Direct Torque Control Of Asynchronous Machine Based On Model-Based Design For DSP

With the development of power electronic technology and the upgrading of the microprocessor, motor control gradually gets more efficient and optimized. Though DSP is a good tool for digital motor control, there are many weaknesses in traditional DSP programming such as tedious code, the complexity of program modification arid low efficiency. While model-based design is a new programming method which provides a development program for users from the algorithm design to the hardware implementation. There are lots of advantages in this new programming method, such as simpler programming, easier modification and higher efficiency. In addition, it does not need developers to master the knowledge of underlying hardware.The direct torque control (DTC) system has been widely used in AC speed control systems with its simple structure, less dependency to rotor parameters and excellent dynamic and static performance. Based on the analysis of DTC theory, the software design of asynchronous machine DTC system is realized by applying model-based design in this thesis and with the DSP TMS320F2812 as main control chip. Firstly, the mathematical model of asynchronous motor under two-phase static coordinate system is built and the principles of DTC are elaborated in the paper. And then, the hardware platform of DTC system for asynchronous motor is built in this thesis, based on which a simulation model that can analyse and verify the feasibility of the algorithm design is created. Finally, modeling program design of the system is constructed to control the speed of asynchronous motor by integrating the DSP embedded module in the Matlab/Simulink into the validated algorithm by computer simulation. The code generation can be accomplished automatically by Real-Time Workshop (RTW) and downloaded into the DSP board by JTAG emulator.In the experiment, the motor runs smoothly and the starting and braking performance is excellent, so the validity and efficiency of the proposed method is proved and it also provides a new development method and platform for the development of motor control system as well as the research of control algorithm.