Equal Interval Compression Algorithm Based on Cosine Angle Dynamic Recognition for Wind Power Time Serial Output

In the electrical system, the wind power is generally measured every 15 min, resulting in 35040 time series datain one year. It will be a time-consuming task to calculate the subsequent optimization and predication. Therefore, this paper proposes a strong random time domain signal flow equal interval compression algorithm based on the cosine angle dynamic recognition. This method extracts the extreme points in the original time series by dynamically calculating the cosine angle, and fills the data with the original data equal interval under the premise of retaining the original volatility to the greatest extent. It maintains the characteristics of the original data acquisition, and replaces the original time series, which is conducive to subsequent analysis and research. Taking 35040 wind power data of Ningxia in 2019 as an example, the data were compressed by using 1 h, 2h and 3 h as the time interval respectively. The results showed that the similarities of the power probability distribution in these three intervals were above 0.997. More apparently, when a time interval of 1 hour was used, the data was reduced to 8117 points from 35040 points. The power probability distribution similarity is 99.95%. It indicated that the method presented in this paper can effectively reduce the data and maintain the important information of the origin wind power data.