Research On Intelligent Cover Anti-Theft And Damage Monitoring System Based On Sound Spectrum Pattern Recognition Technology

Well covers play an important role in ensuring the improvement of underground facilities and protecting the safety of pedestrians and vehicles.However,the phenomenon of stealing and damaging well covers has occurred in recent years.Theft of well covers not only seriously threatens the safety of life and property of passing pedestrians and vehicles,but also it is an illegal act that destroys public property.As an important member of the intelligent municipal structure,the intelligent well cover can suppress the occurrence of the phenomenon of stealing to a certain extent,but its vibration sensor can not effectively distinguish the normal vibration caused by pedestrians and vehicles from the abnormal vibration caused by theft,and it is easy to produce false alarm.Through observation,the sound produced in the two vibration states is different,so it is feasible to use the difference of sound to monitor and identify the damage of the cover.According to the principle of sound recognition technology,this paper first analyzed the types of feature parameters that can be extracted from sound signals,compared the real-time,accuracy and easy-to-hardware requirements of the Internet of Things system,selected appropriate feature parameters,and decided to use the multi-layer screening method to recognize sound signals;similarly,the existing recognition algorithm is compared based on the above requirements,and an improvement scheme is proposed according to the deficiency of the selected algorithm in combination with the requirements of the Internet of Things.Finally,after selecting the characteristic parameters and recognition algorithm,the sound identification and theft-proof monitoring system of the Internet of Things is designed,and the feasibility and reliability of the whole monitoring platform are verified by experiments.The main work of this study is as follows:1)Comparing the feature parameters that can be extracted by the sound signal from the recognition rate,the difficulty of extraction and whether it is easy for hardware implementation,and finally selecting loudness,frequency and timbre to recognize the sound of the well cover;using 3-step screening method to recognize the sound can not only ensure the recognition accuracy,but also greatly reduce the recognition time.2)There are some shortcomings in the application of traditional DTW algorithm to sound recognition,such as insufficient recognition rate and recognition speed is not fast enough.This paper improved the accuracy and recognition speed of DTW algorithm by increasing the number of templates,adding distortion threshold and improving distortion contrast strategy.3)Designed intelligent well cover anti-theft and damage monitoring system based on sound recognition technology.The system used STM32 platform as the core,and used decibel detection module,audio module and wireless transmission module.The software used LabVIEW graphical programming environment as the development platform,and made the human-computer interactive control interface.The main functions of the system include parameter setting,real-time data acquisition and storage.The part of feature parameters and pattern recognition is processed by using MATLAB software.Since the traditional intelligent covers can not effectively distinguish the normal and abnormal vibration,this paper solved the problem by using the method of sound recognition based on the difference between the two vibration sounds.In this paper,loudness,frequency and timbre are used as characteristic parameters to perform 3 layers of sound screening and recognition.In timbre recognition part,improved DTW algorithm is used to discriminate.An intelligent monitoring system is designed and manufactured,which is for acoustic signal identification and theft-proof damage of well covers based on IoT.Experiments showed that the system can achieve the goal of sound recognition,basically meet the application requirements of theft-proof damage of well covers,and can be used for other sound-based applications,and can provide certain data reference and technical support for other intelligent municipal and security systems based on sound recognition.