基于压缩感知的电力监控系统研究

Power Monitoring System Based on Compressive Sensing

使用无线传感器网络对电网进行监控能够节省大量的人力成本并提高数据采集实时性,然而实时采样产生大量数据会给通讯链路带来压力,同时也会增加节点能量消耗。由此提出了一种压缩传感模型,采用一种改进的两步迭代阈值算法压缩电力数据,通过减少数据采样率降低对数据链路和通讯系统的压力并有效减少时延,并在此基础上进行数据重构。实验结果表明,相比于现有BP计算复杂度和IST的低收敛速度,该模型能够有效降低数据采样率并降低节点功耗。

Using WSN to monitor the grid can save a lot of manpower costs and improve data collection in real time. However, the real-time sampling produces large amounts of data and will put pressure on the communication link and also increase energy consumption. Therefore, a compression sensing model is proposed. An improved two-step iterative threshold algorithm is used to compress the power data. By reducing the data sampling rate, the pressure on data links and communication systems is reduced and the delay is effectively reduced. And the data reconstruction is performed on this basis. The experimental results show that,compared with the computational complexity of back propagation(BP) and the low convergence rate of iterative soft thresholding(IST), the model can effectively reduce the data sampling rate and power consumption.