CAJ | 학술논문

针对传统时间序列法的超短期负荷预测模型对波动性较大的负荷预测精度低的问题，提出了一种含误差修正的时间序列法的超短期负荷预测方法。该方法对负荷时序建立超短期时间序列预测模型进行初步预测，并应用BP 神经网络预测时序模型所产生的误差，从而对预测值进行修正。仿真结果表明，与传统时序法相比，改进方法有效地提高了超短期负荷预测的精度。
침대전통시간서렬법적초단기부하예측모형대파동성교대적부하예측정도저적문제，제출료일충함오차수정적시간서렬법적초단기부하예측방법。해방법대부하시서건립초단기시간서렬예측모형진행초보예측，병응용BP 신경망락예측시서모형소산생적오차，종이대예측치진행수정。방진결과표명，여전통시서법상비，개진방법유효지제고료초단기부하예측적정도。
According to the problem of forecasting precision reduced by volatile load of the traditional time series method for ultra short term load forecasting model, a forecasting method of time series ultra short-term load that contains error correction is proposed in this paper. This method establishes the time series prediction model and preliminarily forecast the load, and the error produced by time series method with BP neural network to modify predicted value. The experimental result indicates that compared to the traditional forecasting model, the improved forecasting model can enhance the forecasting precision effectively.