Research On Intelligent Order Decision Support System

The intelligent order decision support system of manufacturing enterprises is an indispensable component of intelligent manufacturing.Its development and promotion are of great significance in improving enterprise decision-making efficiency,reducing decision-making cost and reducing wrong decision-making.The intelligent order decision support system is the only way for enterprises to solve the problem of order acceptance in the order-to-order production(MTO)mode caused by people's individual needs.Although the order decision support system has been developed for many years,it still has problems of inefficiency and inability to adapt to a large amount of data,and its decision1making efficiency still has room for improvement.In this paper,the intelligent order decision support system is built by web crawler technology,LSTM technology,case-based reasoning technology(CBR)and BP neural network technology.The main work is as follows:(1)Based on the CBR-BP estimation model.Based on the current cost and historical data construction order cost and its price estimation attribute model and Scrapy data capture model,combined with CBR technology and BP neural network technology,the CBR-BP order cost and its price estimation model are constructed.And carried out an example verification analysis to fully verify the feasibility of the estimated model,and the estimated error is within 5%,(2)Based on the LSTM estimation model construction,Based on the order cost and its price estimation attribute model and Scrapy data capture model,the LSTM order cost and its price estimation model are constructed,and an example verification analysis is carried out to prove the feasibility of this estimation model.Under the premise of large data volume,it has high efficiency,and the estimation error is within 3%.(3)Based on the web crawler Scrapy model.Based on the design of web crawler technology,the Scrapy-based web crawling and data storage model is built.The raw material price data in the order and the data needed to calculate the producer price index(PPI)are climbed,and the PPI correction is brought by the time change.The price fluctuates.(4)The construction of the future trend forecasting model of the order cost and its price.Using the second exponential smoothing prediction method,based on the LSTM estimation model,the order cost and its price future trend prediction model are constructed.(5)Implementation of intelligent order decision support system.The LSTM estimation model with better estimation effect is used to construct the decision support system.Matlab App Designer is used to realize the main interface construction of the intelligent order decision support system and the integrated display interface for the network crawler and LSTM estimation model.The database is completed.The method library and the model library management function interface are constructed,and the decision function of whether the new arrival order is received by the enterprise is realized by calling the deep learning algorithm(LSTM).