predictive control based on desired glucose levels, insulin amount delivered and measured glucose levels of the subject. The controller is also configured to deliver insulin using a tuning factor (R) for a model predictive controller in the microcontroller as a conservative setting otherwise the system maintains a current tuning factor (R) for the controller.

A direct metal deposition (DMD) process is stabilized by monitoring the temperature and the shape of the melt pool during deposition, applying a recursive least square (RLS) model estimation algorithm to adaptively identify process characteristics in accordance with the temperature and the shape of the melt pool, and delivering the process characteristics to a generalized predictive controller with input constraints to control the process. The process may be controlled by adjusting laser power or by adjusting the speed of the movement of the laser. In the preferred embodiment the temperature is monitored using a two-color pyrometer, and the shape of the melt pool is monitored by detecting the edge of the melt pool with a camera and / or photodetector.

A fast and reliable technique for tuning multivariable model predictive controllers (MPCs) that accounts for performance and robustness is provided. Specifically, the technique automatically yields tuning weights for the MPC based on performance and robustness requirements. The tuning weights are parameters of closed-loop transfer functions which are directly linked to performance and robustness requirements. Automatically searching the tuning parameters in their proper ranges assures that the controller is optimal and robust. This technique will deliver the traditional requirements of stability, performance and robustness, while at the same time enabling users to design their closed-loop behavior in terms of the physical domain. The method permits the user to favor one measurement over another, or to use one actuator more than another.

A rate based model predictive controller for air path control for a diesel engine regulates turbine lift and EGR valve flow rate to specified set points by coordinated control of intake manifold air pressure and EGR valve flow rate. The controller utilizes a rate-based reduced order linear model for model predictive control.

A method for controlling a non-linear process includes the steps of first providing a controller that is operable to receive inputs representing measured variables of the process and predicting on an output of the controller predicted control values for manipulatible variables that control the process. An expert system is provided that models the actions of an operator of the process over an operating region of the process that represents a set of rules for actions to be taken by an operator upon the occurrence of predetermined conditions in the operation of the process. The operation of the controller is modified with the expert system when one of the predetermined conditions exists.

The invention discloses a predictive control method and system based on a multi-model generalized predictive controller. While process disturbance is inhibited, a preset desired output value is enabled to track the optimum set value track, and dynamic characteristics of a system are distinguished in parallel by adopting a plurality of fixed models and a plurality of adaptive models so that an actual output value and an optimum input control quantity of the system can be obtained. The invention also provides a predictive control system which is of a DRTO (Dynamic Real-Time Optimization) dual-layer structure, and adopts a multi-mode generalized predictive controller to replace an existing single-model generalized predictive controller. The predictive control method has the following beneficial effects of well matching the actual process characteristic in the production, reducing the system cost consumption, increasing the system economic benefit, improving the system transient state performance and the system regulating capacity when a system model parameter hops, being capable of effectively eliminating the interference of disturbance to system output, and reducing the influence of inconsistency of models of an optimization layer and a control layer in the DRTO dual-layer structure to the economic benefit.

The invention relates to raw material grinding in the field of cement process industries, and aims to provide an advanced control method and system for a vertical mill based on model identification and predictive control. The method comprises the following steps of: acquiring real-time data from a distributed control system (DCS) monitoring system; analyzing a variation trend of the operation and technology parameters, and then invoking a pathological working condition expert database for performing trend matching; if a pathological working condition appears, issuing early warning display and giving qualitative adjustment suggestion remind; giving an optimal target set value according to the basic operation condition of the vertical mill and the variation situation of the product quality requirement, and writing into a predictive controller; setting an optimal controlled quantity output according to the optimal target set value, and outputting to the DCS monitoring system to control a field actuator to take action. By adopting the invention, the qualitative adjustment suggestion can be precisely given; a mathematical model of the grinding process of the vertical mill is established and updated in real time; the steady-state error of the control system is reduced; and the grinding process of the vertical mill is instructed, so that the mill can operate stably for long term at a maximum efficiency point, and stable margin is maintained.

The present invention relates to a steady state optimization method incorporating dynamic variance correction for dynamic variations of both independent variables and dependent variables of a dynamic system. The dynamic variance correction is based on measured variance of the variables and a weighing factor for each of the variables. The dynamic variance correction offers an effective method of dynamic violations avoidance of controlled variables for a model predictive controller without having to constantly adjust the tuning weights in response to changing dynamical conditions.

A model predictive controller (MPC) for predictive control of nonlinear processes utilizing an EKF (Extended Kalman Filter) and a nominal trajectory generator. The nominal trajectory generator includes another instance of EKF, a linear corrector, and a time-varying deviation model. A nominal control trajectory can be predicted and an optimal solution for the time-varying deviation model can be computed based on an approximation of a system inverse known as signal de-convolution. The EKF can be utilized to estimate a current process state by supplying a measured output and to predict a future nominal trajectory by supplying a reference output. A Kalman smoother can also be utilized for the signal de-convolution in order to obtain enhanced trajectory estimates.

Entering a demand prediction for a discrete component product manufactured in a supply chain into a model predictive controller, entering inventory information for an inventory in the supply chain that is associated with the discrete component product into the model predictive controller, and controlling the inventory with the model predictive controller based on the demand prediction is disclosed.

A modular formulation of model predictive controller is proposed by this invention (400). The modular formulation offers a robust model predictive controller with unified method of tuning. Explicit use of material and energy balance provides stable and robust performance (412). Use of economic based tuning offers a unified tuning method (411). The unified tuning method establishes consistency of control actions resulting from the steady state optimization and the dynamic control (400). Further, a method for formulating and tuning integrated optimization and control using multiple modular model predictive controllers is presented in this invention (100). The method incorporates a direct and explicit method of integration of a number of modular model predictive controllers. The integration of the modular controllers is achieved seamlessly without requiring any further tuning adjustments (200). The resulting integrated system of multiple modular predictive controllers performs robustly, permitting each of the modular model predictive controllers to perform its role in the context of the state of operation of rest of the system. The proposed integration can be made to include lower level modular controllers or at same level controllers or both (300).

An aero-engine online optimization and multivariable control design method based on model predictive control achieves the control and online optimization of multiple variables of an aero-engine undera constraint and according to requirements of thrust and speed. A control system consists of two parts. The first part is a prediction model acquisition layer. Based on the actual working state of each control cycle and the external environmental parameters of the aero-engine, an engine small deviation linear model near different steady-state points is continuously established, and model parameters are provided for a model predictive controller. The second part is a control law decision layer. A closed-loop structure is formed by the model predictive controller and an external output feedback.The model predictive controller, based on an engine model in the current state, a control command and relevant constraint limits, determines the output of the controller at the next moment by solvinga linear optimization problem. The external output feedback introduces the aero-engine actual output into the decision on future controlled quantity of the controller to compensate for the influenceof model mismatch and external disturbances.