Generator stator end part winding structure optimization method based on particle swarm algorithm and support vector machine

Abstract

The invention provides a generator stator end part winding structure optimization method based on particle swarm algorithm and support vector machine, which comprises the following steps of: (1) carrying out finite element analysis, recording design variables and structural responses of each analysis as input variables and output variables respectively to obtain a sample point required for establishing an approximate model of a support vector machine; (2) using the obtained sample point, taking SVM parameters as design variables and carrying out initialization; (3) carrying out SVM parameter selection to establish a support vector machine approximate model with optimal parameters to predict structural response; (4) carrying out optimization combined with the approximate model of the support vector machine by using PSO as an optimization algorithm to obtain an optimal solution of the approximate model; (5) substituting the obtained optimal design variable values into the support vectormachine approximate model to call a finite element program for calculation, and comparing the calculation result with the finite element accurate calculation result, and judging the optimal solution.According to the invention, an accurate approximate model is established by using the SVM instead of the finite element analysis, and the optimization efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of complex structure optimization algorithms, in particular to a method for optimizing the structure of a generator stator end winding based on a particle swarm algorithm and a support vector machine. Background technique [0002] With the development of modern science and technology, in order to rationally use energy, improve economic benefits, protect the environment, and better meet people's production and living needs, domestic and foreign power systems are increasingly developing towards large units, ultra-high voltage and long-distance power transmission. Therefore, the capacity of the power grid continues to increase, and the capacity of a single unit also increases accordingly. Large capacity and large output of generators are an inevitable trend in future development. Large-scale generator sets are the core of the power system, and the two are interdependent and interdependent. The scale of the pow...

AI Technical Summary

Problems solved by technology

The finite element of the end winding structure is usually relatively complex, and the structural response obtained through finite element analysis often takes a long time. In the optimization process, it is necessary to continuously call the finite element process for iterative calculation, which increases the time cost of the optimization process.

Images