基于最大Lyapunov指数的改进预测模型及其在城市用水量短期预测中的应用

Improved Predicting Model Based on the Largest Lyapunov Exponent and Its Application to Short-Term Forecasting for Urban Water Consumption

参照计算Lyapunov指数的Wolf方法,考虑预测中心点与邻近点和上一个演化点的夹角,对混沌理论基于最大Lyapunov指数的预测方法进行了改进.通过对城市用水量短期预测的实例研究,将改进算法与传统算法进行比较.结果表明,与传统算法相比,改进算法的预测精度在整个预测周期内提高了10.2%,在最大可预测时间尺度内提高了1.1%.

According to the Wolf algorithm, the traditional predicting method based on the largest Lyapunov exponent was developed, in which the angle formed by the center point with the adjacent point and the evolutionary point was considered. The improved method was compared with the traditional method in the case of short-term forecasting for urban water consumption. Results show that the predicting precision of the improved method is 10.2% up on that of the traditional method in the whole forecasting period, and 1.1% in the forecasting time scale maximum.