混合智能技术在激光淬火工艺优化中的应用

The appliance of mixed intelligent technique to optimization of procedure parameter of laser surface quenching

为探索激光淬火工艺优化设计的新方法,以GCr15激光淬火为例,首先建立神经网络模块,以提取激光淬火(多)工艺参数与(多)性能指标之间的函数映射关系,然后建立遗传算法模块,通过遗传算法的进化搜索来实现激光淬火工艺参数的优化设计.实验结果表明,本研究建立的工艺优化设计模型具有较好的可靠性;将神经网络与遗传算法的混合智能技术引入激光淬火领域,为解决激光淬火工艺优化设计问题提供了一条先进、合理的途径.

The mixed intelligent technique based on neural network and Genetic Algorithms has been used to the field of laser surface quenching, in order to find a better way of designing and optimizing the technological parameters. Firstly, we build a neural network module, which abstracts the function mapping between multi-parameters and multi-targets of laser surface quenching. Then, we build the genetic algorithms module. We accomplish the design and optimization of technological parameters by evolving and searching of genetic algorithms module. The results show that these modules have fine accuracy. This technigue provides an advanced and reasonable way for optimizing the technological parameters in laser surface quenching.