采用神经网络模糊滑模控制的液压驱动活塞位置控制研究

针对液压驱动活塞运动轨迹和压力精度跟踪误差较大问题,设计了神经网络模糊滑模控制,并对控制精度进行仿真验证。创建了液压缸驱动平面简图,推导出液压缸腔室内部参数变化方程式。分析了液压缸驱动压力和位置的变化,采用线性模型建立输入和输出变换方程式。引用滑模控制方法,采用神经网络算法对滑模控制进行逼近,通过模糊切换规则对滑模控制进行自适应调整。采用MATLAB对液压缸活塞轨迹和腔室压力跟踪进行仿真验证,并且与滑模控制输出效果进行比较和分析。结果表明:采用滑模控制方法,液压缸活塞运动轨迹和腔室压力跟踪误差较大;而采用神经网络模糊滑模控制方法,液压缸活塞运动轨迹和腔室压力跟踪误差较小。采用神经网络模糊滑模控制方法,液压缸控制系统自适应调节能力较强,从而提高了活塞运动轨迹和腔室压力跟踪精度。

Research on Positioning Control of Hydraulic Driving Piston with Neural Network Fuzzy Sliding Mode Control

In view of the large tracking error of hydraulic driving piston motion and pressure accuracy, a neural network fuzzy sliding mode control is designed and the control accuracy is verified by simulation.The driving plane diagram of the hydraulic cylinder was created, and the internal parametric variation equation of the hydraulic cylinder chamber was deduced. The change of driving pressure and position of hydraulic cylinder was analyzed, and the transformation equations of input and output were established by using linear model. Using the sliding mode control method, the neural network algorithm was used to approach the sliding mode control, and the fuzzy switching rule was used to adjust the sliding mode control adaptively. MATLAB was used to simulate and verify the hydraulic cylinder piston track and chamber pressure tracking, and the output effect was compared and analyzed with the sliding mode control. The results show that the tracking error of piston motion track and chamber pressure is larger by using sliding mode control method, while the tracking error of piston motion track and chamber pressure is smaller by using neural network fuzzy sliding mode control method. By using the neural network fuzzy sliding mode control method, the hydraulic cylinder control system has a strong adaptive regulation ability, thus improving the piston motion track and chamber pressure tracking accuracy.