Penalty function and adaptive control of constrained finite Markov chains

In this paper we consider the adaptive control of constrained finite ergodic controller Markov chains whose transition probabilities are unknown. The control policy is designed to achieve the minimization of a loss function under a set of inequality constraints. The average values of conditional mathematical expectations of this loss function and constraints are also assumed to be unknown. A regularized penalty function is introduced to derive an adaptive control algorithm. In this algorithm the transition probabilities of the Markov chain and the average values of the constraints are estimated at each time n. The control policy is adjusted using the Bush–Mosteller reinforcement scheme as a stochastic approximation procedure. Its asymptotic properties are stated. We establish that the optimal convergence rate is equal to n^-1/3+δ (δ is any small positive parameter). © 1998 John Wiley & Sons, Ltd.     

Markov chains for multipartitioning large power system state estimation networks

This paper presents a new and efficient algorithm for multipartitioning an observable power system state estimation network into observable subnetworks. The partitioning algorithm, which uses the spanning tree of an observable network, is based on Markov chains and has a stochastic basis, rather than a heuristic derivation. This algorithm is faster and provides all the possible optimal partitions of a spanning tree. Once the spanning tree is optimally partitioned into full rank subspanning trees, the interconnected lines between the partitioned subnetworks can be obtained directly from the original network. Computational examples using large power networks are given, to illustrate the properties of the proposed algorithm.     

An iterative learning control method with application for CNCmachine tools

A proportional, integral, and derivative (PID) type iterative learning controller is proposed for precise tracking control of industrial robots and computer numerical controller (CNC) machine tools performing repetitive tasks. The convergence of the output error by the proposed learning controller is guaranteed under a certain condition even when the system parameters are not known exactly and unknown external disturbances exist. As the proposed learning controller is repeatedly applied to the industrial robot or the CNC machine tool with the path-dependent repetitive task, the distance difference between the desired path and the actual tracked or machined path, which is one of the most significant factors in the evaluation of control performance, is progressively reduced. The experimental results demonstrate that the proposed learning controller can improve machining accuracy when the CNC machine tool performs repetitive machining tasks     

鲁棒迭代学习控制律的优化设计方法

针对非参数不确定性系统难以实施非因果学习控制的问题,提出一种鲁棒迭代学习控制律的优化设计方法。采用基于名义性能指标的优化设计方法,预留调整加权,并将其演变成频域描述下具有系统伪逆形式的迭代学习控制律。进一步建立μ分析下的系统框架,分析加权系数对系统收敛速度和鲁棒性的影响。根据上述分析来确定初始的加权系数,并采用零相位滤波器和μ分析来修正加权,使系统满足鲁棒收敛性条件,从而得到优化的鲁棒迭代学习控制律。文中附有算例,将时域设计应用到频域分析框架,可以实现非因果稳定系统的鲁棒收敛性分析,仿真结果验证了方法的有效性。     

网络系统的马尔科夫时滞预测控制

针对非线性网络控制系统存在网络诱导时延的问题,提出一种时滞补偿控制方法。首先建立非线性网络系统的T-S模糊模型,将网络诱导时延视为系统输入时滞;利用马尔科夫链对网络诱导时滞进行建模,在此基础上计算状态转移概率矩阵,依据最大转移概率法对下一步的时滞进行预测;采用遗传算法对状态反馈控制器进行寻优,得到T-S模糊模型每个子系统针对各个时滞的最优状态反馈控制律;通过时滞时间的隶属度函数融合得到非线性网络控制系统的模糊预测控制器;仿真算例验证该方法的可行性与有效性。     

A numerical filtering method for linear state-space models with Markov switching

A class of discrete-time random processes arising in engineering and econometrics applications consists of a linear state-space model whose parameters are modulated by the state of a finite-state Markov chain. Typical filtering approaches are collapsing methods, which approximate filtered distributions by mixtures of Gaussians, each Gaussian corresponding to one possibility of the recent history of the Markov chain, and particle methods. This article presents an alternative approach to filtering these processes based on keeping track of the values of the underlying filtered density and its characteristic function on grids. We prove that it has favorable convergence properties under certain assumptions. On the other hand, as a grid method, it suffers from the curse of dimensionality, and so is only suitable for low-dimensional systems. We compare our method to collapsing filters and a particle filter with examples, and find that it can outperform them on 1- and 2-dimensional problems, but loses its speed advantage on 3-dimensional systems. Meanwhile, our method has a proven theoretical convergence rate that is probably not achieved by collapsing and particle methods.     

