Performance Of Model Predictive Control Controller Research Articles

MPC-based thermal management for water-cooled proton exchange membrane fuel cells

The proper operating temperature is inextricably linked to the efficient operation of a proton exchange membrane fuel cell (PEMFC). In this study, a model predictive control (MPC) method for PEMFC water cooling system was proposed to control fans and pump. Firstly, the electrochemical model of PEMFC and thermal model of the system were established, and the electrochemical model was verified by experimental data. Then the MPC controller is designed by linearizing and discretizing the thermal model, and a joint simulation model was built through MATLAB/Simulink and GT Suite. Finally, the performance of MPC controller was evaluated under three different operating conditions (step load, dynamic load, and variable target situation), and its effectiveness was compared to that of a rule-based controller. From the co-simulation results, the MPC controller can accurately adjust the speed of the pump and fans, keep the stack temperature within ±0.5 °C of the setpoint, and realize the tracking control of the change target.

A Soft Constrained MPC Formulation Enabling Learning From Trajectories With Constraint Violations

In practical model predictive control (MPC) implementations, constraints on the states are typically softened to ensure feasibility despite unmodeled disturbances. In this work, we propose a soft constrained MPC formulation supporting polytopic terminal sets in half-space and vertex representation, which significantly increases the feasible set while maintaining asymptotic stability in case of constraint violations. The proposed formulation allows for leveraging system trajectories that violate state constraints to iteratively improve the MPC controller's performance. To this end, we apply convex optimization techniques to obtain a data-driven terminal cost and set, which result in a quadratic MPC problem.

Adaptive Linear Grey-Box Models for Model Predictive Controller of Residential Buildings

Model predictive control (MPC) is an advanced optimal control technique to minimize a control objective while satisfying a set of constraints and is well suited to activate the building energy flexibility. The MPC controller performance depends on the accuracy of the model prediction. Inaccurate predictions can directly lead to low control performance. Linear time-invariant (LTI) models are often used in MPC in buildings. However, LTI models do not adapt to the weather conditions varying throughout the whole space-heating season, which makes the MPC based on LTI models not perform well over a long period of time. Therefore, this study introduces an adaptive MPC where the parameters of a linear grey-box model are continuously updated in real-time. Two alternative versions of this adaptive control are investigated. The first one, called partially adaptive MPC, only updates the effective window area of the grey-box model, while the second one, called fully adaptive MPC, updates all the parameters of the grey-box model. Results show that the partially adaptive MPC is not able to deliver satisfactory prediction performance. The fully adaptive MPC shows better performance compared to the other models when implemented in a MPC, especially in avoiding thermal comfort violation.

Application and Performance Evaluation of Model Predictive Controllers on Driving Force Based Reactive Distillation Processes

In this work, integrated design and control of Reactive Distillation (RD) column operating under Model Predictive Controller (MPC) is presented. At first, a design target is defined, and the RD column is designed based on the element concept at the maximum driving force. After that, the steady state simulation is carried out in ASPEN PLUS that satisfies the design target. Next, the flow driven dynamic system is linearized in ASPEN PLUS Dynamics and the state space matrices are obtained. Using the matrices, MPC closed-loop simulation is carried out in MATLAB. To quantify both the steady state and dynamic operation, a multi objective function is defined that takes into account both the design (total energy consumption, total carbon footprint), and the control performance (Integral Absolute Error, total variation of input, relative gain array, Niederlinski Index). For further verification, an alternative design based on non-maximal driving force is also obtained, and the corresponding steady state and dynamic simulation is performed. The closed-loop performance of MPC controller under feed disturbance shows that the process that is designed at the maximum driving force exhibit better control performance compared to the design alternative that does not take into account the maximum available driving force.
 Chemical Engineering Research Bulletin 21(2020) 49-53

Autonomous Collision Avoidance Using MPC with LQR-Based Weight Transformation.

