负荷预报的混沌时间序列分析方法

某些非线性确定性系统表现出混沌行为，由这些系统产生的时间序列又表现出随机性，但是，这些混沌时间序列在短期内可有较为精确的预测，可以为其建立确定性的预测模型。为了探讨负荷时间序列中的混沌特性，作首先介绍计算时间序列中的多种混沌征数的方法，在发现华中电网发电负荷时间序列中的混沌特性的基础上，给出一种基于混沌理论的局部预测算法。该预测方法的优点是，如果时间序列比较长，就可以通过某些参数的调整来控制预测精度。最后以华中电网的发电负荷为例检验了本中的算法，最终的预测模型的预测效果良好。

CHAOTIC TIME SERIES ANALYSIS TO LOAD PREDICTION

Certain deterministic non-linear systems may show chaotic behavior. Time series derived from such systems seem stochastic, however, these chaotic ti me series can be predicted precisely in a short term, thus, by which a determini stic model can be constructed. To search chaos in load time series, the algorith m of several chaos indexes in time series is presented firstly. Based on the ris ing of chaos in generation load time series of Central China Power Grid, a local prediction model is established based on chaos theory. The advantage of this me thod is that the predicted error can be controlled by adjusting some parameters if user has a very long time series. Taking the generation load time series of C entral China Power Grid for example, the presented algorithm is checked and the predicted result is perfect.