基于单位门函数小波包的电力负荷预测研究

The Research of Power Load Prediction Based on Wavelet Packet of Unit Gate Function

电力负荷预测在电力系统规划过程中,具有重要的引领和指导作用.传统电力负荷预测方法,结果误差较大,不能准确的反映出电力负荷情况.基于小波包函数的短期负荷预测数学模型,能够完整的表现出电力负荷信号的低频部分和高频部分.本文通过构建长期电力负荷预测模型,对加权型非线性模型,直线回归型,对数回归型等三种模型进行比较,得出加权型非线性模型更能够反应出电力负荷的规律,在电力负荷长期预测方面是一种比较准确的数学模型.通过实际预测结果和误差分析可知,基于小波包函数的电力负荷预测模型能有效地提高电力负荷预测的误差精度.

Power load forecasting has an important role in leading and guiding in the planning process of power system. Using traditional means to predict the power load forecasting, prediction error is large, does not accurately reflect the situation of power load. The establishment of a mathematical model of short-term load forecasting based on wavelet packet function can verify the accuracy of the model in the power load forecasting. Wavelet packet functions can complete exhibit low frequency part and high frequency part of the signal power load. Build a long-term power load forecasting model, the actual consumption nonlinear model, predictive weighted linear regression model, logarithmic regression model, three for comparison, the weighted nonlinear model is more able reactive power load of the law, in the power of long-term load forecasting is a more accurate mathematical model. Actual prediction and error analysis can clearly draw wavelet packet function in power load forecasting error can effectively improve the accuracy. Therefore, the wavelet packet functions can more accurately reflect the power load patterns.