考虑滞后的寡头发电商报价动态模型及其复杂性

Delayed Dynamic Bidding Model and Its Complexity in Oligopoly Power Market

在发电商报价古诺模型基础上,运用有限理性动态与自适应动态,建立了考虑滞后的电力报价动态模型,然后对动态模型的稳定域、倍岔、混沌等进行仿真分析;在此基础上,分别分析了滞后因子与不同报价动态对发电商的平均收益的影响,然后运用状态滞后反馈控制方法,对滞后报价动态模型进行了混沌控制分析;分析结果表明:1)引入滞后因子有利于报价系统的稳定,但并不一定能增加发电商的收益;2)在均衡状态下发电商获得最大的平均收益,在Nash均衡形成前,发电商宜采用自适应动态,但失稳后则宜采用有限理性动态;3)选择合适的控制参数,可维持报价系统Nash均衡的稳定,但若发电商仅考虑短期收益,则没有积极性去采用滞后控制策略。

In the paper, with bounded rational and adaptive dynamics, the delayed dynamic bidding model is presented based on Cournot game, and the stable region, bifurcation and stranger attractor of the Nash equilibrium are displayed respectively by numerical simulations, then the impact of delayed factor and different bidding dynamics on average payment of generators are analyzed. In addition, chaos control of delayed dynamic bidding model is discussed with state delayed feedback control method, and numerical simulations indicated that, 1） delayed factor is useful to maintain the stability of Nash equilibrium, but it is not certain to improve the payment of generators ; 2） the maximal average payment is attained to generators in equilibrium state, and the optimal adjustment strategy of dynamics is to use adaptive dynamics before equilibrium ,but bounded rational dynamics once loss of equilibrium; 3） Nash equilibrium of dylayed dynamic modle enables to be maintainable if suitable control parameter is chosen, but itis diffcult to myopic generator.