基于数据可视化的电力数据融合及优化研究

Research on Power Data Fusion and Optimization Based on Data Visualization

为提高电力系统智能数据融合性,使系统软件服务更加强大,在原有数据融合基础上,引入BP神经网络技术,形成基于BP神经网络的电力数据可视化融合软件系统,利用其功能和结构特征对数据进行融合,并对系统性能进行测试。研究发现:利用矩阵迭代法对神经网络进行训练能够达到更好的收敛效果;设计算法相比其他算法,风电场功率预测值更接近实际值,误差值更小,数据处理精度更高;BP神经网络能够从容应对大数据的处理,当数据量达到100GB时,处理时间更短,算法优势更明显。设计的算法在电力数据融合处理方面具有较好的性能,对软件服务优化起到一定的促进作用,为电力大数据信息融合提供了新的机遇。

In order to improve the intelligent data fusion of the power system and make the system software services more powerful,based on the original data fusion,BP(Back Propagation)neural network technology was introduced to form a power data visualization fusion software system based on BP neural network.The data was fused using its functional and structural features,and the system performance was tested.The following results are obtained:using the matrix iteration method to train the neural network can achieve better convergence effect.Compared with other algorithms,the predicted value of wind farm power is closer to the actual value,the error value is smaller,and the data processing accuracy is higher.BP neural network can deal with the processing of big data calmly.When the amount of data reaches 100GB,the processing time is shorter and the algorithm advantages are more obvious.Therefore,the designed algorithm has good performance in power data fusion processing,plays a certain role in promoting software service optimization,and provides new opportunities for power big data information fusion.