Swarm Intelligence Based Parameter Identification Of Doubly-Fed Wind Power Generator

With the increasing consumption of energy such as coal and fossil fuels, the contradiction between economic growth and energy shortage has become a serious problem, which results in the large scale integration and rapid development of renewable energies such as wind power technologyin power grid. Due to the simple structure, high reliability, reasonable parameter configuration and skilled technology etc, doubly-fed wind power generator(DFIG) has been widely applied as the mainstream model in wind power generation system. In an actual wind farm, the stator and rotor resistance as well as the inductance of DFIG can be easily influenced by some factors such as long-time running and mechanical losses of the generator, which may have a negative effect on in the speed of DFIG. Therefore, this paper mainly focuses on parameter identification methods to obtain accurate motor parameters, summarized as following:(1) Firstly, the current situation of doubly-fed wind power generator parameter identification methods and the basic structure and working principle of the double-fed motorare presented. Then, the mathematical model of DFIG, as a theoretical foundation for the parameter identification, was established under the three-phase static coordinate system, which was transformed into two phase rotating coordinate model through coordinate system transformation. Finally, according to the principle of DFIG based on stator flux vector control theory, inaccurate motor parameters play a negative role on the control performance, which apparently shows the importance of parameter identification.(2) In order to ensure the accuracy of doubly-fed wind power generator parameter and improve the control performance of the generator, a hybrid quantum-behaved particle swarm optimization for parameter identification was proposed. Firstly, a parameter identification model of DFIG at two phase rotating coordinate was established based on the basic mathematical model of DFIG. Secondly, quantum-behaved particle swarm optimization was improved and then mixed with simulated annealing algorithm. Lastly, the proposed algorithm was compared with particle swarm optimization(PSO), QPSO and improved QPSO, which were applied to parameter identification of DFIG in Matlab/Simulink. Simulation results showed that the proposed algorithm can improve the identification accuracy of five parameters including stator resistance, stator inductance, rotor resistance, rotor inductance, and mutual inductance of stator and rotor.(3) In view of the problem of DFIG multiple parameters identification, an improved competitive particle swarm optimization(ICPSO) for parameter identification was represented. In order to overcome the slow convergence rate of competitive particle swarm optimization(CPSO), the improvements of competition mechanism and learning methods are introduced. Superior particles still need to learn from the personal best positions and the global best positions, which can improve the convergence rate and accuracy. The proposed approach was used to identify the parameters of DFIG based on the wind power grid simulation model in the Matlab/Simulink, and the Simulation results showed that the proposed approach outperforms other algorithms such as PSO, QPSO and CPSO.