摘要
研究一种基于模糊逻辑推理的多响应稳健参数优化方法。利用满意度函数计算含噪声干扰过程的各个响应的满意度值,然后经过模糊逻辑推理,将多个响应的满意度值转化为模糊推理等级;进而根据主效应法选取最优参数组合。构建包含可控因素和噪声因素的神经网络预测模型,对最优参数组合处的噪声实验进行预测,得到优化参数的响应质量指标值;并且计算出优化参数的信噪比。将该方法运用于铣切削工艺过程多响应参数稳健优化。结果表明:基于模糊逻辑推理的多响应稳健参数优化方法得到的最优参数组合,不仅有效地降低了铣切削表面的粗糙程度和粗糙高度,而且对噪声因素的影响具有较好的稳健性。
A robust parameter optimization method for multi-responses was investigated based on fuzzy logic inference. The desirability value of each responses with noise interference is calculated by desirability function method. Then using the fuzzy logic inference to convert the desirability values of multi-responses into a fuzzy reasoning grade,and the optimal parameter combination is selected according to the main effect chart. A neural network prediction model with controllable factors and noise factors is constructed to predict the optimal parameter combination under the noise interference. The response quality index of the optimized parameters is obtained,and the signal to noise rate( SNG) of the optimized parameters is calculated. This method is applied to optimize the end milling process for multi-responses. The results show that the optimal parameter combination of multi-response robust parameter optimization based on fuzzy logic inference not only effectively reduces the surface roughness and height roughness,but also the results have a good robustness under the noise interference.
出处
《科学技术与工程》
北大核心
2018年第6期142-149,共8页
Science Technology and Engineering
基金
国家自然科学基金(U1404702)
航空科学基金(2014ZG55021)
河南省科技攻关计划(162102210083)
郑州航院大学生科技创新基金(Y2016L09)
郑州航院研究生教育创新计划基金(2017CX014)资助
关键词
模糊逻辑推理
满意度函数
多响应过程
稳健优化
fuzzy logic inference
desirability function
multi-responses process
robust optimization
