基于软测量与遗传算法优化的电磁电器设计技术

Design technique of electromagnetic apparatus based on soft measurement and genetic algorithm optimization

本文利用样机的性能测试数据,建立了具有全局映射的神经网络软测量模型,并将该模型与遗传算法优化方法有机结合,形成基于软测量与遗传算法优化的电磁电器设计技术。该项技术不仅解决了由电器动态过程数值分析的复杂性带来的遗传算法计算工作量大的问题;而且在优化设计过程中,充分体现了产品的性能状况,为"一次性"设计奠定了技术基础。

The soft measurement model based on neural network is established by the application of sample's testing characteristic data. By the integration of above model and generic algorithm optimization method, a new design technology of electromagnetic apparatus based on soft measurement and generic algorithm optimization is come into being. This technology not only solves the great deal of computing for generic algorithm, which is due to the complex numerical analyses of dynamic process of electromagnetic apparatus, but also embodies the characteristics of products in the process of optimization design, which establishes the technical base of 'for once design'. 