Warning System For Preventing Construction Machinery Damage Of Underground Pipeline Based On Vibration Signal Analysis

Today,underground pipelines such as cable,water,communications and gas criss-cross each other.Due to the lack of effective warning device on the ground above the underground pipeline,and the lack of contact with municipal departments,some construction units do not understand the layout of the underground pipeline,so as to arbitrarily construct,which increases the "digging rate" of "cutting belly".Therefore,the protection of underground pipelines from external damage has become an urgent problem to be solved by municipal departments.In view of the above problems,this paper studies and designs a kind of underground pipeline anti-damage warning system of construction machinery based on vibration signal analysis.When the construction signal of construction machinery is detected,the sound and light alarm will be triggered.The system has the advantages of high early warning rate,simple installation and all-day real-time monitoring.Firstly,the propagation theory of seismic wave is introduced,and the related characteristics of Rayleigh wave,the main energy carrier of seismic wave,are analyzed.Then,the acquisition principle and noise source of vibration signal are described.The acquisition signal mainly includes pedestrian passing,vehicle passing and ground vibration signal during excavator construction.Secondly,aiming at the vibration signals generated by ground moving targets are easily affected by noise,median filtering is first used to remove baseline drift of signals,and then wavelet packet denoising is applied.Compared with EMD denoising and wavelet transform denoising methods,it is found that wavelet packet denoising retains the effectiveness of signals and has a higher signal-to-noise ratio.Then feature extraction is carried out for denoised signals,and feature vector groups are constructed and normalized,so as to facilitate the training and prediction of subsequent recognition models.Then,the sparrow algorithm（SSA）is used to find the optimal parameters of KELM model: penalty factor C and kernel parameter S.The results show that the SSA-KELM model has a good effect on vibration signal recognition.The model can effectively identify the ground vibration signals of pedestrians,vehicles and construction machinery during construction.Finally,a warning system against construction machinery damage based on vibration signal analysis is designed,including hardware composition and the design of data acquisition,data processing,identification and warning software.