短期电力负荷的智能预测方法研究

为了提高电力系统短期负荷预测的精度，提出将基于模拟退火思想的改进粒子群优化（SAPSO）算法和误差反向传播（BP）算法相结合构成SAPSO—BP混合算法用于训练人工神经网络，对短期电力负荷进行预测。经实际算例验证，该混合算法能有效克服常规BP和PSO算法独立训练神经网络的缺陷，其收敛速度快于BP及PSO—BP算法，并且具有较高的短期电力负荷预测精度。

Intelligent Pre-Estimation Method Study on Short-Period Power Load

In order to raise the accuracy of power system short-period load pre-estimation, it was raised to combine SAPSO algorithm with BP algorithm to confi gurate SAPSO-BP mixed algorithm to train artifi cial neural network to carry out pre-estimation for short-peri-od power load. After verifi cation to actual algorithm case, the mixed algorithm can overcome defection of normal BP and PSO algorithm in independent neural network training, its restraining speed is faster than BP and BP-PSO algorithm and with higher short-period power load pre-estimation accuracy.