A direct solution method for carrier transport using Markov process representation

A direct solution method for semi-classical carrier transport problems by representing the transport process as a Markov process (also known as Continuous-Time Markov chain) is introduced to evaluate the carrier mobility in semiconductors. Comparison of the results from our direct method to conventional Monte Carlo method shows that this method outperforms the Monte Carlo method in both computational efficiency and accuracy.     

A method of designing ac/dc thyristor converter current control systems by using discrete time MRAC with deterministic disturbances

Two design methods of ac/dc thyristor converter current control systems with MRAC (Model Reference Adaptive Control) have been designed. To linearize the input/output relation of the systems, the constant term of the model is considered in different ways. In the first method, a linear transformation of the control input is made while in the second method, the constant term and compensated modeling error are neglected by integral adaptation. Computer simulations and experiments have shown that the control performances of the MRAC systems are much better than that of PI controllers. However, in the first method, since the variable vector is not persistently spanning, the convergence rate of estimated parameters is not as high as expected. The second method has better dynamic characteristics but its stability has not been proved in theory. This paper suggests a new design method in which the constant term is considered as a deterministic disturbance. The feature of this method is that by introducing an equality equation which makes the term of disturbances apparently unobservable, the effect of disturbances can be eliminated and the existing adaptive control scheme without disturbances whose stability has been proved theoretically can be used. Computer simulations and experiments show that the new method can obtain better responses than the old ones.     

A deterministic Wigner approach for superposed states

Journal of Computational Electronics - The Wigner formalism is a convenient way of describing quantum mechanical effects through a framework of distribution functions in phase space. Currently,...     

Adaptive control for time-varying systems: A combination of Martingale and Markov chain techniques

Adaptive control problems of a first-order randomly time-varying stochastic system are considered. A class of adaptive controllers based on the Kalman filter is introduced and is analysed using a combination of martingale and Markov chain techniques. It is shown that both the expected value and sample path averages of the square of the output of the closed-loop system remain bounded and that the long-run cost is a continuous functional of the parameters of the controller and the distribution of the disturbance process. These results hold even when the Gaussian assumption used in previous papers is removed and the a priori estimate of the noise variance is incorrect.     

基于双马尔科夫链的新型能源互联网虚假数据注入攻击检测

新型能源互联网中各环节信息的高度融合和开放削弱了其防御外界攻击的能力.针对具有隐蔽性特征的虚假数据注入攻击(FDIA),提出了适用于新型能源互联网的基于双马尔科夫链的FDIA检测方法.考虑到新型能源互联网的FDIA原理、特征以及新型能源互联网包含大量的测量数据和多变的运行状态,将待检测的数据映射到2个不同的状态空间,并生成2个不同的马尔科夫链模型;根据该模型估计所得的能源互联网运行状态的精确度,生成FDIA的检测器.通过实验案例验证所提检测方法的正确性和有效性.实验结果表明:所提FDIA检测方法具有优秀的检测概率和较少的检测计算量,检测概率可达98.60％,虚警概率仅为1.35％,计算量相比于支持向量机方法降低了1个数量级,能满足新型能源互联网的应用要求.     

一种柔性直流输电系统PID-ANFIS优化控制方法

针对柔性直流输电系统(Voltage Source Converter based High-Voltage Direct-Current,VSC-HVDC)双闭环控制中PI控制存在参数整定困难及控制器数量过多等问题,提出一种具有PID功能的自适应神经元模糊推理系统(Adaptive Neuro-Fuzzy Inference System with PID function,PID-ANFIS)控制器用于该系统控制。其中,PID-ANFIS控制器兼有神经网络控制、二阶模糊控制及PID功能;同时提出的基于三重合作粒子群算法(Treble Cooperative Particle Swarm Optimization,TCPSO)用于优化该控制器中神经网络参数。TCPSO采用由降维合作、分组合作与记忆合作组成的三重合作策略,极大程度上提升了神经网络参数优化的精度。深入研究了TCPSO优化PID-ANFIS控制器参数的步骤。基于TCPSO优化的PID-ANFIS控制器能够实现VSC-HVDC系统的直接功率控制效果。仿真结果表明该控制器具有控制速度快、超调量小、抗干扰能力强等优点,是VSC-HVDC控制系统的一个可行方案。     

Markov跳变系统的滚动时域鲁棒双模控制

针对一类具有输入和状态联合约束的不确定离散时间Markov跳变系统,提出滚动时域鲁棒双模控制的方法.基于椭圆不变集的性质,在椭圆不变集以外使用滚动时域控制,驱动状态进入椭圆不变集.一旦状态进入椭圆不变集,改为使用线性状态反馈控制以保证状态始终位于椭圆不变集内,并使系统均方稳定.控制器的求解转化为以线性矩阵不等式形式给出的半正定规划问题.为改进半正定规划问题的可行性,将自由终端成本的上界作为椭圆不变集的边界,在优化成本的同时得到一组压缩椭圆集,相对于固定椭圆不变集降低了保守性.仿真实例验证了所提出方法的可行性和有效性.     

