Extended Prediction Self-Adaptive Control Research Articles

Context and Driver Dependent Hybrid Electrical Vehicle Operation

This paper studies the driver and context changes during the operation of a hybrid electric vehicle (HEV) and their influence on fuel consumption. Firstly, a context estimation model to recognize driving styles is developed based on machine learning techniques, for which a realistic scenario with simulation of urban mobility (SUMO) and car modeling platform (IPG Carmaker) integration is designed. Secondly, a novel context-aware control strategy based on model predictive control with extended prediction self-adaptive control (MPC-EPSAC) strategy is proposed. The control objective is to achieve optimal torque-split distribution, while optimizing fuel consumption in the parallel HEV. The simulation results suggest that an improvement in fuel economy can be achieved when the driving style in the control loop is adequately considered. Copyright (C) 2020 The Authors.

A Low Computational Cost, Prioritized, Multi-Objective Optimization Procedure for Predictive Control Towards Cyber Physical Systems

Cyber physical systems consist of heterogeneous elements with multiple dynamic features. Consequently, multiple objectives in the optimality of the overall system may be relevant at various times or during certain context conditions. Low cost, efficient implementations of such multi-objective optimization procedures are necessary when dealing with complex systems with interactions. This work proposes a sequential implementation of a multi-objective optimization procedure suitable for industrial settings and cyber physical systems with strong interaction dynamics. The methodology is used in the context of an Extended Prediction self-adaptive Control (EPSAC) strategy with prioritized objectives. The analysis indicates that the proposed algorithm is significantly lighter in terms of computational time. The combination with an input-output formulation for predictive control makes these algorithms suitable for implementation with standardized process control units. Three simulation examples from different application fields indicate the relevance and feasibility of the proposed algorithm.

Distributed Model Predictive Control of Steam/Water Loop in Large Scale Ships

In modern steam power plants, the ever-increasing complexity requires great reliability and flexibility of the control system. Hence, in this paper, the feasibility of a distributed model predictive control (DiMPC) strategy with an extended prediction self-adaptive control (EPSAC) framework is studied, in which the multiple controllers allow each sub-loop to have its own requirement flexibility. Meanwhile, the model predictive control can guarantee a good performance for the system with constraints. The performance is compared against a decentralized model predictive control (DeMPC) and a centralized model predictive control (CMPC). In order to improve the computing speed, a multiple objective model predictive control (MOMPC) is proposed. For the stability of the control system, the convergence of the DiMPC is discussed. Simulation tests are performed on the five different sub-loops of steam/water loop. The results indicate that the DiMPC may achieve similar performance as CMPC while outperforming the DeMPC method.

Nonlinear Predictive Control Applied to Steam/Water Loop in Large Scale Ships

Abstract In steam/water loop for large scale ships, there are mainly five sub-loops posing different dynamics in the complete process. When optimization is involved, it is necessary to select different prediction horizons for each loop. In this work, the effect of prediction horizon for Multiple-Input Multiple-Output (MIMO) system is studied. Firstly, Nonlinear Extended Prediction Self-Adaptive Controller (NEPSAC) is designed for the steam/water loop system. Secondly, different prediction horizons are simulated within the NEPSAC algorithm. Based on simulation results, we conclude that specific tuning of prediction horizons based on loop’s dynamic outperforms the case when a trade-off is made and a single valued prediction horizon is used for all the loops.

Effect of Control Horizon in Model Predictive Control for Steam/Water Loop in Large-Scale Ships

This paper presents an extensive analysis of the properties of different control horizon sets in an Extended Prediction Self-Adaptive Control (EPSAC) model predictive control framework. Analysis is performed on the linear multivariable model of the steam/water loop in large-scale watercraft/ships. The results indicate that larger control horizon values lead to better loop performance, at the cost of computational complexity. Hence, it is necessary to find a good trade-off between the performance of the system and allocated or available computational complexity. In this original work, this problem is explicitly treated as an optimization task, leading to the optimal control horizon sets for the steam/water loop example. Based on simulation results, it is concluded that specific tuning of control horizons outperforms the case when only a single valued control horizon is used for all the loops.

