電動(dòng)負載模擬器的非線(xiàn)性因素分析及補償
中國測試付夢(mèng)瑤1，2， 楊瑞峰1，2， 郭晨霞1，2， 張 鵬1，2， 張新華3
摘 要：為提高電動(dòng)負載模擬器系統的動(dòng)態(tài)性能和信號跟蹤準確度，提出針對系統摩擦非線(xiàn)性和間隙非線(xiàn)性進(jìn)行補償的方法。分析系統存在的非線(xiàn)性因素及其對系統造成的影響，在此基礎上建立其非線(xiàn)性數學(xué)模型。采用基于小波神經(jīng)網(wǎng)絡(luò )的PID控制器實(shí)現摩擦非線(xiàn)性補償，同時(shí)利用間隙逆模型針對間隙非線(xiàn)性進(jìn)行補償。利用Matlab軟件對補償結果進(jìn)行仿真驗證，仿真結果顯示經(jīng)過(guò)補償后系統正弦響應曲線(xiàn)跟隨性能變好，跟蹤誤差明顯減小，準確度得到很大改善。仿真結果證明：基于小波神經(jīng)網(wǎng)絡(luò )的PID控制器和間隙逆模型分別對摩擦非線(xiàn)性和間隙非線(xiàn)性有明顯的抑制效果，系統動(dòng)態(tài)性能得到提高。
關(guān)鍵詞：電動(dòng)負載模擬器；非線(xiàn)性；摩擦補償；神經(jīng)網(wǎng)絡(luò )
文獻標志碼：A 文章編號：1674-5124（2016）01-0096-06
Analysis and compensation for the nonlinearity of electric load simulator
FU Mengyao1，2， YANG Ruifeng1，2， GUO Chenxia1，2， ZHANG Peng1，2， ZHANG Xinhua3
（1. School of Instrument and Electronics，North University of China，Taiyuan 030051，China；
2. Key Laboratory of Instrumentation Science & Dynamic Measurement，Ministry of Education，
North University of China，Taiyuan 030051，China；
3. Beijing Automation Control Equipment Research Institute，Beijing 100000，China）
Abstract: To improve the dynamic performance and signal tracking accuracy of electric load simulator systems， a method have been proposed for compensating the friction nonlinearity and the gap nonlinearity of the system. Particularly， the nonlinear factors of the system and its impact are analyzed and a nonlinear mathematical model has been established. The friction nonlinearity is compensated with a PID controller based on the wavelet neural network and the gap nonlinearity is compensated through a gap inverse model. The results are verified with MATLAB software. The simulation test shows that， after compensation， the tracing performance of the sinusoidal response curve of the system is improved and the tracking error is largely reduced. Also， the simulation results indicate that the PID controller and the gap inverse model have significantly inhibited friction nonlinearity and gap nonlinearity so as to enhance the dynamic performance of the system.
Keywords: electric load simulator； nonlinearity； friction compensation； neural network
