Research On The Performance Evaluation Of Household Appliance Enterprises Supply Chain Based On Improved BP Neural Network

In the condition of economic globalization, household appliance enterprises have presented its features, such as the variety product of customization, the product of high energy-consumption and intelligentialize, the business diversification, as well as the technology and product changed rapidly in small household appliance. The questions—how to reduce the cost of production and operation, how to improve environmental protection level and competitiveness of the enterprises—become more severe. The traditional way of vertical and simply horizontal development cannot be changed in time to the complex market for household appliance enterprises, which made the enterprises associate with its horizontal enterprises to form a supply chain system according to the market demand. By optimizing the whole operation process of supply chain, it makes logistics,information flow and capital flow operate effectively, and by optimized supply chain it can make raw materials to product for customer soon. As a way to evaluate dynamic alliance of supply chain, performance evaluation of supply chain plays a significant role in calculating supply chain profit, realizing the environment protection target, providing suggestions for managers in the diverse and complex market.On the basis of summarizing the domestic and foreign documents, the paper put forward the questions of establishing the index system, and presented the key factors which were affected by supply chain. At the same time, at the foundation of analyzing index system of supply chain performance, combining with the traditional balanced scorecard(BSC), adding two factors which were environmental protection and the service of E-commerce, the paper put forward a six dimensional balanced scorecard model. In the method of performance evaluation, firstly, the paper analyzed the methods of the different scholars(such as fuzzy analytic hierarchy process method, genetic algorithm, data envelopment analysis(DEA), etc). Secondly, it analyzed comparatively the different ways and models. Financially, BP neutral network was chosen in the paper which is objective and dynamic relatively. In case study, this paper had 18 groups data from the first month in 2014 to the sixth month in 2015 for a household appliance enterprise as the research sample, and made use of the software Matlab2012 b,then built BP neural network model including 3 training functions(traingd, traingdx,.trainscg). At last, the results of the model were obtained. With the minimal number of training steps and the minimum error, trainscg function was best. Some suggestions were put forward to the problems in the household appliance supply chain.The main conclusions of this thesis are:(1) Six dimensions Balanced Scorecard Model were established, including finance, internal processes, learning and innovation, customers, environmental protection, and the ability of E-business service.(2) Improved BP neural network model was chosen as a model of household appliance enterprises supply chain performance evaluation, and the example was proved that the model is feasible.(3) Some suggestions were proposed to the problem of the household appliance enterprises supply chain, which was the cost, environmental issues, and E-business management.