Integrated B2B‐NMPC control strategy for batch/semibatch crystallization processes

The uncertainty in crystallization kinetics is of major concern in manufacturing processes, which can result in deterioration of most model‐based control strategies. In this study, uncertainties in crystallization kinetic parameters were characterized by Bayesian probability distributions. An integrated B2B‐NMPC control strategy was proposed to first update the kinetic parameters from batch to batch using a multiway partial least‐squares (MPLS) model, which described the variances of kinetic parameters from that of process variables and batch‐end product qualities. The process model with updated kinetic parameters was then incorporated into an NMPC design, the extended prediction self‐adaptive control (EPSAC), for online control of the final product qualities. Promising performance of the proposed integrated strategy was demonstrated in a simulated semibatch pH‐shift reactive crystallization process to handle major crystallization kinetic uncertainties of L‐glutamic acid, wherein smoother and faster convergences than the conventional B2B control were observed when process dynamics were shifted among three scenarios of kinetic uncertainties. © 2017 American Institute of Chemical Engineers AIChE J, 2017

Dynamic and Reactive Walking for Humanoid Robots Based on Foot Placement Control

This paper presents a novel online walking control that replans the gait pattern based on our proposed foot placement control using the actual center of mass (COM) state feedback. The analytic solution of foot placement is formulated based on the linear inverted pendulum model (LIPM) to recover the walking velocity and to reject external disturbances. The foot placement control predicts where and when to place the foothold in order to modulate the gait given the desired gait parameters. The zero moment point (ZMP) references and foot trajectories are replanned online according to the updated foothold prediction. Hence, only desired gait parameters are required instead of predefined or fixed gait patterns. Given the new ZMP references, the extended prediction self-adaptive control (EPSAC) approach to model predictive control (MPC) is used to minimize the ZMP response errors considering the acceleration constraints. Furthermore, to ensure smooth gait transitions, the conditions for the gait initiation and termination are also presented. The effectiveness of the presented gait control is validated by extensive disturbance rejection studies ranging from single mass simulation to a full body humanoid robot COMAN in a physics based simulator. The versatility is demonstrated by the control of reactive gaits as well as reactive stepping from standing posture. We present the data of the applied disturbances, the prediction of sagittal/lateral foot placements, the replanning of the foot/ZMP trajectories, and the COM responses.

Just-in-Time-Learning based Extended Prediction Self-Adaptive Control for batch processes

This article presents a new Extended Prediction Self-Adaptive Control (EPSAC) algorithm based on the Just-in-Time Learning (JITL) method. In the proposed JITL-based EPSAC design, linearization of the process model is achieved by a set of local state-space models, each of which can be independently and simultaneously identified by the JITL method along the base trajectory. For the end-product quality control for a simulated semi-batch pH-shift reactive crystallization process where shrinking prediction and control horizons are essential, the proposed EPSAC algorithm not only simplifies the control weight tuning but also provides better and more robust closed-loop control performance than its previous counterpart.

Real-Time Optimization of Organic Rankine Cycle Systems by Extremum-Seeking Control

In this paper, the optimal operation of a stationary sub-critical 11 kW el organic Rankine cycle (ORC) unit for waste heat recovery (WHR) applications is investigated, both in terms of energy production and safety conditions. Simulation results of a validated dynamic model of the ORC power unit are used to derive a correlation for the evaporating temperature, which maximizes the power generation for a range of operating conditions. This idea is further extended using a perturbation-based extremum seeking (ES) algorithm to identify online the optimal evaporating temperature. Regarding safety conditions, we propose the use of the extended prediction self-adaptive control (EPSAC) approach to constrained model predictive control (MPC). Since it uses input/output models for prediction, it avoids the need for state estimators, making it a suitable tool for industrial applications. The performance of the proposed control strategy is compared to PID-like schemes. Results show that EPSAC-MPC is a more effective control strategy, as it allows a safer and more efficient operation of the ORC unit, as it can handle constraints in a natural way, operating close to the boundary conditions where power generation is maximized.

A Portable Implementation on Industrial Devices of a Predictive Controller Using Graphical Programming

This paper presents an approach for developing an extended prediction self-adaptive controller employing graphical programming of industrial standard devices for controlling fast processes. For comparison purposes, the algorithm has been implemented on three different field-programmable gate arrays (FPGAs) chips. This paper presents research aspects regarding graphical-programming controller design, showing that a single advanced control application can run on different targets without requiring significant program modifications. Based on the time needed for processing the control signal and on the application, one can efficiently and easily select the most appropriate device. To exemplify the procedure, a conclusive case study is presented.