Model predictive control (MPC) is a multi-objective control technique that can handle system constraints. However, the performance of an MPC controller highly relies on a proper prioritization weight for each objective, which highlights the need for a precise weight tuning technique. In this paper, we propose an analytical tuning technique by matching the MPC controller performance with the performance of a linear quadratic regulator (LQR) controller. The proposed methodology derives the transformation of a LQR weighting matrix with a fixed weighting factor using a discrete algebraic Riccati equation (DARE) and designs an MPC controller using the idea of a discrete time linear quadratic tracking problem (LQT) in the presence of constraints. The proposed methodology ensures optimal performance between unconstrained MPC and LQR controllers and provides a sub-optimal solution while the constraints are active during transient operations. The resulting MPC behaves as the discrete time LQR by selecting an appropriate weighting matrix in the MPC control problem and ensures the asymptotic stability of the system. In this paper, the effectiveness of the proposed technique is investigated in the application of a novel vehicle collision avoidance system that is designed in the form of linear inequality constraints within MPC. The simulation results confirm the potency of the proposed MPC control technique in performing a safe, feasible and collision-free path while respecting the inputs, states and collision avoidance constraints.

Optimized Model Predictive Control With Dead-Time Voltage Vector for PMSM Drives

In order to improve the current steady-state control performance of the model predictive control (MPC), a dead-time voltage-vector based MPC method is proposed in this article. First, the effect of the dead-time on MPC is introduced, and the formation process of the dead-time voltage vector existed in MPC is analyzed. Furthermore, the beneficial dead-time voltage vector and nonbeneficial dead-time voltage vector for the current steady-state control performance of MPC is distinguished, and the advantage of the beneficial dead-time voltage vector is analyzed. Then, the MPC method based on dead-time voltage vector is presented, which optimizes the action time of dead-time. Finally, the experimental results prove that the current steady-state control performance of the proposed MPC method is better than the conventional MPC method without increasing the switching frequency.

Practical factors of envelope model setup and their effects on the performance of model predictive control for building heating, ventilating, and air conditioning systems

Model predictive control (MPC) for buildings is attracting significant attention in research and industry due to its potential to address a number of challenges facing the building industry, including energy cost reduction, grid integration, and occupant connectivity. However, the strategy has not yet been implemented at any scale, largely due to the significant effort required to configure and calibrate the model used in the MPC controller. While many studies have focused on methods to expedite model configuration and improve model accuracy, few have studied the impact a wide range of factors have on the accuracy of the resulting model. In addition, few have continued on to analyze these factors’ impact on MPC controller performance in terms of final operating costs. Therefore, this study first identifies the practical factors affecting model setup, specifically focusing on the thermal envelope. The seven that are identified are building design, model structure, model order, data set, data quality, identification algorithm and initial guesses, and software tool-chain. Then, through a large number of trials, it analyzes each factor’s influence on model accuracy, focusing on grey-box models for a single zone building envelope. Finally, this study implements a subset of the models identified with these factor variations in heating, ventilating, and air conditioning MPC controllers, and tests them in simulation of a representative case that aims to optimally cool a single-zone building with time-varying electricity prices. It is found that a difference of up to 20% in cooling cost for the cases studied can occur between the best performing model and the worst performing model. The primary factors attributing to this were model structure and initial parameter guesses during parameter estimation of the model.

Experimental study on control performance comparison between model predictive control and proportion-integral-derivative control for radiant ceiling cooling integrated with underfloor ventilation system

Radiant ceiling cooling integrated with underfloor ventilation (RCCUV) system has energy-saving potential in buildings. However, control problem results in its energy-saving effect lower than expected and restricts its application. Conventional control is inadequate due to its poor performance, control lag and low economy, etc. The objectives of this study are to: (1) develop advanced model predictive control (MPC) controller of RCCUV system; and (2) experimentally discuss MPC controller performance compared with conventional proportional-integral-derivative (PID) controller performance. The experimental results indicated that under experimental step setpoint variations, the adjusting time of controlled variables using MPC controller was less than that of using PID controller. The time to reach recommended thermal comfort range using MPC controller was 3 min, and was 13 min using PID controller under experimental step setpoint variations. Meanwhile, MPC controller reduced energy consumption of RCCUV system by 13.2% compared with PID controller during 2 h of experimental period. Thus, control performance of MPC controller is superior to that of PID controller in every aspect. The study provides an advanced superior MPC controller for RCCUV system.

