层次分析法和神经网络的电网安全评估

Power grid security assessment combining AHP with neural network

影响电网安全的因素种类繁多,为了使电网安全评估系统的准确性提高,人为主观误差减少,创造了一种由电网、调节运行的运行方式、继电保护装置、电力通信、调度自动化及各种电器设备构成的评估指标体系,使用层次分析法对评估指标体系进行分析,同时使繁多的安全指标简化,通过神经网络的非线性映射和自学能力特性构建神经网络的安全评估模型。实验结果表明,这种模型减少了安全指标的数量,使得安全评估工作变得简单,避免了评估中人为主观因素的影响,提高了安全评估的准确性,使得评估结果更加具有科学性。

The factors influencing on power grid safety are various. In order to improve the accuracy of the power grid secu- rity assessment system, and reduce the subjective human error, an assessment indicator system composed of power grid, opera- tion mode regulating the running, relay protection device, electric power communication, scheduling automation and other elec- trical devices was developed. The analytic hierarchy process （AHP） is used to analyze the assessment indicator system, and sim- plify the manifold safety indicators. The safety evaluation model of neural network was constructed by means of the nonlinear mapping and self-learning ability of neural network. The experimental results show this model can reduce the quantity of safety indicators, simplify the security assessment, avoid the impact of subjective human factors, improve the accuracy of the safety as- sessment, and make the assessment result more scientific.