Application Of Predictive Control Research Articles

Industrial, large-scale model predictive control with structured neural networks

The design of neural networks (NNs) is presented for treating large, linear model predictive control (MPC) applications that are out of reach with available quadratic programming (QP) solvers. First, we introduce a new feedforward network architecture that enables practitioners to obtain offset-free closed-loop performance with NNs. Second, we discuss the data generation procedure to sample the state space relevant to training the NNs based on anticipated online setpoint changes and plant disturbances. Third, we use the input-to-state stability results available in the MPC literature and establish robustness properties of NN controllers. Finally, we present illustrative simulation studies on process control examples. We apply the NN design approach and compare the performance with online QP based MPC on an industrial crude distillation unit model with 252 states, 32 control inputs, and a control-sample horizon length of 140. Parallel computing is used for data generation and graphical processing units are used for network training. Anticipated plant operational scenarios with setpoints and disturbances that may change during operation must be sampled for NN training. After the offline design phase, NNs execute MPC three to five orders of magnitude faster than an available QP solver with less than 1% loss in the closed-loop performance.

Innovative Application of Model-Based Predictive Control for Low-Voltage Power Distribution Grids with Significant Distributed Generation

In past decades, the deployment of renewable-energy-based power generators, namely solar photovoltaic (PV) power generators, has been projected to cause a number of new difficulties in planning, monitoring, and control of power distribution grids. In this paper, a control scheme for flexible asset management is proposed with the aim of closing the gap between power supply and demand in a suburban low-voltage power distribution grid with significant penetration of solar PV power generation while respecting the different systems’ operational constraints, in addition to the voltage constraints prescribed by the French distribution grid operator (ENEDIS). The premise of the proposed strategy is the use of a model-based predictive control (MPC) scheme. The flexible assets used in the case study are a biogas plant and a water tower. The mixed-integer nonlinear programming (MINLP) setting due to the water tower ON/OFF controller greatly increases the computational complexity of the optimisation problem. Thus, one of the contributions of the paper is a new formulation that solves the MINLP problem as a smooth continuous one without having recourse to relaxation. To determine the most adequate size for the proposed scheme’s sliding window, a sensitivity analysis is carried out. Then, results given by the scheme using the previously determined window size are analysed and compared to two reference strategies based on a relaxed problem formulation: a single optimisation yielding a weekly operation planning and a MPC scheme. The proposed problem formulation proves effective in terms of performance and maintenance of acceptable computational complexity. For the chosen sliding window, the control scheme drives the power supply/demand gap down from the initial one up to 38%.

Eddy current de-tumbling large geostationary debris based on feedback linearization model predictive control

In this paper, we investigate the application of Feedback Linearization Model Predictive Control (FLMPC) for the eddy current de-tumbling of space tumbling targets. Based on the Clohessy-Wiltshire (C-W) equation and the Euler rotational equations of motion, the dynamic model of the eddy current de-tumbling process is established, and the error dynamic model is further derived. After that, the safety constraint of the eddy current de-tumbling process is deduced and linearized. On this basis, real-time optimal trajectory is obtained with the optimal equilibrium solver, and FLMPC is proposed to control the system. In order to ensure the simplicity and rapidity of the controller, the feedback linearization input constraint is directly used as the global prediction constraint and the MPC constraint problem is converted into the form of Quadratic Programming (QP) which can be effectively solved. Simulation results show that the proposed algorithm can realize quick and safe de-tumbling of the target with the various constraints ensured and a very small range of steady-state errors.

Back Cover: Advanced Real‐Time Process Analytics for Multistep Synthesis in Continuous Flow (Angew. Chem. Int. Ed. 15/2021)

Advanced process analytics are becoming an increasingly important part of continuous flow chemistry. In their Research Article on page 8139, Jason D. Williams, C. Oliver Kappe and co-workers describe the use of four different process analytical technologies, controlled by one central system, in a multistep continuous flow process. Advanced data processing methods, including artificial intelligence, are used to accurately quantify up to nine process species in real time, toward self-optimization and model predictive control applications.

Application of Economic Model Predictive Control in Integrated Energy System

Due to the uncertainty of the energy supply and the power demand, the stability and economic performance of the integrated energy system has become a key problem. In this paper, economic model predictive control with augmented model was directly applied to optimize the performance index while responding power demand. Based on the prices of power and hot water, the economic objective function was designed and two modes of operation of heating have been studied which include providing domestic hot water and space heating. The simulation result shows, compared with traditional model predictive control, economic model predictive control could improve economic performance of system by 20% while providing domestic hot water, and showed similar performance while working on space heating.

