基于支持向量机的大电网运行风险量化评估

Quantitative assessment of large power grid operation risk based on support vector machine

文章建立了综合考虑电网结构、运行方式、设备工况、运行环境和保电需求的电网运行风险评估指标体系。基于风险诱因、发生概率、影响对象、风险后果、持续时间等方面因素从电网的安全、经济、优质3个方面对电网运行风险进行全面分析。利用支持向量机(support vector machine,SVM)在有限样本下能较好地对非线性模型进行映射的优点,构建了大电网运行风险量化评估模型,并利用专家层次分析法自适应调整SVM决策函数的精度。算例结果验证了该方法的合理性。

The risk assessment index system of power network is established, which considers the structure, operation mode, operating environment, equipment condition and power demand. Based on the risk factors in- eluding the risk incentives, occurrence probability, impact object, risk consequence, duration and so on, the comprehensive analysis of the power network operation risk is carried out in view of the network security, e- conomy and high quality. The support vector machine （SVM） is used to map the nonlinear model in finite ~mple, the model of risk assessment for large power grid is constructed, and the expert-analytic hierarchy process（AHP） is used to adaptively adjust the accuracy of decision function of SVM. The results of numerical examples show the rationality of the proposed method.