| As an ancient price discovery mechanism and resource allocation measure, auction has already last for two thousand years in history, owns its unique attraction. With the development of the practice and theory, auction has been widely used in fresh agricultural products market. The result of auction is influenced by strategies of all participant, agricultural production have its own diversity and perishability, so agricultural products auction is full of uncertainty, and fuzzy theory has its unique advantage in dealing with uncertain problem. This paper proposes a fuzzy logic based strategy in continuous double auction which stabilizing the volume and the price of auction market.The paper focuses on the decision of continuous double auctions. First,we propose a new algorithm based on the recent developments of fuzzy auction decisions models, market environment was described by membership function, then the algorithm redivide the space of the input and output of fuzzy reasoning, put forward a series of reasoning rules. A new fuzzy logic strategy (NFL-S) was proposed by improving the bidding mechanisms and algorithms with fuzzy logic. The new NFL-S is adaptive and do not need to adjust the bidding decision parameters frequently. With the characteristics of relative membership function, unsensitive parameter, dynamic space of input and output, adaptive risk attitude, optimal bidding strategies are proposed by the heuristic fuzzy rules and fuzzy reasoning. Second, this paper simulates auction environment by MATLAB and its fuzzy toolbox. The numerical simulation experiments show that our new strategies are superior to several other commonly used strategies. By changing the experimental set-up, a detailed description of the performance is put forward in different market size, competitive environment, and policy structure. From the simulation results this paper draws the following conclusions:In equilibrious market, NFL-S gets most profit in different auction size.In adverse market environment, NFL-S gets most profit in rational strategies, but less than irrational strategy. With the growth of NFL-S percentage, more and more profit were obtained by NFL-S group. Finally, this paper discusses the application of NFL-S in fresh agricultural auction market. Different products have different time requirements on auction, by adjusting time parameter in simulation, NFL-S was confirm that it has superiority under time-constraint. According to the conclusion of experiment analyses, some suggestions is put forward for development of the fresh agricultural auction market. |
