基于混沌神经网络模型的电力系统混沌预测与控制

由于电力系统的日趋复杂和庞大，电力系统除了因负阻尼引起的低频振荡外，还存在PSS不能消除的混沌振荡的危机。为及早判断和抑制电力系统的混沌振荡，提高电力系统稳定性，根据电力系统的负荷时间序列重构吸引子相空间，计算相空间饱和嵌入维数和最大Lyapunov指数，并以此为指导，建立混沌神经网络预测模型，该模型即便在电力系统负荷含有部分坏数据输入的情况下，仍能对电力系统的混沌特性进行精确地判断和预测。如果判断出系统存在混沌现象，则设计模糊神经网络预测控制器，实现了对电力系统混沌振荡的预测控制。仿真结果表明，该方案对抑制电力系统混沌振荡具有显著效果。

Chaos forecast and control of power systems based on chaos neural network model

Because of power systems being more and more complicated, besides low frequency oscillation by minus damp, there is also chaos oscillation that can not be dispelled by PSS in power systems. In order to early detect and suppress the chaos oscillation and improve power systems stability, a chaotic attractors space is reconstructed by the power system load time series, systemic embed dimension and maximal Lyapunov exponent are calculated in this paper. By above all, a chaos neural network model is constructed to detect and forecast chaos characters of the power systems even by the inputs including part outlier. If there is chaos oscillation in the power systems, a fuzzy-neural forecast controller is designed and forecast control is realized to the chaos oscillation. The validity of the project is proved by simulate results.