Time-varying multi-objective region iterative learning motion control

Wan Xu,Tingwei Chen,Di Xiao 대한기계학회 2023 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.37 No.1

In the multi-axis motion control field, the point-to-point multi-objective learning algorithm, taking target point tracking accuracy as the major control objective, limits the optimization of other control objectives such as motion distance and energy consumption of the system. Therefore, this paper initially proposes multi-objective region iterative learning control algorithm, extending point-to-point tracking to region-to-region tracking and further optimizing other performance indexes via flexibly setting the region size on the grounds of the tracking requirements. In view that the fixed weight coefficient of each object in the multi-objective iterative learning algorithm is not suitable for occasions with dynamic control requirements, this paper further presents time-varying multi-objective region iterative learning control algorithm and unifies point-to-point tracking and region-to-region tracking under one algorithm framework. In addition, by means of designing time-varying weight matrix, strengthening major control objectives, weakening minor control objectives and satisfying the ever-changing control requirements, experimental results show that time-varying multi-objective region iterative learning algorithm is capable of flexibly adjusting the weight size of each objective in line with various control needs so that the control system is endowed with higher tracking accuracy in trajectory tracking segment as well as lower energy consumption and shorter motion distance in region tracking segment, thus having good flexibility.

다중 MR 감쇠기의 효과적인 동시제어를 위한 제어알고리즘 개발

김현수,강주원,Kim, Hyun-Su,Kang, Joo-Won 한국공간구조학회 2013 한국공간구조학회지 Vol.13 No.3

A multi-input single-output (MISO) semi-active control systems were studied by many researchers. For more improved vibration control performance, a structure requires more than one control device. In this paper, multi-input multi-output (MIMO) semi-active fuzzy controller has been proposed for vibration control of seismically excited small-scale buildings. The MIMO fuzzy controller was optimized by multi-objective genetic algorithm. For numerical simulation, five-story example building structure is used and two MR dampers are employed. For comparison purpose, a clipped-optimal control strategy based on acceleration feedback is employed for controlling MR dampers to reduce structural responses due to seismic loads. Numerical simulation results show that the MIMO fuzzy control algorithm can provide superior control performance to the clipped-optimal control algorithm.

다목적 유전자알고리즘을 이용한 첨단기술산업 시설물의 스마트 미진동제어

김현수,강주원,김영식,Kim, Hyun-Su,Kang, Joo-Won,Kim, Young-Sik 한국공간구조학회 2013 한국공간구조학회지 Vol.13 No.2

Reduction of microvibration is regarded as important in high-technology facilities with high precision equipments. In this paper, smart control technology is used to improve the microvibration control performance. Mr damper is used to make a smart base isolation system amd fuzzy logic control algorithm is employed to appropriately control the MR damper. In order to develop optimal fuzzy control algorithm, a multi-objective genetic algorithm is used in this study. As an excitation, a train-induced ground acceleration is used for time history analysis and three-story example building structure is employed. Microvibration control performance of passive and smart base isolation systems have been investigated in this study. Numerical simulation results show that the multi-objective genetic algorithm can provide optimal fuzzy logic controllers for smart base isolation system and the smart control system can effectively reduce microvibration of a high-technology facility subjected to train-induced excitation.

Robust Multi Objective H₂/H<SUB>∞</SUB> Control of MIMO Nonlinear Uncertain systems via T-S Fuzzy Model

Vahid Azimi,Peyman Akhlaghi,Mohammad Hossein Kazemi 제어로봇시스템학회 2011 제어로봇시스템학회 국제학술대회 논문집 Vol.2011 No.10

This paper describes robust H₂/H∞ multi-objective state feedback controller for nonlinear uncertain systems. To apply the H₂/H∞ multi-objective state feedback method, the nonlinear dynamics is represented by a T-S fuzzy model. First, uncertain parameters and Quantification of uncertainty on physical parameters is defined by affine parameter-dependent systems method. Next, the Takagi and Sugeno"s fuzzy linear model is utilized to approximate uncertain nonlinear systems. Then, some states (error of tracking) are augmented to the system in order to improve tracking control. Finally, based on fuzzy linear model with augmented state, a H₂/H∞ multi-objective state feedback controller is developed to achieve the robustness design of nonlinear uncertain systems. LMI (Linear Matrix Inequality) method and PDC (Parallel Distributed Compensation) are used to design the controller for the whole system. The results show that the proposed method can effectively meet the performance requirements like robustness, disturbance rejection and tracking for the 3-phase permanent magnet synchronous motor (PMSM).

