基于改进支持向量机的发电企业滥用市场力违规识别

Market Power Abuse Identification of Power Generation Enterprises Based on Improved Support Vector Machine

目前我国正在稳步推进电力市场改革,为了构建健康高效的电力市场,迫切需要加强市场主体的信用评价。对发电企业滥用市场力的违规行为进行识别是电力市场主体信用评价的一项关键性工作。传统的发电企业滥用市场力违规行为识别主要依赖于专家决策,随着市场交易量越来越大,专家决策无法满足工作需要,必须提出适用于计算机分析的智能识别方法。首先分析了发电企业滥用市场力的原理,提出了滥用市场力的量化定义。然后考虑到电力市场具体的数据特点提出了改进的支持向量机发电企业违规识别方法,并将其与定义结合起来形成了一个系统的识别方法。最后采用电力市场仿真实验构造训练集,进行训练和测试。测试结果表明发电企业滥用市场力行为能够准确识别出来,验证了所提方法的有效性。

As China is steadily advancing power marketization,it is urgent to strengthen the credit evaluation of market players in terms of market power abuse to maintain a healthy and efficient power market.Traditionally,identification of market power abuse mainly relies on experts’judgement,which does not live up to the expectation of the increasing market volume.Thus,this paper put forward an intelligent identification by computer analysis.Firstly,we defined market power abuse by quantitatively studying its cause.Then,we proposed an identification method based on improved support vector machine(SVM)considering the data characteristics of the power market.The definition and method above together form a systematic identification method.Finally,we conducted power market simulation experiment to construct a training set for training and testing.The test results have showed that the proposed method is accurate and effective in identifying market power abuse committed by power generation enterprises.