基于BP神经网络的发动机机体工时定额研究

Research on Work Quota for Diesel Engine Body Based on BP Neural Network

针对新发动机机体工时定额难度大,以经验估计为主,且对报工的工时核定难度大,通过以S型机体铣床工序为例,运用BP神经网络模型,从工艺文件的结构参数中抽取部分数据作为训练集,构建了发动机机体工时定额神经网络模型。同时运用线性回归方法对工时定额进行了预测,通过对比分析发现,BP神经网络计算的误差较小,适用于机体加工前期对工时定额的初步、快速估算。

It is difficult for the new diesel engine body to fix the work quota. It is usually estimated based on experience. And it is difficult to verify the work quota. Taking the Milling process of the S-type diesel engine body for example, the method of BP neural network model based on technology file is proposed. The data selected from the part of technology file is taken as the train set. The work quota model is established by using the method of BP neural network and the linear regression. By contrast, BP neural network is better than the linear regression method. It is suitable for quick estimation of the work quota on the body in the early stage.