基于卡尔曼滤波迭代学习的交通信号控制方法

由于城市交通流具有高度的复杂性,路网内存在的非重复性干扰会使迭代学习的交通控制系统动态性能变差。因此,提出了一种卡尔曼滤波器和迭代学习的交通信号复合控制方法,以进一步改善控制系统的鲁棒性和抗干扰能力。该控制方法首先利用卡尔曼滤波器对系统的状态进行观测,在含有随机噪声干扰的情况下,估计系统的最优状态;其次设计了带遗忘因子的迭代学习控制方法,遗忘因子可增强对大幅扰动的抗干扰能力,再通过迭代学习逐渐跟踪系统的参考轨迹;最后,对该算法的收敛性进行了数学证明,而仿真的实验结果也表明在扰动环境下利用提出的方法可以降低干扰对控制系统的影响,在一定程度上提高了道路通行能力、减少了交通拥堵。     

SVC平衡控制方法及其所需信号的检测

本文探讨了静止无功补偿器（SVC）在三相对称和不对称、电压电流波形有畸变情况下的平衡控制方法和控制信号检测的方法。对于控制所需要的信号，包括电压和无功电流的有效值、无功功率和功率因数等，给出了基于瞬时功率理论的信号检测方法。对所提出的方法进行仿真并将该方法应用于近期研制的SVC样机的设计，结果表明，它们是快速有效和准确的，具有实际的工程应用价值。     

Reinforcement learning for reactive power control

This paper presents a Reinforcement Learning (RL) method for network constrained setting of control variables. The RL method formulates the constrained load flow problem as a multistage decision problem. More specifically, the model-free learning algorithm (Q-learning) learns by experience how to adjust a closed-loop control rule mapping states (load flow solutions) to control actions (offline control settings) by means of reward values. Rewards are chosen to express how well control actions cause satisfaction of operating constraints. The Q-learning algorithm is applied to the IEEE 14 busbar and to the IEEE 136 busbar system for constrained reactive power control. The results are compared with those given by the probabilistic constrained load flow based on sensitivity analysis demonstrating the advantages and flexibility of the Q-learning algorithm. Computing times with another heuristic method is also compared.     

A design method for continuous deadbeat control systems

Deadbeat control of discrete‐time systems has attracted particular interest because it ensures setting the deviation response to zero within finite time and affords fast responsiveness. By using controllers which include a delay element, however, the deadbeat characteristics can be realized in the continuous‐time domain, and such a tracking system has been obtained. In this paper, we propose a new design method of continuous deadbeat control systems without using delay elements. Our new controller is composed of a digital deadbeat compensator using state feedback and a series smoothing filter. Illustrative numerical examples are presented. The method proposed here is very simple compared with other methods and it seems to be useful for practical applications. © 1999 Scripta Technica, Electr Eng Jpn, 127(2): 42–48, 1999     

适用于闭环控制的快速相量计算方法

近年来,大量电力电子设备在电力系统电源、电网与负荷侧广泛应用,导致电力系统跨区域、跨电压等级的系统性连锁故障逐渐增多,亟需精细化闭环控制。这要求同步相量测量装置(SynchrophasorMeasurementUnits,PMUs)在保证测量精度的同时,具有快速的响应速度。针对这一问题,提出了一种适用于闭环控制的快速相量测量方法。该方法分析了传统DFT算法在系统动态条件下的测量误差特性,揭示了时标位置对相量测量精度与上传延时的影响。为减少PMU上传延时,研究了将时标打在时间窗尾部时相量测量误差与动态相量模型参数的规律,提出了相量修正方法,在减少上传延时的同时兼顾了测量精度。仿真测试验证了所提方法的测量精度与上传延时远高于PMU标准对保护控制类PMU的要求,可用于复杂电力系统闭环控制应用。     

行程控制与速度给定方法

阐述了采用八段折线图取代“S”型速度控制对各行程点进行计算及速度给定产生的方法 ,既简化了复杂的控制过程计算又解决了五段折线速度控制中切换冲击的问题。