Dynamic Supplier Selection Method And System For Multi-Stage Iterative Sourcing Of Power Materials

Safe and steady operation for power system is related to people’s life safety. With the rapid development of economic society, the scale of power grid and equipment demand are dramatically increased, and people also put forward higher request to safe operation of power grids. But grid companies purchase goods to decide whether the power grid safe operation. Grid companies purchase goods to a large number of suppliers, and procurement costs often account for more than 70% of its investment. Power supplies procurement has the following features:(1) Centralized procurement model. The power industry is very important for a country. Through superior centralized control material quality and quantity, power companies can better select vendors and can less spend procurement costs compared with decentralized procurement. (2) Using bidding way. The nature of public infrastructure makes electricity materials purchase often use bidding way. In turn, bidding can improve transparency and reduce integrity risks.By summarizing current power materials centralized bid procurement process, we found the following problems:(1) For large quantities of demand in one bidding, bidder deliberately use lower price to win the bidding. (2) Supplier evaluation system is not perfect, lacking of pertinence. (3) Supplier selection and service processes are static: it will lead to supply fluctuation that the dynamic change of raw material prices and production, so that, suppliers can not ensure the quality of long-cycle supply.To deal with the unsteady supply problem of power materials which is caused by the centralized procurement adopting high-volume and long-period purchasing, a multi-stage iterative revising sourcing model is put forwarded. The main idea is to dynamically adjust the sourcing amount from suppliers based on continuous sampling inspection. (1) According to the economic order quantity model (EOQ), consider the cost, demand, base price and other factors to determine the best bidding batches and divided package. (2) And then begin to stage tender by invited bidding, power material supplier evaluation index system being built based on the actual situation of bidding and Analytic Network Process (ANP) being applied in determining bidding suppliers’weights. (3) By relying on mobile Internet information system, sampling observation of quality is captured real-timely in the each tender stage access/delivery quality sampling processes, and then using the current and historical data to predict the power materials quality of future supplies’. A dynamic standard package allocation model that aim at maximizing the total value of purchasing (TVP) as well as minimizing the total cost of purchase (TCP) is built to finish the bid evaluation and scaling.A demonstrative case is adopted to exemplify the usage of the above model, and a series of sensitivity analyses are then conducted to provide instructive guidance for both material purchaser and suppliers. At last, to ensure the data of quality sampling reliably, a mobile internet-of-things-based electric power material monitoring system for is developed to construct a feasible working environmental for the proposed model, and the work of the full text summary and outlook are described.