Reliability analysis of a model for servo hydraulic system by utilizing Markov chain Monte Carlo method
Electrohydraulic servo systems are comn;only used in industry because of their high accuracy and large payload capacity. For decades modeling and control of such systems have been the focus of research, since the systems are often nonlinear and have parameters difficult to determine. The validity of models has usually been studied by approximative methods based on linearization, which do not properly reveal the success of parameter identification. In the recent 15 years Markov chain Monte Carlo (MCMC) methods have emerged as a tool to create statistical analysis for nonlinear models. In this study the MCMC approach is applied to analyze a model of an electrohydraulic position servo system. The model is identified using physical data with instrumental noise. The identifiability of the model parameters is quantified as probability distributions of the parameters. The model structure is developed until acceptable statistical results are achieved. The results show how the model is optimized and parameters identified. The reliability of the model predictions is analyzed as well.
Junhong Liu Heikki Handroos Heikki Haario Huapeng Wu
Faculty of Technology, Lappeenranta University of Technology, Lappeenranta, Finland
国际会议
杭州
英文
847-851
2009-04-08（万方平台首次上网日期，不代表论文的发表时间）