Research And Application Of Discrete Enterprise Of Production Forecast Based On APS

In order to achieve the dynamic management,the enterprise adopts Advanced Planning and Scheduling technology, it is a optimization plan, it can consider real constraint conditions such as time, order, inventory, etc. Get all kind of dynamic changes at any time to adjust production to cater to the market changes in enterprise production. To help enterprises solve cannot real-time material requirement of dynamic balance and resource problems, provides effective support for manufacturing enterprises. By using a mathematical model to predict production plan to provide reference for the enterprise, it has great practical significance in the APS. But enterprise’s production forecast model has complex, multilayered, openness, non-static characteristics. It is very difficult to build model.Based on the discrete enterprise background, the enterprise is a manufacturing valve company, it is used more batch、small batch-multiple batch production mode,make the enterprise more flexible to environmental change. To meet customer requirements to achieving maximum interests under the minimum cost, we must adopt more effective production organization way, and reasonable use of inventory. In order to improve the enterprise’s own existing APS system, enterprises will establish discrete production line production prediction as the main direction of enterprise development.The paper through access to enterprise database to get the historical data of valve parts, found that the data on the production line has no regularity, contain noise, complex and coupling between the variables features. First of all, get the original data for data cleaning, data integration, normalization processing. Using the BP neural network, RBF neural network algorithm to yield prediction model, by using the genetic algorithm has global searching and fast convergence speed characteristics of the optimized model, analyzes the simulation results, finally chose a kind of optimal algorithm is applied in the enterprise, and played a major role in the actual scheduling of the APS.