Research On Contour Error Control Technology Of Linear Motor Based On Neural Network

With the rapid development of science and technology, the development of CNC machines towards high-speed and high-precision. Linear motors as a kind of structure of direct drive have unique advantages of speed and accuracy. As a result, linear motors have been used in CNC machines, and begain to replace the traditional servo mechanism composed with rotary motor and ball screw gradually. It produces two kinds of errors in the process running. One is tracking error of single-axis system, and the other is contour error of multi-axis system. Two errors especially contour error affect the contour accuracy of the workpiece and the feed velocity directly. It is important to achieve the contour error control.This paper’s focus on contour error control technology for XY platform of CNC machines driven by linear motors. Firstly, it describes the project’s research status and development prospects at home and abroad. Introduces the basic principle of linear motors, and sets up its mathematical model. Secondly, the theoretical basis of neural network control is analyzed, and a control method for single-axis servo system of linear motors is designed. Lastly, thie paper establishes the contour error model. Analyzes and confirms the factors affecting the contour error by experiments, and designs a control method for multi-axis system.The main research results of this paper are as follows:(1) Combing with the mathematical model of linear motors, designs three closed-loop feedback control system based on the conventional PID control principle for single-axis. Quoting intelligent control theory, designs the BP neural network PID controller. Simulation experiments confirm the system based on BP neural network PID controller has higher response speed, better stability, and higher tracking accuracy than which based on the conventional PID controller.(2) Combining with the contour error model, the general formula of contour error expressed by tracking error is obtained. The analysis and calculation are carried out from the motion of the linear interpolation and circular interpolation, which summarizes the influence of open-loop gain coefficient and speed on contour error. It is shown in simulation experiments that the more mismatch of open-loop gain coefficients, the larger the contour error, the contour error is proportional to the speed of motion.(3) Researching on contour error control technology, a variable gain cross-coupling controller is designed for the system. Combining the neural network control theory, a neural network cross-coupling controller is designed. It is shown in simulation experiments that the contour error of system based on neural network cross-coupling control is the lowest.