分数阶区间系统的鲁棒稳定性分析

由于实际的物理系统常会存在各种不确定结构,这些不确定结构给系统的稳定性分析带来了困难和一定的影响,因此对这类带有不确定结构的分数阶系统控制问题的研究,具有广泛的实际应用意义。本文主要针对带有区间不确定性的分数阶控制系统,分析了该系统的鲁棒稳定性问题,首先将区间矩阵用分式线性变换(LFT)表示,然后基于分析得出了此类系统鲁棒稳定的充分必要条件。并给出两个数值算例阐释该方法,仿真结果表明了该方法的有效性。     

A neural net-based time-delay compensation scheme and disturbance rejection for pneumatic systems

In practical electro-pneumatic or hydraulic servo mechanism, there is usually an unavoidable operating delay time in the solenoid valves. This may sometimes cause closed-loop instability if the time delay is not treated carefully in the control system designs. Furthermore, pneumatic actuators can offer high performance at low cost, but are often suffered from external disturbances depending on the operating conditions. This paper proposes a non-model-based design approach for pneumatic actuating systems. In the proposed control system, a gain scheduled fuzzy-PID controller ensures tracking performance an adaptable wavelet neuro compensator compensates for time-delay of the control valve, and a disturbance rejecter diminishes influence from load changes. A condition ensuring the closed-loop stability is derived. The proposed design is experimentally verified to show its effectiveness.     

线性时滞系统稳定性分析综述

时滞在工程领域广泛存在,对此综述了线性时滞系统的稳定性研究方法.从频域和时域两个角度详细介绍了各种方法的特点,着重讨论基于线性矩阵不等式(LMI)的分析方法,指出保守性是分析的重点.对现有结果的保守性进行比较和评述,并提出了改进的思路.     

基于最小二乘支持向量机的非线性广义预测控制

通过中值定理将一类非线性系统近似为时变线性系统,然后将提出的在线最小二乘支持向量机回归(OLSSVMR)与广义预测控制相结合,提出了一种基于OLS-SVMR的自适应直接广义预测控制.利用OLS-SVMR直接设计预测控制器,并基于广义误差估计对控制器参数和广义误差估计中的未知向量进行自适应调整.理论证明了该方法可使广义误差估计值收敛到原点的一个小邻域内.仿真算例也验证了该方法的有效性.     

Stability analysis of neutral systems with mixed delays

The stability of neutral systems with mixed delays is studied in this paper. A new discretized Lyapunov functional method is proposed. The method in this paper is less conservative than the existing ones and the results are very close to the analytical results. In addition, this method allows the coefficient matrix of the neutral term to have time-varying uncertainties.     

A new delay system approach to network-based control

This paper presents a new delay system approach to network-based control. This approach is based on a new time-delay model proposed recently, which contains multiple successive delay components in the state. Firstly, new results on stability and H_∞ performance are proposed for systems with two successive delay components, by exploiting a new Lyapunov-Krasovskii functional and by making use of novel techniques for time-delay systems. An illustrative example is provided to show the advantage of these results. The second part of this paper utilizes the new model to investigate the problem of network-based control, which has emerged as a topic of significant interest in the control community. A sampled-data networked control system with simultaneous consideration of network induced delays, data packet dropouts and measurement quantization is modeled as a nonlinear time-delay system with two successive delay components in the state and, the problem of network-based H_∞ control is solved accordingly. Illustrative examples are provided to show the advantage and applicability of the developed results for network-based controller design.     

Theoretical justification of approximate norm minimization method for intelligent digital redesign

This paper investigates the theoretical justification of intelligent digital redesign (IDR) methods based on the approximate discrete-time models. Its main purpose is to design the digital fuzzy controller achieving the minimum norm distance between the trajectories of the approximate closed-loop systems under the analog and the digital controls. It is verified that, if there exist the digital gains such that the approximate norm distance is sufficiently small, the redesigned controller makes the trajectories of the exact closed-loop systems close, or, at least, asymptotically stabilizes the plant.     

Stability analysis of neutral systems with distributed delays

The stability of neutral systems with distributed delays is investigated in this paper. A modified Lyapunov-Krasovskii functional is constructed to study this class of systems. The proposed stability criterion is discrete-, distributed- and neutral-delay-dependent. In addition, by this method one can study the case when the coefficient matrix of the neutral delay term is time-varying uncertain. The reduced conservatism is illustrated in a numerical example.     

Adaptive IIR/FIR fusion filter and its application to the INS/GPS integrated system

Motivated by the complementary features of the IIR-type filter and the FIR-type filter, this paper proposes a robust IIR/FIR fusion filter and an INS/GPS integrated system designed with the fusion filter. In the fusion filter, an IIR-type filter (SPKF) and a FIR-type filter (MRHKF filter) are processed independently, and then the two filters are merged using the mixing probability calculated using the residuals and residual covariance information of the two filters. The merits of the SPKF and the MRHKF filter are integrated and the demerits of the filters are diminished through the filter fusion. Consequently, the proposed fusion filter shows robustness against model uncertainty, temporary disturbing noise, large initial estimation error, etc. The stability of the fusion filter is verified by showing the closeness of two filters in the mixing/redistribution process and the upper bound of the error covariance matrices. This fusion filter is applied to an INS/GPS integrated system. The performance of the INS/GPS integrated system designed using the fusion filter is verified through a simulation under various error environments and is experimentally confirmed.