A Novel Multi-vector Model Predictive Current Control of Three-Phase Active Power Filter

This paper proposes the application of a novel finite control set model predictive control (FCS-MPC) strategy in active power filter (APF). In the process of APF compensating harmonic and reactive power, the traditional single vector model predictive current control (MPCC) has low tracking accuracy to harmonic current, while the multi-vector MPCC has the problems of complex calculation and long calculation time, a new multi-vector MPCC control method has proposed in this paper. Firstly, the harmonic reference value is transformed into d-q coordinate system, according to the sector, the slope is calculated and the action time is obtained. Six new expected vectors are synthesized from six effective vectors and zero vectors. The value function is established to loop and calculate the optimal virtual vector, which is applied to APF. Compared with single vector control and traditional multi-vector control, it has a wider vector action area and faster calculation speed. The compensation results and dynamic performance are improved. The simulation results show that the total harmonic distortion (THD) is low.

Application of feedforward predictive control in DC furnace coordination system

In order to solve the control difficulties such as large dynamic delay, serious coupling between machine and furnace and strong nonlinear, the feedforward and feedback predictive control is proposed to control the coordinated system on the basis of multi-model predictive control, and the static feedforward part designs the feed line feedforward of coal quantity against coal quality disturbance on the basis of the accurate balance of energy. Thus, the reference total fuel quantity under the target load instruction is determined, and the dynamic part fully considers the dynamic compensation of the unit energy storage, gives a certain overshoot fuel quantity to the boiler regulation, speeds up its response, and ensures the rapidity of the unit. The real-time simulation results of virtual DPU (Distributed Processing Unit) show that the designed feedforward-feedback predictive control can quickly respond to load change, improve the resistance to fuel disturbance, reduce the fluctuation of main steam pressure, and better ensure the safe, economic and stable operation of the coordination system.

Fast distributed model predictive control combining ADMM, IPM and Riccati iteration

Abstract In model predictive control, the control action is found at each sampling time by solving an online optimization problem. Computationally, this step is very demanding, especially if compared to the evaluation of traditional control laws. This has limited the application of model predictive control to systems with slow dynamics for many years. Recently, several methods have been proposed in the literature which promise a substantial reduction of the computation time by either running the computation in parallel (distributed model predictive control) or exploiting the problem structure (fast model predictive control). A combination of these methods has not yet been considered in the literature. To achieve this goal, different optimization techniques need to be employed at once. The order of how these methods are applied matters. This paper considers fast distributed model predictive control combining the alternating direction method of multipliers (ADMM), the interior point method (IPM) and the Riccati iteration for a particular class of multi-agent systems for which the order of the methods can be arbitrarily changed. This leads to two different solver schemes where a trade-off arises between computation time and number of communications required to reach consensus. A simplified problem involving the formation control of a fleet of vehicles is considered at the end.

Model predictive control in pharmaceutical continuous manufacturing: A review from a user’s perspective

Pharmaceutical continuous manufacturing is considered as an emerging technology by the regulatory agencies, which have defined a framework guided by an effective quality risk management. With the understanding of process dynamics and the appropriate control strategy, pharmaceutical continuous manufacturing is able to tackle the Quality-by-Design paradigm that paves the way to the future smart manufacturing described by Quality-by-Control. The introduction of soft sensors seems to be a helpful tool to reach smart manufacturing. In fact, soft sensors have the ability to keep the quality attributes of the final drug product as close as possible to their references set by regulatory agencies and to mitigate the undesired events by potentially discard out of specification products. Within this review, challenges related to implementing these technologies are discussed. Then, automation control strategies for pharmaceutical continuous manufacturing are presented and discussed: current control tools such as the proportional integral derivative controllers are compared to advanced control techniques like model predictive control, which holds promise to be an advanced automation concept for pharmaceutical continuous manufacturing. Finally, industrial applications of model predictive control in pharmaceutical continuous manufacturing are outlined. Simulations studies as well as real implementation on pharmaceutical plant are gathered from the control of one single operation unit such as the tablet press to the control of a full direct compaction line. Model predictive control is a key to enable the industrial revolution or Industry 4.0.

