| As market competition intensified, manufacturers are facing the shorting of product life cycle and unpredictability of the market changes. In order to meet the market demand, manufacturers tend to adopt multiple specification and small batch production. Due to the product variety, small differences between species and states change quickly, logistics control of workshop becomes difficult. Material feeding of workshop is the key to ensure smooth production, and the process is that, according to demand, timely and accurately to deliver material to the workstation which needs feed material. Therefore, at assembly workshop of adopting multiple specification and small batch production, finding efficient methods of material feeding has an important significance.This paper's research field is material feeding at assembly workshop by adopting multiple specification and small batch production. This paper puts forward a dynamic material feeding model which solves the problems of feeding material untimely and inaccurately due to lack of real-time information of production, based on real-time information dynamic material distribution model through analysis of the traditional material feeding method. Firstly, according to the distribution mode, material is divided into three categories: JIT material; buffer material; standard material, and puts forward the corresponding dynamic feeding algorithm: accurate feeding algorithm; buffer feeding algorithm; rough feeding algorithm. Then, based on dynamic real-time process of production material distribution and logistics transparency, using RFID to collect real-time information of production and material demand, integration of production plan, production process information and materials, the solution feeding material to workshop by dynamic feeding algorithm referred above. This solution effectively controls the flow of raw materials, and figures out the problem caused by traditional material feeding method. It can reduce production cost, improve the quality and improve production efficiency. Finally, aiming at the vehicle scheduling problem, this paper uses an improved genetic algorithm to solve the problem, and analyzes its experimental results. Based on the above research, the author designs a RFID-based dynamic material feeding sub-system, which has been applied in a motorcycle company in Chongqing. The effect of the application shows that the research of this paper is feasible and effective. |