Research On Nonlinear Reconstruction Algorithm For Mechanical Vibration Signal

Online monitoring and collection of mechanical vibration signal is a key technology in fault diagnostics.The compression acquisition of mechanical vibration signals can break through the limitation of Shannon-Nyquist sampling theorem,this method turn sampling of the signal into the sampling of the information,which effectively solve the problem of traditional sampling methods to generate huge amounts of data as well as does not cause the loss of information.Applying compressive sensing theory into mechanical vibration signal monitoring,the reconstruction algorithm for mechanical vibration signal are mainly studied.The main achievements of this paper are as follows:(1)Reconstruction algorithm for vibration signal is a very important part of compression acquisition vibration signal technology,it directly affect the reconstruction accuracy of the vibration signal.Aiming at the reconstruction of mechanical vibration signal,the adaptability of mechanical vibration signal based on the traditional classical reconstruction algorithm is studied.Firstly,some reconstruction algorithms are studied,the basic principle of OMP(Orthogonal matching pursuit),StOMP(Stagewise orthogonal matching pursuit),SP(Subspace pursuit),BP(Basis pursuit),GP(Gradient projection),on this basis the mechanical vibration signal is reconstructed,the reconstruction results are mainly studied from two aspects: reconstruction accuracy and reconstruction time.The overall relationship between the reconstruction error and the running time of different reconstruction algorithms with the change in the number of measured values obtained by the simulation test,the performance of typical reconstruction algorithms for the reconstruction of mechanical vibration signals is analyzed.(2)On the basis of the above research,in the noiseless environment because of lower computational complexity and higher reconstruction accuracy of StOMP reconstruction algorithm,it is suitable for the recovery of the actual vibration signals,and the value of threshold parametersst has a great influence on the performance of the algorithm.However,the optimal value of threshold parametersst of StOMP algorithm for mechanical vibration signal is not clear.Through theoretical analysis and simulation experiment,the optimal value of threshold parameters for mechanical vibration signals are determined.The simulation results show that the algorithm canachieve the highest reconstruction accuracy under the optimal value.(3)Aiming at the reconstruction of mechanical vibration signal in noise environment,the multipath matching pursuit algorithm(MMP)is studied.The reconstruction accuracy of MMP in noise and noiseless environment is better than some previous greedy algorithms.However,its drawback is that the computation is large and the operation speed is slow,and it is necessary to know precisely the sparsity of signals.So this should be restricted in practical application.In view of the above shortcomings of MMP,a novel sparsity adaptive stagewise multipath matching pursuit method is proposed.SASt MMP makes use of the low computational complexity of StOMP to guarantee the rapidity of the algorithm,at the same time introduces the idea of backtracking,to avoid decline in the accuracy of algorithm reconstruction.The simulation results show that the recovery performance of SAStMMP is similar with MMP,but its overall running time is greatly reduced.