Constrained Multivariable Predictive Control of a Train of Cryogenic 13C Separation Columns

This work presents a linear, constrained, multivariable predictive control strategy, i.e. EPSAC (Extended Prediction Self-Adaptive Control) applied to a train of three distillation columns. The columns are used to obtain the carbon isotope 13C used widely in medicine and specific industries. The oversimplified models of the three columns stem from a real-life plant designed and constructed in the National Institute for Research and Development for Isotopic and Molecular Technologies, in Cluj Napoca, Romania. The simulation results suggest the strategy is applicable to this process.

Enhancing the robustness of the EPSAC predictive control using a Singular Value Decomposition approach

In this paper, we present a Robust Model Predictive Control (MPC) based on the Singular Value Decomposition (SVD) analysis to handle long prediction and control horizons where numerical instability might appear. The proposed method is developed following the Extended Prediction Self-Adaptive Control (EPSAC) algorithm. The performance of the controller is evaluated in simulation using a 4th order mass–spring–damper system, and the dynamic walking of the humanoid COMAN. The stability of the closed-loop system is analysed using root-locus and Bode plots whilst robustness tests are performed by introducing modelling errors in the prediction model. The results show that the proposed extension increases the robustness of the feedback control, and therefore the operational range of the system.

A constrained EPSAC approach to inventory control for a benchmark supply chain system

The design of an appropriate inventory control policy for a supply chain (SC) plays an essential role in tempering inventory instability and bullwhip effect. Several constraints are commonly encountered in actual operations so managers are required to take these physical restrictions into account when designing the inventory control policy. Model predictive control (MPC) appears as a promising solution to this issue, due to its capability of finding optimal control actions for a constrained SC system. Therefore, the inventory control problem for a benchmark SC is solved using the extended prediction self-adaptive control approach to MPC. To extend methodologies in our previous work, the control framework relies on generic process model and incorporates the physical constraints arising from practical operations to form the general constrained optimisation problems. The managers can choose from decentralised and centralised control structures according to specific informational and organisational factors of their SCs. The proposed control schemes in this study may be appropriate for industrial practice because the designed policy can bring a reduction of over 30% in operating cost and a significant increase of customer satisfaction level compared with that of the conventional policy.

Quantifying and mitigating the bullwhip effect in a benchmark supply chain system by an extended prediction self-adaptive control ordering policy

An undesired observation known as the bullwhip effect in supply chain management leads to excessive oscillations of the inventory and order levels. This paper presents how to quantify and mitigate the bullwhip effect by introducing model predictive control (MPC) strategy into the ordering policy for a benchmark supply chain system. Instead of quantifying the bullwhip effect with commonly used statistical measure, we derive equivalently the expression of bullwhip metric via control-theoretic approach by applying discrete Fourier transform and (inverse) z-transform when the demand signal is stationary stochastic. A four-echelon supply chain is formulated and its dynamical features are analyzed to give the discrete model. An extended prediction self-adaptive control (EPSAC) approach to the multi-step predictor is implemented in the development of MPC formulation. The closed-form solution to MPC problem is derived by minimizing a specified objective function. The transfer function for MPC ordering policy is then obtained graphically from an equivalent representation of this closed-form solution. A numerical simulation shows that MPC ordering policy outperforms the traditional ordering policies on reducing bullwhip effect.

Decentralized and centralized model predictive control to reduce the bullwhip effect in supply chain management

Mitigating the bullwhip effect is one of crucial problems in supply chain management. In this research, centralized and decentralized model predictive control strategies are applied to control inventory positions and to reduce the bullwhip effect in a benchmark four-echelon supply chain. The supply chain under consideration is described by discrete dynamic models characterized by balance equations on product and information flows with an ordering policy serving as the control schemes. In the decentralized control strategy, a MPC-EPSAC (Extended Prediction Self-Adaptive Control) approach is used to predict the changes in the inventory position levels. A closed-form solution of an optimal ordering decision for each echelon is obtained by locally minimizing a cost function, which consists of the errors between predicted inventory position levels and their setpoints, and a weighting function that penalizes orders. The single model predictive controller used in centralized control strategy optimizes globally and finds an optimal ordering policy for each echelon. The controller relies on a linear discrete-time state-space model to predict system outputs. But the predictions are approached by either of two multi-step predictors depending on whether the states of the controller model are directly observed or not. The objective function takes a quadratic form and thus the resulting optimization problem can be solved via standard quadratic programming method. The comparisons on performances of the two MPC strategies are illustrated with a numerical example.

