The Study Of Design Optimization For Dual Stator-winding Wind Power Induction Generator

The Dual Stator-Winding Induction Generator (DWIG) is a new type of ACmachine which is capable of regulating the excitation reactive power continuously. Itnot only inherits the advantages of simple structure, high reliability and easymaintenance of cage rotor induction machine, but also overcomes the disadvantages oftraditional induction generator system, such as large capacity and volume of powerelectronics, abundant harmonic components. Therefore, there is a broad applicationpotential of DWIG in the field of independent power supply and wind power generation.However, it is lack of the research about DWIG’s optimal design in wind power systemat present. This thesis focuses on the research of the cage rotor DWIG from the aspectsof electromagnetic design and optimization algorithm respectively.Firstly, the basic structure and the operating principle of DWIG as well as itspower generation system are briefly introduced. The magnetomotive forces are analyzedin detail. Then, the equivalent circuit of DWIG is deduced, the power balancerelationship and the calculation methods of control winding current are derived.Secondly, combined with the electromagnetic design characteristics of the DWIG, theelectromagnetic design program is developed by the platform of Visual Basic6.0andthe initial design scheme is obtained. And then, the impacts of main parameters to theperformances of DWIG are discussed and a better design scheme is obtained. Thedesign results provide the basic parameters for DWIG’s optimal design. Thirdly, thecharacteristics of the wind power DWIG are analyzed, and the operation characteristicsat rated speed as well as variable speed are discussed in detail. Furthermore, theselection of excitation capacitor in the optimal design of DWIG is obtained. Fourthly,based on the introduction of the basic features and the current improvements of ParticleSwarm Optimization (PSO) algorithm and Genetic algorithm (GA) respectively, a newkind of Genetic-Particle Swarm Memetic Algorithm (GPSMA) is proposed, and thevalidity of GPSMA is verified by4classic Benchmark functions. Finally, the DWIGoptimal design model is established combined with the characteristics of wind powergeneration system. And then, the GPSMA is applied to optimize the electromagneticdesign of DWIG, and the optimization results are analyzed.The results show that the derived electromagnetic design formulae and operatingcharacteristics calculation methods of DWIG are correct and effective. The proposed GPSMA can not only keep the advantage of fast convergence of PSO algorithm, butalso overcome the disadvantage of premature convergence of PSO algorithm effectively.It has a stronger global optimization ability, and suitable for complex multivariableoptimization problems. The obtained optimization schemes of DWIG reduce the controlwinding current or cost or increase the efficiency respectively, which suggests that theoptimization objective has been achieved.The study of this thesis provides the theoretical support for the electromagneticdesign optimization of DWIG, and the optimized design schemes provide the referencefor the practice design of a DWIG prototype.