Ultra-Short-Term Forecasting of Wind Power Based on 1DCNN-DACLSTM Model
Ultra-short-term prediction of wind power is an important basis for the situational awareness of power grid operation.Taking advantage of the high-frequency sampling of PMU,a novel ultra-short-term power prediction method based on the combination of 1DCNN and DACLSTM is proposed to handle the high dynamic disturbance problem of wind power with random fluctuation on power grid operation.Firstly,PMU with high precision and high-frequency sampling is introduced to measure the ultra-short-term wind power in real time.Secondly,by utilizing the advantages of 1DCNN in feature extraction and temporal convolution reducing computation com-plexity,the key features for wind power and its related factors sampled by PMU can be extracted,and then DACLSTM model is used to analyze the relationship between wind power and the input features.Wind power prediction based on the 1DCNN-DACLSTM model can realize the high dynamic trend prediction of wind power.Finally,the validity and feasibility of the proposed method are verified by tak-ing an actual wind farm with the configured PMU as an example.
国家科技期刊平台
NSTL
万方数据