Distributed Pareto Reinforcement Learning for Multi-objective Smart Generation Control of Multi-area Interconnected Power Systems

Yin Linfei,Cao Xinghui,Sun Zhixiang 대한전기학회 2022 Journal of Electrical Engineering & Technology Vol.17 No.5

Multiple objectives of smart generation control (SGC) are considered. The paper proposes a distributed Pareto reinforcement learning (DPRL) based on game theory to address the multi-objective control problem (MOCP) of SGC of multi-area interconnected power systems (MAIPSs). The proposed DPRL consists of the framework of reinforcement learning and multiple Q matrices, which can provide multiple outputs for MOCPs. As a control algorithm based on the Markov decision process, the stability of the proposed approach can be guaranteed with the ability to provide dynamic control strategies online. With the Pareto concept of optimization algorithms introduced into a control algorithm, the proposed approach can obtain multiple comprehensive objectives. The case studies under the two-area and practical four-area power systems show that: compared with four reinforcement learning methods and a proportional-integral controller, the proposed DPRL can achieve the highest control performance with multiple objectives in MAIPSs. The case studies of SGC under two MAIPSs verify the feasibility and eff ectiveness of the DPRL for MOCPs.

Design and Application of Hierarchical Multi-objective Predictive Control for Continuous Flow Stirred Tank Reactor

Fan Zhang,Yi Zhang,Yali Xue 제어·로봇·시스템학회 2022 International Journal of Control, Automation, and Vol.20 No.5

In order to deal with various control objectives in process, such as economic and regulating preferences, the multi-objective predictive control (MOPC) is under investigation in this manuscript. The proposed method has the hierarchical structure: in the upper layer fuzzy control is devised for the compromise point of the process under multi-objective consideration at steady-state, which provides stability constraint for the MOPC; the formulated MOPC in the lower layer directly optimizes distance to the utopia point. The proposed fuzzy control based multiobjective predictive control can simultaneously reduce the computation burden and increase the multi-objective performance. Effectiveness and merits are verified on the typical continuous flow stirred-tank reactor process.

Metaheuristics-based Multi-objective Design of Global Robust Optimal Sliding Mode Control of Discrete Uncertain Systems

Wafa Boukadida,Anouar Benamor,Hassani Messaoud,Patrick Siarry 제어·로봇·시스템학회 2019 International Journal of Control, Automation, and Vol.17 No.6

The paper introduces a new design control strategy for finite time stabilization of uncertain discrete systems. The proposed control method uses the First Order Sliding Mode Control (FOSMC), the Linear Quadratic Regulator (LQR) and Genetic Algorithms (GA). The key idea of combining two control strategies lies in the fact that a robust controller tackles the uncertainties when the optimal controller performances are unaffected. Firstly, as the performances of the Sliding Mode Control (SMC) greatly depends on the choice of the sliding surface, a novel method based on the resolution of a Sylvester equation is proposed. Secondly, to ensure the robustness in the whole space, the Integral Sliding Mode Control (ISMC) is combined with the above LQR method. Rigorous stability analysis for the closed-loop system is given and proves that the system states are always bounded even the unknown disturbances. In order to handle the problem accompanying the LQR synthesis as well as the gain of the robust compensator, a multi-objective optimization problem considering a set of competing objectives is solved using GA.For that,a novel dynamically aggregated objective function is proposed. A set of non dominated optimal solutions are provided to the control engineer and then may select the most adequate solution among the resulting optimal Pareto set. The novel method has successfully solved the robust stabilization of two known benchmarks and the different results asserts not merely the efficiency as well as the quality of the designed controller but also its supremacy compared to some existing ones.

