The Research On Differential Evolution Algorithm And Its Application In Motor Parameter Identification

Differential evolution(DE) algorithm is applied extensively in many engineering field owing to its advantage of simple theory,few parameter and easily to be achieved.However,like many other intelligent optimization algorithm the problem of premature convergence easily exists in DE due to the species diversity decreased.To improve the DE’s disadvantage of premature convergence easily and the poor local search ability,this paper makes some improvements on differential evolution algorithm according to its characteristic,artificial immune system is inserted in differential evolution process then an immune differential evolution algorithm using clone selection is proposed in this paper.PMSM is applied extensively in modern industrial field because of its feature of simple mechanism,small volume and well performance. It is very necessary to identify the parameter of PMSM because motor parameter identification is extremely important to perfecting the motor model.What’s more,whether the parameter of PMSM can be identified accurately will affect the performance of the motor system and the accurate monitoring of motor status.In this paper,the motor parameter identification problem would be transform into a optimization problem,the new algorithm proposed in this paper will be used to solve the parameter identification problem of PMSM and the ability of this algorithm to solving the complex engineering optimization problem will be showed.The main contents are outlined as following:1. Firstly,in this paper the basic principle of DE is researched carefully,then to search the breakthrough of improving DE,the reason of DE falling into local optimal value easily and DE’s characteristic are analyzed.Secondly,in order to enhance the optimizing ability of DE,the influence of control parameters and mutation strategies on the differential evolution algorithm is researched and the most applicable control parameters and mutation are selected through simulation experiment.2.In order to solve the defects of differential evolution algorithm,a novel hybrid optimization algorithm base on Artificial immune system and differential evolution si proposed.In this algorithm,clonal selection and receptor edit mechanism are inserted into the differential evolution process to enhance the convergence speed and the population diversity of the algorithm.What’s more,local enhancement operator is used in the late of evolution to further improve the precision of the algorithm. According to the experiment based on six standard function tests,the simulation results show that this algorithm not only effectively avoids the premature convergence,but also significantly improves the global optimization ability and convergence speed.3.To prove the algorithm researched in the paper has some engineering application value,we apply it to PMSM parameter identification and temperature monitoring.Firstly,a full rank identification model is established according to motor mathematical model,the parameter identification problem is transformed into an optimization problem.We use the strong global optimization ability of the algorithm search the optimal solution quickly and accurately,then the motor parameters will be identified.The result of many experiment prove that the algorithm researched in this paper not identify the motor’s several parameters accurately under the condition of motor temperature variation,but also achieve online monitoring of the temperature changes.