Research On Machine Tool Feed Servo System Based On Fuzzy Neural Network Control

Servo products are widely used in machine tool industry, so the prospect of the machine tool industry relies on the servo technology development. At present, the technology of permanent magnet synchronous motor servo system is more mature, able to meet the requirements for high performance servo drive, with permanent magnet synchronous motor as the servo drive device is widely used in the CNC machine tool feed system. The location of the feed servo system requires high tracking accuracy and positioning accuracy, it demands the highest in the CNC machine tool servo system, the performance of permanent magnet synchronous motor servo system control strategy research, will largely influence the performance of the feed system and the control effect is good or bad. Based on this, this paper presents a fuzzy neural network and servo control method used in machine tool feed servo system.First, this thesis introduces the structure of the machine tool feed servo system, classification and the requirement of servo motor. Then starting from the structure of permanent magnet synchronous motor, the rotating coordinate system is obtained under the mathematical model of establishing simulation model all-digital fuzzy ontology, and analyzes the principle and algorithm of space vector pulse width modulation technology. On this basis, the all-digital fuzzy field oriented vector control simulation. Then further study of the fuzzy logic and neural network control theory, gives the design process of fuzzy controller and neural network set up steps. For permanent magnet synchronous motor servo system is nonlinear, strong coupling, multi-variable and time-varying characteristics, on the basis of the vector control of permanent magnet synchronous motor, and puts forward the fuzzy neural network control strategy is used in position loop, it is a kind of fuzzy logic reasoning, adaptive and self-learning ability of the integration of the hybrid intelligent controller, and presents a fuzzy neural network structure and learning algorithm, and then use the MATLAB simulation tools to position loop simulation under different control strategies, draw a all-digital fuzzy position loop servo system by using fuzzy neural network controller, will have better response speed and performance.Through theoretical research and the analysis of simulation results, finally has carried on the system hardware and software design based on DSP, select produced by TI company TMS320F2812 chip as the control core of the whole system.The hardware design mainly includes the main circuit, control circuit, position control mode design and buttons and the display circuit design;And the software design includes the main program and interrupt subroutine, SVPWM subroutine and power protection interrupt subroutine.In the end, the article through the work are summarized, further shows the subsequent work arrangement of the subject and the vision of the future. In conclusion, based on the fuzzy neural network control of machine tool feed servo system quick response, high control precision, good stability, high performance motor drive system is realized.