基于BP神经网络和遗传算法的大功率脉冲电源优化设计
High power supply optimization design based on BP neural network and genetic algorithm
-
摘要: 短电弧脉冲电源的性能指标对短电弧加工工件工艺性有着十分重要的影响.文中通过建立脉冲电源的Matlab仿真模型,利用模型仿真数据对电源神经网络模型进行训练,结合基于目标加权法的遗传算法对电源参数进行优化,指导短电弧样机电源参数的修改,并进行了启动性试验和负载突变试验.结果表明,采用BP-GA算法优化的短电弧电源不仅可以减少了系统调试时间、节约调试和试验成本,并且具有输出电压平稳,受外界干扰小,响应速度快等优点.Abstract: Precision and machining efficiency are the unity of opposites, and ultimately depends on the performance indicators of pulse power supply in short electric arc machining (SEAM). In the article, power supply's neural network model using Matlab simulation data on this basis, the power supply parameters are optimized by combing with genetic algorithm based on objective weighting method to guide the power parameter changes to meet the requirements. Then, test results of simulation and tests, results show that the power supply designed by this optimization method have the advantages of stable output voltage and fast response speed, which meet the expectant targets and test requirements.