Power Systems Stability and Security Assessments Using Artifical Neural Networks

这项研究正在调查新兴的 人工神经网络技术作为一种强有力工具， 线路稳定性诊断 因为神经网络 可训练，该技术是自适应的，适合在线 应用程序以及规划。 经过充分训练的神经 网络，可以快速评估电力系统的稳定性 计算时间相对较少。 的技术概念 正在研究的是一个大型的，复杂的神经网络是否可以 当电力系统处于稳定或不稳定状态时， 网络被提供有关于系统条件（负载， 代等）。 利用目前的分析工具，这种稳定性 评估需要大量的计算时间。 成功 可训练人工神经网络的实现将导致 系统稳定性和安全性的真实的实时评估。 能力是 电力行业非常需要。

This research is investigating the potential use of the emerging technology of artificial neural networks as a powerful tool for on line stability diagnostics. Because the neural networks are trainable, the technique is adaptive and suitable for on-line applications as well as for planning. With adequately trained neural networks, the stability of the power system can be quickly assessed with relatively little computational time. The technological concept being investigated is whether a large, complex neural net can be trained to indicate whether a power system is stable or unstable when the net is provided with input data on the system conditions (loads, generation, etc.). With present analytical tools, this stability assessment requires massive computational time. Successful implementation of a trainable artificial neural network will result in real time assessment of system stability and security. The ability is very much needed in the power industry.