Forecasting Research And System Design For The Auto Parts Demands

Automotive industry can be seen as the pillar-based industries of the national economy, it plays an important role in the developing of the economy. But auto parts is the basic of the automotive industry, in recent years, as the rapid growth of the automotive industry, the auto parts industry in our country is also expanding rapidly, and there is more competitive in this industry. At the same time, as the level of information improving continuously, the auto parts demanding data which the business manager grasp is increasing fast, for the purpose of keeping reasonable inventory and improving the efficiency of funding, the forecasting of auto parts demand is necessary.Based on the current situation of the forecasting of auto parts demand, and combining the existing material of vehicle equipment management system, this paper researches the theory and methods of the forecasting of the auto parts demand.First, this paper analyzes the forecasting methods in and abroad the country and the current situation of the forecasting of the auto parts, and it introduces the methods and basic theory of the demand forecasting, then, it presents the traditional forecasting models, the related requirements and current situations of the creation of the present forecasting models.On this basis, this paper analyzes the forecasting for the auto parts demands. According to the different category features of the auto parts, we can divide it into Special Purpose Vehicle Parts, General vehicle parts, Seasonal accessories and Accessories with regional. The author adopts the combined forecasting method to implement the model structure of the forecasting for the auto parts demands. This paper establishes Multi-dimensional linear regression prediction model, Gray system forecasting model, Exponential smoothing forecasting model, Seasonal forecasting model, Neural network forecasting model according to different types of auto parts, and then, it verifies the effectiveness of the model using experiments.For the purpose of improving the forecasting accuracy, this paper adopts the combined forecasting method that the IOWGA operator based on to apply in the forecast system of auto parts demand. Based on the actual historical data and large simulation experiments, the author selects each type of individual forecasting methods with weighted combination, and establish the related forecasting model for the auto parts demands. Simulation results show that, choosing the different forecasting method for different type of auto parts is reasonable, the combined forecasting method in this paper is better than single forecasting methods, which improves the accuracy and reliability of the forecasting.In the last, this paper studies the structure of the forecasting subsystem of auto parts demand and composition as well as function of each module. In the SOL Server2000environment, using C#development system, the author achieves the computation of the forecasting algorithm in the background using MATLAB, and establishes the forecast subsystem of auto parts demand, which was based on the IOWGA operator.