改进D-S证据理论在电动汽车锂电池故障诊断中的应用

针对电动汽车电池系统的故障采用基于神经网络的改进D-S证据理论组合规则完成诊断过程。为了避免单一途径的诊断可能造成故障漏检误检的状况,决策层采用D-S证据理论组合规则来确定基于BP网络和RBF网络两种故障诊断算法结果。然而为了克服D-S证据理论处理高度冲突证据的缺陷,本文提出了一种基于神经网络改进的D-S证据理论组合规则。首先,采用神经网络对电池故障进行初步诊断,结合网络诊断准确率来分配不确定信息并构造证据体,又引入了证据间的支持矩阵来确定新的加权证据体。然后,把各个焦元的信任度融入D-S证据理论组合规则,从而融合神经网络证据体及新加权证据体。最后,依据决策准则确定锂电池系统的故障状态。通过仿真实验验证了本文提出的改进D-S证据理论融合诊断方法在电动汽车锂电池故障诊断中的有效性。     

Considerations about synchrophasors measurement in dynamic system conditions

This paper focuses on some concepts that are to be considered to ensure interoperability of phasor measurement units (PMU) during dynamic phenomena. Interoperability of devices in dynamic system conditions is needed to allow accurate analysis and, in the near future, control of phenomena like power system oscillations. The paper emphasizes the need to extend the IEEE standard to ensure the measurement latency and accuracy are known during field measurements. It also introduces a sequential logic aiming at providing the user with more robust measurements, together with more information about the measurement quality.     

An actuator fault reconstruction scheme for linear systems

This paper presents an actuator fault reconstruction scheme that can be classified into an algebraic decomposition approach in which the original system is decomposed into two subsystems: a fault-free subsystem and a fault-dependent one. From the result of such algebraic decomposition, the fault estimate is obtained as a set of combinational functions of the system input, the measurement output, the estimate of certain linear functionals involving the state of the fault-free subsystem and the first-order derivative of the measurement output. Here a functional observer and robust exact differentiators are adopted to provide the estimate of the linear functionals of the state of the fault-free subsystem and the derivative of the measurement output, respectively. The proposed method is verified through two numerical examples of seventh-order aircraft and fourth-order crane models.     

Two-layer observer based control for a class of uncertain systems with multi-frequency disturbances

A novel disturbance estimation approach is presented for a class of uncertain systems subject to multiple-sinusoidal disturbances with unknown frequencies. Different from existing results on disturbance observer based control (DOBC), a new methodology with a two-layer observer structure is developed to effectively estimate and reject the disturbances. In the proposed control architecture, an auxiliary observer is derived to generate a disturbance representation in a parametric uncertainty form. Furthermore, the unknown parameters can be reduced to a constant vector with the dimension of the number of harmonic components in the disturbances. Then an augmented observer is designed to estimate the corresponding unknown parameters of the disturbances. As a result, the uncertain systems with disturbances constituting of multiple unknown-frequency sinusoidal components can be controlled within the DOBC framework, where asymptotic stability can be guaranteed. The proposed approach is successfully validated on a robotic manipulating example.     

混合神经网络在变压器故障诊断中的应用

针对变压器故障诊断准确率低的问题提出了粒子群-自组织映射-学习矢量化(PSO-SOM-LVQ)混合神经网络算法。为了获取更加有效的SOM神经网络拓扑结构,首先采用PSO算法对SOM神经网络的权值向量加以改进,在此基础上融入LVQ神经网络,弥补了无监督学习SOM神经网络的不足。这种PSO、SOM和LVQ相结合的混合神经网络算法提高了变压器故障诊断的精度,减少了故障诊断的误差。通过仿真,对SOM、PSO-SOM和PSO-SOM-LVQ这3种算法进行了对比。对比结果表明,PSO-SOM-LVQ混合神经网络算法准确度最高,其故障诊断准确率为100%。由此可见,采用PSO-SOM-LVQ混合神经网络算法可有效提高变压器故障诊断的性能。     

Decentralized adaptive neural control of nonlinear interconnected large-scale systems with unknown time delays and input saturation

In this paper, a novel decentralized adaptive neural control scheme is proposed for a class of interconnected large-scale uncertain nonlinear time-delay systems with input saturation. RBF neural networks (NNs) are used to tackle unknown nonlinear functions, then the decentralized adaptive NN tracking controller is constructed by combining Lyapunov–Krasovskii functions and the dynamic surface control (DSC) technique along with the minimal-learning-parameters (MLP) algorithm. The stability analysis subject to the effect of input saturation constrains are conducted with the help of an auxiliary design system based on the Lyapunov–Krasovskii method. The proposed controller guarantees uniform ultimate boundedness (UUB) of all the signals in the closed-loop large-scale system, while the tracking errors converge to a small neighborhood of the origin. An advantage of the proposed control scheme lies in that the number of adaptive parameters for each subsystem is reduced to one, and three problems of “computational explosion”, “dimension curse” and “controller singularity” are solved, respectively. Finally, a numerical simulation is presented to demonstrate the effectiveness and performance of the proposed scheme.