Model Predictive Control Based on Linear Parameter-Varying Models of Active Magnetic Bearing Systems

Active magnetic bearing (AMB) system has been recently employed widely as an ideal equipment for high-speed rotating machines. The inherent challenges to control the system include instability, nonlinearity and constricted range of operation. Therefore, advanced control technology is essential to optimize AMB system performance. This paper presents an application of model predictive control (MPC) based on linear parameter-varying (LPV) models to control an AMB system subject to input and state constraints. For this purpose, an LPV model representation is derived from the nonlinear dynamic model of the AMB system. In order to provide stability guarantees and since the obtained LPV model has a large number of scheduling parameters, the parameter set mapping (PSM) technique is used to reduce their number. Based on the reduced model, a terminal cost and an ellipsoidal terminal set are determined offline and included into the MPC optimization problem which are the essential ingredients for guaranteeing the closed-loop asymptotic stability. Moreover, for recursive feasibility of the MPC optimization problem, a slack variable is included into its cost function. The goal of the proposed feedback control system is twofold. First is to demonstrate high performance by achieving stable levitation of the rotor shaft as well as high capability of reference tracking without violating input and state constraints, which increases the overall safety of the system under disturbances effects. Second is to provide a computationally tractable LPVMPC algorithm, which is a substantial requirement in practice for operating the AMB system with high performance over its full range. Therefore, we propose an LPVMPC scheme with frozen scheduling parameter over the prediction horizon of the MPC. Furthermore, we demonstrate in simulation that such frozen LPVMPC can achieve comparable performance to a more sophisticated LPVMPC scheme developed recently and a standard NL MPC (NMPC) approach. Moreover, to verify the performance of the proposed frozen LPVMPC, a comparison with a classical controller, which is commonly applied to the system in practice, is provided.

Simplified Predictive Stator Current Phase Angle Control of Induction Motor With a Reference Manipulation Technique

Finite control set model predictive control (FCS-MPC) is a simple method and has an appropriate dynamic response for the drive applications. Applying additional control objects, e.g., the maximum torque per ampere (MTPA), is easy in FCS-MPC because of its characteristics. A direct application of FCS-MPC to MTPA is the predictive direct angle control method. Though this method eased the MTPA process, the good result is highly sensitive to the proper selection of the weighting factor. Furthermore, finding the best phase angle needs a complicated optimization process. In this paper, the application of simplified predictive control is proposed for angle control. By this proposed method, not only the weighting factor is eliminated but also the constraints of the motor are considered in the control strategy. In this way, the phase angle is automatically controlled in the proper value due to the torque while tedious computation is avoided. Therefore, the proposed method is valid in a wide range of operating points while no optimization process is performed due to changes in speed and torque. This proposed method is evaluated by simulations and experiments.

Application of Model Predictive Control to the reactive extrusion of e-Caprolactone in a twin-screw extruder

Abstract In this work the control of the reactive extrusion of e-Caprolactone in a twin-screw extruder using nonlinear model predictive control with a tracking objective is investigated. For this, the modeling of the extrusion process using a one-dimensional mechanistic model is presented and extended to reactive extrusion systems. A novel modeling approach describing the pressure as a differential state is proposed to be able to use efficient optimization methods. Results for two scenarios, a change of the product quality and a change of the throughput are shown. The tuning of the controller is discussed and the benefits over traditional control methods are elaborated.

Advanced Coolant Temperature Control Study With Integrated Thermal Management System Valve

Abstract Objective of the vehicle thermal management system is to achieve precise coolant temperature control that ensures operation of the combustion engine in the region with the maximum thermal efficiency. In this study it is achieved by a combined effect of coolant temperature control algorithm that is operating an advanced electronic valve with one input and three outputs that balances distribution of the coolant flow between the heat source and sink components. The objective is to present an application of a nonlinear model predictive control (NMPC) approach for coolant temperature control. It is a simulation study to quantify benefits of precise coolant temperature control with the possible inclusion of preview information.

Comparative Application of Model Predictive Control and Particle Swarm Optimization in Optimum Operation of a Large-Scale Water Transfer System

This research evaluates the application and performance of two methods of Model Predictive Control (MPC) and Particle Swarm Optimization (PSO) in real time control and operational management of a large-scale water transfer system. Metaheuristic approaches are computationally common but do not guarantee the global optimality of their solution. On the other hand, a main limitation for these algorithms is the inability to solve large-scale optimization problems due to curse of dimensionality. Model Predictive Control is a modern control method that has been developed for industrial problems and owes its success to easy and effective applicability of constraints on state and control variables. In this research, a multi-objective optimization problem is designed for the Zarrinehrood water transfer system, the largest water transfer line in Middle East, and is solved using MPC and PSO. Two objective functions of maintaining the safe stored water in reservoirs and reducing the fluctuations in pump stations are defined. Results show that the proposed MPC model yields high capability to satisfy all constraints while fluctuations in pump stations are decreased and the volume of stored water in reservoirs meets the lowest error in the vicinity of the safe volume. Performance of MPC in optimizing management of a large-scale water transfer systems like Zarrinehrood with spontaneity in demand sector is better than PSO. However, defining the equations required to perform MPC with respect to the metaheuristic algorithm is more complicated and needs high accuracy to correctly look for the optimal solution.

