Abstract: To overcome the impact of time-varying and nonlinear factors on the performance of pneumatic servo systems, online identification is employed to obtain time-varying parameters, and a Kalman filter is used to estimate the system state. To reduce the impact of disturbances and improve positioning accuracy, an LQG optimal self-tuning controller with integral is designed. Simulation and experiments show that the LQG self-tuning control with integral can effectively suppress disturbances and improve the dynamic and static performance of pneumatic position servo control.
Keywords : Pneumatic position servo system; Linear quadratic Gaussian optimal control; Self-tuning
0 Introduction
There are three types of power drives: electric, hydraulic, and pneumatic. Each has its own advantages and disadvantages, and different drive methods are selected for different applications. Pneumatic systems have many advantages, such as low cost and simple structure. However, in the past, due to the significant influence of temperature, pressure, and volume on gas, pneumatic systems were considered time-varying, inherently nonlinear systems, only capable of point-to-point (PTP) logic control. With the rapid development and application of electronic technology, control theory, and computer technology, servo control can also be achieved using pneumatic systems. Currently, pneumatic transmission and control occupy an important position in the field of automation and are widely used in various automated equipment. Pneumatic servo control is also increasingly being applied in many high-requirement automated equipment.
1. Mathematical Model
Linear Quadratic Gaussian (LQG) control is optimal control using a quadratic performance function as the optimization index. This method is based on a linear model, and the controller parameters are directly related to the model parameters. The aerodynamic system is a time-varying, nonlinear system. For simplicity in analysis, a linearized model is used to approximate the actual aerodynamic system, and real-time identified parameters are used to reflect the system's variable characteristics. According to the references, the linearized aerodynamic position servo system can be approximated as a third-order system. A third-order variable-parameter linear model is used here, employing a controlled autoregressive translational averaging process (CARMA) model.
For details, please click: Research on Linear Quadratic Gaussian LQG Self-Tuning Control of Pneumatic Position Servo System
