基于Zeuthen决策与贝叶斯学习的发电权双边交易模型

A Generation Rights Bilateral Trade Model Based on Zeuthen Strategy and Bayesian Learning

能源供应紧张与环境污染严重是中国社会的基本现状,通过发电权交易,可在一定程度上缓解能源与环境问题,同时也为发电权买卖双方创造了获取额外利润的可能性。买卖双方为了获得更高的收益,均希望交易价格趋向己方的期望,由此构建了发电权交易双方的谈判模型。首先,双方针对先验知识估算对方的交易底线,并结合对方的报价对对方的交易底线进行贝叶斯学习,更新对方交易底线的概率分布。其次,基于Zeuthen决策对比双方的最大风险接受度,并作出保持原有报价或者按最小让步幅度改变报价的决策。最后,通过多轮谈判,双方报价将逐步收敛形成交易价格。研究结果表明,交易价格接近纳什积最优解,与谈判的理论最优结果较为一致。

The increasingly intensified energy supply gap and ever severe environmental pollutions are key issues in China.Trading of power generation right could offer a solution to alleviate the above problems and allows for extra profit among the contracting parties.Setting the price is a critical part of the trading process,because the contracting parties both want to maximize their own benefit.A negotiation model is established for this purpose.Firstly,both parties estimate the base price of the opponent according to the prior knowledge,and then update the probability distribution using Bayesian method.Secondly,the risk tolerance of each party is tested and compared via a Zeuthen decision model,and a decision is made whether to remain the price unchanged or to change it within the smallest concession range.Finally,the final contracting price is made over several rounds of negotiations.Study results show that the price calculated via the above model is in keep with the theoretical optimal solution of Nash product.