Application of Model Predictive Control Based on Kalman Filter in Solar Collector Field of Solar Thermal Power Generation

There are two prominent features in the process of temperature control in solar collector field. Firstly, the dynamic model of solar collector field is nonlinear and complex, which needs to be simplified. Secondly, there are a lot of random and uncontrollable, measurable and unmeasurable disturbances in solar collector field. This paper uses Taylor formula and difference approximation method to design a dynamic matrix predictive control (DMC) by linearizing and discretizing the dynamic model of the solar collector field. In addition, the purpose of controlling the stability of the outlet solar field salt temperature is achieved by adjusting the mass flow of molten salt. In order to further improve the ability of the system to suppress unmeasured disturbances, a steady-state Kalman filter is designed to estimate state variables, so that the system has better stability and robustness. The simulation verification results show that the DMC control system based on Kamlan filtering has better control effect than the traditional DMC control system. In the case of large fluctuations in solar radiation intensity and consideration of undetectable interference, the overshoot of the system is reduced by 4% and the rise time remains unchanged.

Application of the Model Predictive Control Based on the Mathematical Theory for Process Industry

Application of the Model Predictive Control Based on the Mathematical Theory for Process Industry

Asynchronous Distributed Model Predictive Control for Optimal Output Consensus of High-Order Multi-Agent Systems

Considering the extensive applications of the model predictive control (MPC), this paper investigates the MPC problem with an application to the optimal output consensus of high-order multi-agent systems. A synchronous distributed optimization algorithm for the MPC problem is proposed, and we further devise its asynchronous version. The algorithm allows agents to choose the uncoordinate step-sizes depending on local information, which improves the flexibility of multi-agent systems. The convergence of the proposed algorithm is guaranteed if the positive uncoordinate step-sizes satisfy the explicit conditions. Compared with the synchronous version that needs a global clock to control all agents for communication and update, the asynchronous algorithm does not require each agent in the multiagent systems to update and communicate simultaneously, and also converges to the optimal global solution to the problem. The numerical simulations demonstrate the availability of the proposed algorithm and the validity of the theoretical results.

Model Predictive Control of the Vertical Gradient Freeze Crystal Growth Process

Abstract This contribution presents the application of nonlinear model predictive control to the Vertical Gradient Freeze crystal growth process. Due to the time-varying spatial extent of the crystal and melt during growth, this process is characterised by two coupled free boundary problems that form a so called two-phase Stefan problem which is of nonlinear nature. To apply model predictive control to this process, a simplified, spatially distributed representation of the system is derived and transferred into a spatially lumped form by means of the finite element method. For this model, a nonlinear control problem is formulated, that takes process limitations into account and tries to satisfy different quality objectives by formulating demands on the systems spatiotemproal temperature distribution. This provides the foundation for the presented predictive control design. Finally, the approximated model and the controller are verified for different real-world scenarios that include model errors and parameter uncertainties.

Sensorless Simplified Finite Control Set Model Predictive Control of SynRM Using Finite Position Set Algorithm

The sensorless application of predictive control in drive applications has been investigated for a decade. Finite control set model predictive control (FCS-MPC) is one of the easy and practical methods in the predictive category. Several methods have been investigated for the sensorless application of FCS-MPC. Since the sensitivity of the predictive method to the speed error is more than that of the classical control methods, sophisticated speed estimators should be used in this method. The model reference adaptive system (MRAS) has been the most successful estimator. The main problem of this estimator is tuning the coefficients in different operating points and the stability of the adaptive function. The finite position set technique is a very recent solution. In this method, the adaptive function is used as the cost function and the optimum rotor position is selected by minimizing that. However, the numerous iteration is a barrier for application to the predictive method. Also, the application of the method for the synchronous reluctance motor (SynRM) is a challenge because of the lack of the rotor model as the adaptive function. In this paper, the finite position technique is modified for the predictive application. The number of iterations is reduced by an optimization method based on sensitivity analysis. Also, a new and simple function is used as the adaptive error function in order to apply the method to the sensorless control of the SynRM. The proposed method is evaluated by simulation and experiment.

Modulated Model Predictive Control Applied to LCL-Filtered Grid-Tied Inverters: A Convex Optimization Approach

Modulated Model Predictive Control Applied to LCL-Filtered Grid-Tied Inverters: A Convex Optimization Approach