Implementation of Robust EPSAC on dynamic walking of COMAN Humanoid

In this paper we present the Model Predictive Control (MPC) with dynamic constraints for generating dynamic walking for the compliant humanoid COMAN. The dynamics of the robot are modeled using the cart-table model which allows the generation of a dynamically balanced gait given a planned walking pattern based on the Zero Moment Point (ZMP). Our simulation study of the MPC's implementation on bipedal walking finds out that a large receding and control horizons are needed to track a predefined walking pattern, leading to numerical instability. Therefore, the Extended Prediction Self-Adaptive Control (EPSAC) approach for MPC has been used and a method based on the analysis of the Singular Value Decomposition (SVD) is presented as new contribution to guarantee feasibility, robustness and stability of the MPC formulation. Study on an inverted pendulum and the COMAN humanoid prove that the proposed strategy improves the robustness and stability of the original EPSAC controller, in both well or ill conditioned systems. The simulation results finally demonstrate that the proposed methodology is well suited to smoothly track a dynamic walking pattern.

Nonlinear Predictive Control of Shank Movement Around the Knee Joint

This paper derives an elementary mathematical model for the movement of the shank around the knee joint. The resulting equation of motion is a nonlinear differential equation. The control purpose is to regulate the angle of the shank by manipulating the torque applied to the knee joint. This model has applications in dynamical knee rigs and active orthoses. As the system is nonlinear with a rather larger range of motion, a nonlinear predictive controller is designed. The implemented control strategy is NEPSAC (Nonlinear Extended Prediction Self-Adaptive Control). The effect of the filter parameter and the prediction horizon in the NEPSAC algorithm are investigated using MATLAB/SIMULINK. The results show that for both the filter parameter and the prediction horizon an optimal value can be chosen as a trade-off between robustness and fastness of the response.

Accelerating animal cell growth in perfusion mode by multivariable control: simulation studies

This study considers the problem of manipulating in an optimal way the perfusion and bleed flow rates of a continuous culture of hybridoma cells, so as to achieve a fast transient start-up and reject potential disturbances. The proposed solution makes use of an analysis of the properties of the steady state solutions of the nonlinear dynamic model of the cell culture, and in particular the relationship between the two main limiting substrates, glucose and glutamine. The solution is implemented using extended prediction self-adaptive control. Simulation results demonstrate the approach potentiality.

Reducing the Bullwhip Effect in Supply Chain Management by Applying a Model Predictive Control Ordering Policy

An undesired situation in supply chain management known as the bullwhip effect, suggests that the demand variability amplifies toward upstream nodes, leading to significant oscillations in the inventory and order levels. The purpose of this paper is to show the ability to reduce the bullwhip effect by using model based predictive control (MPC) algorithms. First, a discrete dynamic model that describes the basic features of a four nodes supply chain is presented. The model is characterized by the material balance and information flow with an ordering policy serving as the control input. An MPC-EPSAC (Extended Prediction Self-Adaptive Control) approach to predicting the changes of inventory position level is used. The closed form solution of optimal ordering is obtained by minimizing a cost function of the errors between the predicted inventory position levels and their setpoints, while using a weighting function that penalizes the excess movement of orders. The ordering rate transfer function is derived by Diophantine equations. Under the condition of a stationary stochastic demand pattern, the link of bullwhip metric in the sense of common statistical measure to control theory metric is made by inverse z -transform. An illustrative example shows that MPC-EPSAC outperforms the traditional ordering policy in reducing the bullwhip effect.

EPSAC Predictive Control Applied to Muscle Relaxant Administration

Abstract An Extended Prediction Self-Adaptive Control (EPSAC) strategy is applied to the rocuronium administration, a neuromuscular blocking drug used in anaesthesia to facilitate endotracheal intubation and to provide skeletal muscle relaxation during surgery. A safe anaesthesia should provide comfort to the patient and best possible working conditions for the surgeon. The muscle relaxant plays an important role and an adequate muscle relaxation, which allows efficient and safe surgery, can be provided using automated control. Moreover, closed-loop control may reduce anesthetist workload and achieve better regulation of muscle relaxation than manual administration, calculating the dose required for each patient and avoiding in this way overdosing. Predictive control requires prior knowledge of system dynamics, therefore a mathematical model is required. A three-compartment model for rocuronium is presented. Several parameter sets are used to simulate different patients and control performance is discussed.