Research On The Tools Consumption Forecasting Based On The Artificial Neural Network

Rational use and consumption of tools place importantly, which in production management, occupation of liquidity, cost and product quality. The preparations of traditional fixed consumptions are by the fixed members in factories, and the calculation is manual. These works are time-consuming and laborious to enterprises, which are multi-species production. These enterprises can not meet the market competition.The Artificial Neural Network model of tools consumption is constructed in this thesis. Draw on traditional methods, the model have six input variables. These variables are the number of processed products, tool life, cutting speed, tool wear rate of accidents, production volume, and technical level of workers. Based on the construction method of Artificial Neural Network, the BP (error back-propagation) and RBF (radial basis function) Artificial Neural Network model of tools consumption are constructed. Samples was trained and tested in Data Processing System, which was collected from tools section of ZMJ.In addition, use the traditional method of technical calculations to calculate tools consumption is mentioned in this thesis. Prediction accuracy of Artificial Neural Network model higher than traditional method is proved in this thesis, by comparison the forecast results of technical calculations and Artificial Neural Network model.