神经网络用于大规模电力系统暂态稳定控制决策

Transient Stability Control Decision Based on Neural Network in Large Scale Power Systems

提出了应用线性神经网络进行特征抽取,并与各子神经网络进行统一训练的PRP共轭梯度法法,将神经网络用于大规模电力系统暂态稳定紧急控制决策。文中对我国西北电网中某发电厂切机控制决策进行了数值仿真研究,其结果表明神经网络用于大规模电力系统局部暂态稳定控制决策是可行的。

To apply neural network to transient stability control decision in large scale power systems, a linear neural network for feature extraction and the sub neural networks dealing with different situations and the Polak-Ribiere-Poiyak conjugate gradient method for simultaneously training the neUral networks are proposed. The simulation results of generator shedding decision for a real 49-machine 248-bus system illustrate the perspective of the application of neural network to transient stability control in large scale practical power systems.