Identification Detection And Early Warning Control Of Danger Sources Of Transmission Line Construction Sites Based On Internet Of Things

Since 2015.the State Grid corporation has been doing a lot of thinking with the Internet.This year,the State Grid Jiangsu Electric Power CO.LTD "two sessions" simultaneously proposed that high-end power grid should be built in the strategic goal of new era development.At the meeting,the provincial company said that the implementation of the strategic goal of new era should focus on the four strategies,one of which is to build high-end power grid.In the background of the continuous development of technology such as Internet of things,large data and cloud computing,We should use new technology for technological innovation in time,support the transformation of grid mode of operation and management way,serve the broad masses of customers better,and enhance the social sense of responsibility continuously,so that we can create high-end and smart grid good,meanwhile,play a positive role in good operation and development of the company.As a central part of the power grid,the transmission line should be strong,self-recoverable and reliable.The traditional way of safety supervision is to conduct personnel patrol.The advantage of this way is that the thinking mode of personnel is netlike,and people can judge the danger more accurately.However,this method has obvious disadvantages.The reliability of the route is dependent on the staff’s own experience,and the staff cannot inspect the whole line all the time and which lacks flexibility.Based on the current situation,this paper proposes the monitoring and warning projects of transmission line hazard source,using sensors to transmit data,using neural network algorithm to simulate the process of human brain judgment and depending on system return data for management and assessment.The main work and achievements of this thesis are as follows:Firstly,based on the work process of the transmission line inspection,this paper puts forward an improvement scheme on the current management mode of the line inspection from the aspects of vehicle management and personnel management through analysis.Secondly,vehicle identification is required in vehicle management.On the basis of the comparison between YOLO and SSD algorithm,an improved YOLOv2 target detection scheme is presented in view of SE module combination.After the training data and experimental analyss,the scheme has improved the accuracy of detection.Thirdly,based on the Internet of things technology,the overall design of the risk control system for construction sites is given.According to the system requirements,the network design of hardware modules including WIFI probe,camera with 4G and license plate recognition function,infrared human induction switch and inductive audio and visual alarm lamp is given and the software design strategy is provided from the aspects of platform architecture and platform function.Fourthly,the early warning control platform of danger sources of transmission line construction sites is realized on the basis of B/S structure(Browsr/Server).The user interface of the model is simple,and the system functions are implemented in the server,which makes the development,operation,maintenance and uses of the system uncomplicated.In order to meet the functional requirements,the APP of the real-time danger sources pre-control system is also developed,so as to fill in the processing situation on site.After six months of actual use,the control platform is basically used to meet all the proposed requirements,and the system is expected to be updated intelligently in the later stage,so as to achieve zero manpower in the work link besides field work.After half a year’s practical use,the platform basically meets all the requrements put forward,and the system is expected to be updated intelligently in the later stage,so as to achieve zero manpower in the work link except the site.