Online Monitoring System For Switchgear Based On Cloud Platform

In today’s rapidly developing economy,people are demanding more and more reliable technologies for power supply in their daily lives and social production.As one of the important components of the power system,switchgear plays a vital role in the entire power transmission process,and the number is constantly increasing.As switchgear is in a high current state for a long time,there are often heat failures in it,which can lead to fires and serious safety accidents such as power outages.Therefore,in view of the above-mentioned problems,it is necessary to develop a system to monitor the temperature rise status of the switching device.(1)The traditional monitoring centre lacks the function to analyse and process the data,and only pre-sets thresholds for the monitoring data,which does not enable predictive temperature rise monitoring.Now by improving the Sparrow Search Algorithm(SSA)and then optimising the BP neural network(Back Propagation,BP),we can predict the temperature rise risk level of t he switchgear temperature rise data,and then grasp the internal temperature rise operation law of the switchgear,and make accurate prediction of the temperature rise operation status of the switchgear.The first method is to improve the sparrow search al gorithm by initializing the population with good point sets and introducing non-linear weight factors,combining differential evolution and elitist strategy,and proposing a sparrow search algorithm that combines differential evolution and hybrid multi-strategy(Differential evolution,Elitist strategy,DEH-SSA).The second method is to improve the discoverer position update formula and introduce the iterated local search strategy at the optimal position after the position update,incorporating the golden sine algorithm as an operator into the sparrow search algorithm,and then proposing the Gold-SA Iterated Local Search(GI-SSA)based on the golden sine and iterated local search.Simulation experiments are conducted to verify that the two improved algorith ms perform better and have more competitive advantages.The improved sparrow algorithm is then effectively fused with the BP network to compensate for the shortcomings of the BP network,such as the tendency to fall into local extrema and slow convergence,and to optimise the performance of the network for application to switchgear temperature rise fault risk level prediction.The simulation experiments were carried out using the actual operating temperature rise data of switchgear cabinets with certain sam ples of temperature rise data,and compared with the Particle Swarm Optimization(PSO)algorithm to verify that the improved Sparrow algorithm optimizes the BP neural network for better temperature rise prediction,which lays the foundation for the subsequent invocation of the temperature rise prediction model in the temperature rise cloud monitoring centre.Basis for subsequent invocation of the Forecast Temperature Increase Model in the Cloud Watch Centre.(2)By developing a switchgear cloud temperature rise monitoring centre,combined with switchgear specific application scenarios,the(Long Range Radio,Lo Ra)radio frequency is combined with 4G communication technology to design a network structure with Lo Ra self-organised network sensing and 4G convergence and uploading to the cloud.The CT electromagnetic induction principle is used to realise the isolated transformation of the power supply and the voltage transformation to supply power to the temperature measurement equipment and to realise the function of self supply of power.The software and hardware parts of the terminal equipment are developed to improve the data processing and storage methods of the traditional switchgear monitoring system by using the edge cloud collaborative data processing mechanism.The adaptive threshold algorithm is used to optimise the edge intelligence processing model,which is applied to the hierarchical transmission of monitoring data to avoid the problem of unsynchronised temperature rise data on the display interface caused by the delay in data upload due to the distance between the cloud server and the temperature measurement terminal node equipment.A monitoring centre cloud platform is developed and designed to invoke tem perature rise prediction models in the cloud for temperature rise risk level prediction and predictive monitoring of temperature rise.Through the establishment of an asynchronous connection of Socket and Web Socketde site gateway and Web side communication,using a TCP/IP connection of 4G module to reach the server,for cloud server communication,to achieve the switchgear temperature rise status data flow visualisation display function.Finally,the feasibility and application value of the technical resear ch of the monitoring system was verified by tests of communication and operational energy consumption in the field.