Revenue-share Under Quality Optimization Of BY Solar PV Power Plants: A Computational Experiment Simulation

The revenue-share with intelligent agents is a complex problem, because of the different attributions, preference and the learning ability during decision making, etc. So the revenue-share with intelligent agents must consider the different agents' preference demand and the effect of accumulation of experience to agents' decision. The computational experiment is a good method to research the complex system problem with intelligent agents. With the microscopic analysis of the system elements, system behavior and the interaction of agents, the researcher can find the regularity of the system evolution from the macro level, which has been widely recognized and applied in many social science research areas. This paper research the revenue-share problem under quality optimization of the BY photovoltaic power(PV) power plant with the computational experiment.First, this paper system analyzed the “investment-construction-operation”, and found the due to the lack of reasonable revenue-share, the contractor and the operator take the goal of maximize their own profit and squeeze the profit of the other as much as possible, which is not only bad for the both, also cause quality problem of the station and is bad for the revenue of the whole system. Then, this paper analyzes the attribution of agents related and the relationship of the game, take the risk of quality optimization, the fairness preference of revenue-share and the learning ability during decision making process into consideration, builds an agent-based model of contractor and operator on revenue-share. Finally we design three experimental scenarios: Only contractor has fairness preference, only operator has fairness preference, both of the two agents have fairness preference, to research about the feasible region, the result and the risk of the revenue-share in medium quality level and high quality level with the Netlogo. The results of the experiment show that the higher the fairness preference is, the smaller the feasible region with more revenue of the agent is. If the operator got high fairness preference, the system would give more revenue to operator and less to contractor, this would lead contractor compress the cost and bigger risk of the system at the same time. If the contractor got high fairness preference, the system would give more revenue to contractor and less to operator, this would lead contractor increase the cost and smaller risk of the system at the same time.