스마트 연결 제어 시스템과 연결 구조물의 통합 최적 설계

김현수,강주원 한국공간구조학회 2019 한국공간구조학회지 Vol.19 No.2

A smart connective control system was invented recently for coupling control of adjacent buildings. Previous studies on this topic focused on development of control algorithm for the smart connective control system and design method of control device. Usually, a smart control devices are applied to building structures after structural design. However, because structural characteristics of building structure with control devices changes, a iterative design is required for optimal design. To defeat this problem, an integrated optimal design method for a smart connective control system and connected buildings was proposed. For this purpose, an artificial seismic load was generated for control performance evaluation of the smart coupling control system. 20-story and 12-story adjacent buildings were used as example structures and an MR (magnetorheological) damper was used as a smart control device to connect adjacent two buildings. NSGA-II was used for multi-objective integrated optimization of structure-smart control device. Numerical simulation results show the integrated optimal design method proposed in this study can provide various optimal designs for smart connective control system and connected buildings presenting good control performance.

인접건물의 준능동 퍼지제어를 위한 유전자알고리즘 기반 다목적 최적설계

김현수(Kim, Hyun-Soo) 한국산학기술학회 2016 한국산학기술학회논문지 Vol.17 No.1

본 연구에서는 지진하중을 받는 인접한 건물의 진동제어를 위한 준능동 제어장치의 제어성능을 검토하였다. 준능동 제어장치로는 MR 감쇠기를 사용하였다. MR 감쇠기로 연결된 인접한 건물을 효과적으로 제어하기 위하여 퍼지제어알고리즘 을 사용하였다. MR 감쇠기로 연결된 인접한 건물의 제어시 한쪽 건물의 응답을 저감시키는 것은 다른 한 쪽 건물의 응답을 증가시키는 효과를 가져온다. 따라서 연결된 건물의 제어는 서로 상충되는 특성이 있기 때문에 다목적 최적화문제로 귀결된 다. 따라서 본 연구에서는 다목적 유전자알고리즘을 사용하여 MR 감쇠기를 제어하는 퍼지제어알고리즘을 최적화하였다. 수 치해석을 통하여 준능동 MR 감쇠기를 이용한 인접건물의 연결제어효과를 검토하였고 매우 우수한 성능을 나타내는 것을 확인하였다. The vibration control performance of a semi-active damper connected to adjacent buildings subjected to seismic loads was investigated. The MR damper was used as a semi-active control device. A fuzzy logic control algorithm was used for effective control of the adjacent buildings connected to the MR damper. In the buildings control coupled with a MR damper, the response reduction of one building results in an increase in the response in another building. Because of these conflict characteristics, multi-objective optimization is required. Therefore, a fuzzy logic control algorithm for the control of a MR damper was optimized using a multi-objective genetic algorithm. Based on numerical analyses, the semi-active fuzzy control of MR damper for adjacent coupled buildings can provide good control performance.

Research on Fractional Order Fuzzy PID Control of the Pneumatic-hydraulic Upper Limb Rehabilitation Training System Based on PSO

Hongyan Wang,Jinbo Lu 제어·로봇·시스템학회 2022 International Journal of Control, Automation, and Vol.20 No.1

In this paper, a new type of pneumatic-hydraulic upper limb rehabilitation training system is designed, and the system modeling and control method are studied. The 2DOF fuzzy fractional PID controller is proposed, which provides a new reference for improving the control accuracy of two degrees of freedom pneumatic system. Due to the increase of the regulating parameters of the fractional order controller, a multi-objective particle swarm optimization algorithm is designed to determine the parameters of the control system and optimize the determination process of the parameters. In order to verify the effectiveness of the control method, an experimental study was carried out by building a pneumatic-hydraulic upper limb rehabilitation training system platform. Compared with PID controller, the overshoot of the system controlled by fractional order fuzzy PID controller is reduced by 17.3%. The peak time is shortened about 11.7%, and the delay time of sine tracking signal is reduced about 10.6%. The results show that the control performance of two degrees of freedom fractional order fuzzy PID controller is better than other controllers.