Research on Feedforward Parameter Identification of Valve-Controlled Cylinder System Based on Kalman Algorithm

The valve-controlled cylinder servo system is an underdamped system with insufficient amplitude margin, which makes overall closed-loop control accuracy low; Also, the open-loop proportional coefficient of the system is a time-varying parameter, which makes it impossible to accurately design feedforward compensation for the system. To resolve above two problems, this paper proposes a feedforward compensation algorithm with parameter auto-tuning function based on Kalman algorithm and local parameter optimization model reference adaptive algorithm. In the paper, we provide the design scheme of the controller and establish a mathematical model, analyze the dynamic characteristics and static characteristics of whole valve-controlled cylinder system, and finally conduct simulation verification. The simulation results shows that the feedforward control has little impact on system stability since the feedforward compensation algorithm we adopt does not rely on the original closed-loop. At the same time, the feedforward parameters have the ability to self tune and can quickly respond to time-varying parameter changes. Ultimately, it greatly improved control accuracy.