基于正交试验-遗传神经网络的陶瓷球面精密磨削参数优化

针对陶瓷等难加工材料的精密加工要求与特点以及球面磨削传统加工模式,分析了氮化硅陶瓷材料球面廓形工件砂轮法向跟踪精密磨削的方法。采用正交试验法设计试验，运用极差法和方差法综合分析相关磨削工艺参数对工件加工质量与效率的影响规律。考虑到当前磨削加工工艺方案选择与优选的难点，利用遗传神经网络算法建立了工件加工质量与效率和相关磨削工艺参数之间的非线性映射关系，并基于正交试验法的分析结果对遗传神经网络算法进行了改进,实现了相关磨削工艺参数的优化，缩短了氮化硅陶瓷材料球面廓形工件数控磨削工艺制定与操作的时间，提高了磨削加工质量和效率。

Parameter Optimization on Precision Grinding of Ceramic Sphere Using Orthogonal Experiment and Genetic Neural Network

According to the requirements and characteristics of ceramics difficult-to-machine materials' precision processing, and traditional processing mode about spherical grinding, the grinding wheel normal tracking was analyzed for the silicon nitride material's spherical profile workpiece of precision grinding. The orthogonal test method was applied for the design of experiment, and the range method and variance method were used to analyse the influence of related grinding parameters on the machining quality and efficiency comprehensively. Considering the difficulties of current grinding scheme's selection and optimization, a nonlinear mapping relationship was established between the machining quality and efficiency and related grinding parameters, with the use of genetic neural network algorithm. Based on the analyses of the orthogonal test method, the genetic neural network algorithm was improved. With the genetic neural network algorithm, the related grinding parameter optimization was realized, thus to shorten the process formulation and operation time of silicon nitride ceramic material's spherical profile workpiece CNC grinding, and to improve grinding quality and efficiency.