The Research Of BLDCM Servo-Control System Based On DSP

The brushless DC motor (BLDCM) is a novel motor. It's control system has excellent characteristics of small volume, high efficiency and economy, and shows great developing potentiality. The digital-signal-processor (DSP) has many advantages such as high precision, reliability, etc. TMS320LF2407A possesses integrated function of controlling BLDCM, it can simplify control circuits, lower the cost, and increase reliability of systems. The hardware circuit of this control system only includes power supplying unit, electrical insulation unit, driver, current feedback unit, etc. The rest of parts are realized by the software. The BLDCM drive system based on a DSP TMS320LF2407A has been researched in this thesis.In this thesis, the development and tendency of BLDCM are described at home and abroad. An brief introducetion to structure and operational principle of BLDCM. And according to the mathematical models the brushless DC motor is modeled by using Matlab/Simulink. The research on the dynamic performance of different operating conditions is carried out.The design procedure of control system and driving system are introduced in this thesis. The TMS320LF2407A DSP minimum system is designed as the main micro-controllor, which includes power management, expanded RAM and COM interface, etc. The driver consists of IGBT-IPM, drived by speedy light-sensitive components. Circuit for current sampling is given to measure the feedback signal of current and voltage.When designing the software, the main function, "speed- current" double closed loop control, are complete. Using DSP, the PWM signal, signal acquisition, reverse control and speed control are achieved successfully. At last, the experimental results of sample motor are presented.In this thesis, a hybrid N-PI speed-variable which combinating the advantage of neural network and PI control strategy is proposed for controlling the speed of BLDCM. The speed-variable can overcome the problems of over-regulating and instantaneous oscillationby updating PI parameters online based on the self-learnning and self adaptationadaptive functions of neural network. The experiment of this control method is carried out in Matlab/Simulink, a good performance of speed tracking has been achieved.