Analyzing the Impacts of System Parameters on MPC-Based Frequency Control for a Stand-Alone Microgrid

Model predictive control (MPC) has been widely studied for regulating frequency in stand-alone microgrids (MGs), owing to the advantages of MPC such as fast response and robustness against the parameter uncertainties. Understanding the impacts of system parameters on the control performance of MPC could be useful for the designing process of the controller to achieve better performance. This study analyzes the impact of system parameters on the control performance of MPC for frequency regulation in a stand-alone MG. The typical stand-alone MG, which consists of a diesel engine generator, an energy storage system (ESS), a wind turbine generator, and a load, is considered in this study. The diesel engine generator is in charge of primary frequency control whereas the ESS is responsible for secondary frequency control. The stand-alone MG is linearized to obtain the dynamic model that is used for designing MPC-based secondary frequency control. The robustness of MPC against the variation of system parameters is also analyzed in this study. A comparison study of MPC and proportional–integral (PI) control is presented. Simulation results show that MPC has a faster response time and lower overshoot compared to PI control. In addition, the robustness of MPC against the system uncertainties is stronger than conventional PI control.

Model predictive control of attitude maneuver of a geostationary flexible satellite based on genetic algorithm

This article presents an application of Model Predictive Controller (MPC) to the attitude control of a geostationary flexible satellite. SIMO model has been used for the geostationary satellite, using the Lagrange equations. Flexibility is also included in the modelling equations. The state space equations are expressed in order to simplify the controller. Naturally there is no specific tuning rule to find the best parameters of an MPC controller which fits the desired controller. Being an intelligence method for optimizing problem, Genetic Algorithm has been used for optimizing the performance of MPC controller by tuning the controller parameter due to minimum rise time, settling time, overshoot of the target point of the flexible structure and its mode shape amplitudes to make large attitude maneuvers possible. The model included geosynchronous orbit environment and geostationary satellite parameters. The simulation results of the flexible satellite with attitude maneuver shows the efficiency of proposed optimization method in comparison with LQR optimal controller.

MODELING AND CONTROL OF MULTIVARIABLE DISTILLATION COLUMN USING MODEL PREDICTIVE CONTROL USING UNISIM

Distillation columns are widely used in chemical industry as unit operation and required advance process control because it has multi input multi output (MIMO) or multi-variable system, which is hard to be controlled. Model predictive control (MPC) is one of alternative controller developed for MIMO system due to loops interaction to be controlled. This study aimed to obtain dynamic model of process control on a distillation column using MPC, and to get the optimum performance of MPC controller. Process control in distillation columns performed by simulating the dynamic models of distillation columns by UNISIM R390.1 software. The optimization process was carried out by tuning the MPC controller parameters such as sampling time (Ts = 1 – 240 s), prediction horizon (P = 1-400), and the control horizon (M=1-400). The comparison between the performance of MPC and PI controller is presented and Integral Absolut Error (IAE) was used as comparison parameter. The results indicate that the performance of MPC was better than PI controller for set point change 0.95 to 0.94 on distillate product composition using a modified model 1 with IAE 0.0584 for MPC controller and 0.0782 for PI controller.

MPC Controller Performance Evaluation and Tuning of Single Inverted Pendulum Device

Inverted pendulum is a non-linear, multivariable and unstable device, a model predictive control (MPC) performance evaluation and tuning method for inverted pendulum device is proposed. MPC was designed to control the inverted pendulum device, and the minimum variance covariance constrained control (MVC 3 ) was applied to evaluate the performance of the MPC controller and tune its parameters. The application results to a single inverted pendulum device have verified the feasibility and effectiveness of the proposed method.

Batch distillation control improvement by novel model predictive control

Novel model predictive control (MPC) based on a sequence of reduced-order models is developed for a ternary batch distillation operated in an optimal reflux policy. A Kalman filter (KF) estimator is employed for estimating process outputs in the presence of disturbance and measurement noise. To handle highly nonlinearity and non-stationary by nature of the batch, a series of local models are developed around different parts of the reference profiles, and further reduced their orders individually to attain only observable and controllable state-contributions. The control performance of MPC based on reduced-order models incorporating with KF has been compared with conventional MPC (based on simple model) incorporating with an extended Kalman filter (EKF). Simulation results demonstrate that the proposed control strategy gives good control performance even in a presence of external disturbance and measurement noise. The main advantage of using the reduced-order models is small computation effort and sampling frequency requirement. In addition, the knowledge of thermodynamic is not necessarily required and augmented states can be easily initialized which is applicable in real situation.