Optimization Of Parameters Identification And DSP Realization In DTC System

The paper entirely summarizes the actuality and development of Direct Torque Control(DTC) system of induction motors,and particularly introduces the basic theory and principles.Then a parameter identification method based on particle aptitude algorithm for DTC system is brought forward and is proved its validity through simulation and experimentation.As a new and efficient AC speed control scheme,Direct Torque Control has merits lying in the simplicity of the control structure,fast torque response,better robustness and being more prone to digital control.Although it has defect in the theory and performance that not good at low speed operation and having serious pulsating torque.Identification of key Parameter is very important to improving system performance,and the development of intelligent control method brings a new way to solve the problems.Combining intelligent control theory with DTC technology,a flux linkage observer and a rotation speed identifier are constructed based on BP neural network optimized by modified particle swarm optimization algorithm(PSO),which improved parameter observing method in DTC.The old algorithms of BP neural network have defect such as low convergent speed and precision.As a new particle intelligent optimization algorithm, PSO can obviously enhance convergent speed and precision of BP network through optimizing its weight value and threshold value.The optimized neural network can identify parameter of system effectually and quickly.The experiment use TMS320F240 DSP chip which is specialized in motor control system. The newest Intelligent Power Module(IPM) is been used as power instrument,its inner protect circuit raise reliability of the system.To verify the new flux linkage observer through experiments,results show that the scheme is feasible and effective.