基于粒子滤波的电液伺服系统故障诊断方法研究

发布时间：2018-03-03 17:00

本文选题：故障诊断　切入点：粒子滤波　出处：《燕山大学》2014年硕士论文　论文类型：学位论文

【摘要】：电液伺服系统是典型的集机、电、液一体化的复杂精密自动控制系统，，应用极其广泛，尤其是在功率重量比大，又要求响应速度快的场合应用更为普遍，难以替代。但同时由于元器件的复杂精密和高度集成导致加工难度高，抗污染能力差，故障时有发生，为此对其进行故障诊断研究具有非常重要的意义。其中基于解析模型的故障诊断方法在解决电液伺服系统的故障诊断问题中得到了广泛应用，但目前这类方法大多都是采取对系统的工作点附近进行线性化来实现故障诊断，而电液伺服系统本质上是一个较强的非线性系统，因此不可避免地会影响故障检测和诊断的准确性。针对现有方法的不足，以及粒子滤波方法在对非线性非高斯问题的处理显现出明显的优越性，为此本文提出将基于粒子滤波的故障诊断方法应用到电液伺服系统中去。本文围绕基于粒子滤波的电液伺服系统故障诊断方法如何实现展开研究，主要研究内容和结论如下：第一、综述了电液伺服系统现有故障诊断方法并分析了其优缺点，同时还总结分析了粒子滤波算法的改进研究现状及其在故障诊断中应用研究现状；第二、详细论述了粒子滤波的基本原理，进而通过实例仿真，对比研究了标准粒子滤波方法与扩展卡尔曼滤波、无迹卡尔曼滤波方法的滤波估计性能，结果表明，不论是非线性高斯模型还是非线性非高斯模型，粒子滤波方法的滤波精度均高于后面两种传统的滤波方法；第三、对比研究了基于粒子滤波的两种检测方法的性能，结果显示基于状态估计和残差平滑的故障检测方法优于基于似然函数的故障检测方法；同时针对通过残差并不容易识别故障的类型问题，将基于信息散度的故障识别方法引入，仿真结果验证了该方法的有效性；第四、以电液位置伺服系统为研究对象，对其建立了非线性模型，进而研究了基于粒子滤波状态估计和残差平滑的故障检测方法和基于信息散度的识别方法应用于对系统典型故障进行检测与识别，仿真结果表明两方法分别能够及时准确地检测故障和识别故障类型；第五、通过液压缸内泄漏故障对基于粒子滤波的故障检测方法和基于信息散度的故障识别方法进行了实验研究，实验结果表明两方法切实有效。
[Abstract]:Electro-hydraulic servo system is a typical complex and precise automatic control system with integration of electricity and fluid, which is widely used, especially in situations where the power / weight ratio is large and the response speed is required. It is difficult to replace. But at the same time, due to the complex precision and high integration of components, it is difficult to process, poor anti-pollution ability, and faults occur from time to time. Therefore, it is of great significance to study the fault diagnosis, in which the analytical model based fault diagnosis method has been widely used to solve the problem of electro-hydraulic servo system fault diagnosis. However, at present, most of these methods adopt linearization near the operating point of the system to realize fault diagnosis, and the electro-hydraulic servo system is essentially a strong nonlinear system. Therefore, the accuracy of fault detection and diagnosis will inevitably be affected. In view of the shortcomings of existing methods and the obvious superiority of particle filter in dealing with nonlinear non-#china_person0# problems, In this paper, the fault diagnosis method based on particle filter is applied to the electro-hydraulic servo system. The main contents and conclusions are as follows: (1) this paper focuses on how to realize the fault diagnosis method of electro-hydraulic servo system based on particle filter. First, the existing fault diagnosis methods of electro-hydraulic servo system are summarized, and their advantages and disadvantages are analyzed. At the same time, the research status of particle filter algorithm improvement and its application in fault diagnosis are summarized and analyzed. Secondly, the basic principle of particle filter is discussed in detail, and the estimation performance of standard particle filter, extended Kalman filter and unscented Kalman filter are compared by simulation. Whether the nonlinear Gao Si model or the nonlinear non-#china_person1# model, the filter accuracy of particle filter is higher than the latter two traditional filtering methods. Thirdly, the performance of two detection methods based on particle filter is compared. The results show that the fault detection method based on state estimation and residual smoothing is better than that based on likelihood function. At the same time, the fault identification method based on information divergence is introduced to solve the problem that the fault type can not be easily identified by residual error. The simulation results verify the effectiveness of the method. In 4th, taking electro-hydraulic position servo system as the research object, a nonlinear model is established. Then the fault detection method based on particle filter state estimation and residual smoothing and the method based on information divergence are studied to detect and identify the typical faults of the system. The simulation results show that the two methods can detect faults and identify fault types in time and accurately. In 5th, the fault detection method based on particle filter and fault identification method based on information divergence are studied experimentally by hydraulic cylinder leakage fault. The experimental results show that the two methods are practical and effective.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TH137;TH165.3

【参考文献】