基于MATLAB的压缩感知重构漏水信息改进算法研究

Research on Improved Algorithm of Compressed Sensing Reconstruction of Water Leakage Information Based on MATLAB

地下供水管道漏水监测时,采集到的漏水信号在传输过程中因节点自身限制,导致丢失部分有用信息,从而影响漏水监测的准确性。通过对该问题提出的改进压缩感知的重构方法,用传感器节点采集地下供水管道漏水声信号,在压缩采样匹配追踪(CoSaMP)算法的基础上,使用自适应方法得到最佳输出信号并改变重构算法中残差的初值,参与计算的残差初值带有更多信息,可得到更加完整的重构结果,从而恢复出更多原始信号的信息,提高漏水监测的准确性。经过多次Matlab仿真实验结果证明,即使在低圧缩比下,该方法也能保留更多有效的采集信息,重构成功率较CoSaMP算法提高了14.17%。

In the leakage monitoring of underground water supply pipelines,some useful information of water leakage signal collected on-site always lost due to the limitation of the node itself during the transmission process,which greatly affects the accuracy of water leakage monitoring.To solve the problem,we propose an improved reconstruction method based on compressed sensing theory in the paper.In the method,the leaking acoustic signal of underground water supply pipes is collected by sensor nodes.Based on the compressive sampling matching pursuit(CoSaMP)algorithm,the self-adaptive method is used for obtaining optimal output signal,and the initial value of the residual in the reconstruction algorithm is changed.Because of the rich information carrying of the initial value of the residual involved in the calculation,the more complete reconstruction result is obtained and more useful information are recovered from the original signal,hence to improve the accuracy of water leakage monitoring.By many times Matlab simulation experiments,it is proved that the method can retain more effective information collected on-site even at low compression ratios and the reconstruction power is increased by 14.17%compared with the CoSaMP algorithm.