Winner Determination Problem Of Reverse Auction Based On Prospect Theory

Reverse auction,as a new procurement method,plays an important role in many economic fields.More and more governments and enterprises began to use online reverse auction to purchase,which can reduce procurement costs and shorten the procurement time,and suppliers can also reduce inventory costs,expanding the market,so the reverse auction allows for both sides to achieve a win-win situation.When buyers face a large number of suppliers to bid,they need to determine the winning supplier,so the Winner Determination Problem(WDP)is one of the key issues in the reverse auction.At present,a lot of research on the winner determination problem is based on the assumption that buyers have rational decision-making behavior.However,the buyers have psychological expected values about procurement costs and procurement time,which shows that the buyers have limited rational decision-making characteristics.How to make the winner supplier on the cost and time in accordance with purchaser's psychological preferences,is one of the important factors for supplier selection decision.At the same time,the prospect theory effectively depicts the psychological characteristics of decision-making in uncertain situations,so my research is based on the prospect theory to study the winner determination problem of reverse auction.First of all,before the auction link,evaluate each supplier to determine which suppliers is qualified to bid,and according to the suppliers evaluation score eliminate some suppliers which is no bidding qualification.According to the actual situation analysis,we select nine indicators,which are divided into four categories:product quality,historical transaction record,service level,and financial status of the company.Using the fuzzy analytic hierarchy process calculate bid qualification score of each supplier,and then based on the value of each supplier's bid qualification score to determine which suppliers is qualified to bid.Secondly,this paper studies the problem of single product multi-item winner determination considering buyer' s behavior about procurement costs and procurement time.We respectively base on cumulative prospect theory,prospect theory and utility value theory to depict buyer's behavior,and then establish three different kinds of mathematical model.Then,according to the characteristics of the problem model,and based on the basic ant colony algorithm,a hybrid ant colony algorithm with maximum and minimum pheromone control and dynamic parameter adjustment is designed,and we use the hybrid ant colony algorithm to solve the model based on cumulative prospect theory.According to three different size instances,we analyze the algorithm parameters to obtain the optimal combination of parameters.Then,we respectively compare the solution with enumeration algorithm and the basic ant colony algorithm.And this proves that the improved ant colony algorithm is better than other algorithms,and it is more accurate and suitable for different size instances.Finally,we analyze the winner determination problem of reverse auction considering the buyer's psychological preference about time and costs.In the three different instances,we analyzed the influence of the expected cost and expected time of the buyer on the final supplier selection scheme,and obtained the influence of the expected cost and expected time of the buyer on the final supplier selection scheme.Then,we use three different kinds of models to solve three different instances and make these results contrast and analyze each other.The result of instances shows the mathematical model based on cumulative prospect theory can better reflect the actual decision-making.