The Application And Research Of The Modified ACO And DTC System

The direct control and the effect of the torque are the important points for Direct Torque Control (DTC) technology. DTC is a new high performance technology which was developed after the developing of vector transformation technology, in the middle of 1980's. It is taken widely attention to and developed quickly for its new control mind, concise system structure and good performance in both static state and dynamic state.In this paper, the development of the AC electromotor and the DTC system are described overall first, and the study status of the speed sensorless technology is analysed briefly. Then the basic principle of the DTC is introduced in detail. After analysing the shortcomings of the DTC system and the speed sensorless technology, through the simulation expriment, using the speed observer based on BP neural network replaces the traditional speed sensor is advanced in this paper, and its validity is proved. The simulation model of DTC system with speed sensorless is built at the MATLAB/SIMULINK, and the simulation experiment of the change tendency of the flux linkage, the torque, and the speed in DTC system is carried out. The result shows that: the speed of the electromotor can be identified by the speed observer based on BP neural network optimized by quick modified ACO, and the static state and the dynamic state performance of the DTC system can be satisfied. So, the feasibility of this method used in DTC system is proved. That the DTC system with speed sensorless is implemented in theory.On the base of the theory analysis and the simulation study, at the intelligent synthesis empolder test-bed whose control core is the digital signal processor(DSP) TMS320F240, the software of the algorithm is designed, and the experiment of the asynchronous motor DTC is implemented. The software of the system, the neural network algorithm programme based on DSP is achieved by mix programming with C and assembly language. At the same time, the speed and the torque of the motor with load are collected and observed. After the union debugging of the hardware and the software of the system, the result shows that: the asynchronous motor DTC system with neural network speed observer based on DSP takes on good performance, and it is feasible, also it provides a useful reference for the application of DTC system in project.