Study Of The Fast Quotation System For The Products That Are Small Batch And Variety

The production mode of the contemporary enterprises has changed from single and massproduction to small batch and variety production in the current market competition. Thetraditional quotation methods based on experience and historical data can’t meet the needs ofmodern production’s any more. The uncertainty of the customer demands lead to series ofuncertainty such as product research and production which require the enterprises to master thecustomer demands timely and make adjustments any time. The fast quotation system is the verymeans by which the enterprises can response to customer demands and make better decision.Consequently, it is of much signification to study the fast quotation system based on thecustomer personalization. Based on this, this paper takes the fast quotation system for the smallbatch and variety production mode as research object which is of very important realisticsignificance.Firstly, this paper analyzes and compares the domestic and abroad studies of quotationsystem utilizing the literature analytic method. And it is found that the fast quotation systembased on product module is more suitable to the small batch and variety production mode thanother quotation systems. In the wake of the comparison of many cost estimation methods’advantages and shortcomings, this research applies BP neural network in estimating the cost ofunknown module and then builds matrix of customer demands’ weights. The paper quantitativelyanalyzes the correlation between the customer demands and technical requirements. Theimportance of technical requirements is gained by applying the house of quality in the qualityfunction deployment. The aim is to make mapping between the customer demands and technicalrequirements and at the same time make mapping between the technical requirements andproducts functional modules. Based on this, this research optimizes the retrieving paths for theknown modules in the quotation system model and proposes optimization solution that retrievingseparately from two ends of the product structure tree. When analyzing the product quotation forthe unknown modules, this research takes elements of the unknown modules mapped bydemands as the input and takes the cost of the unknown modules as the output according to thenonlinear principle of BP neural networks so as to design the net structure. The design of thenetworks’ parameters and the operation are implemented through the neural networks toolbox ofMATLAB. The quotation for the product is available when the costs of the known and the unknown modules are input into the model. Lastly, an empirical analysis is conducted with oneexample of the cost estimation of elevator cars in an elevator manufacturing enterprise. Aftertraining the network with multiple sets of historical data, the paper estimates the cost of theelevator cars. In this way, the practicality of the quotation system is illustrated and the paperanalyzes the research emphasis in next step.