Font Size: a A A

Multi-station Lightning Location Algorithm Optimization Based On Grid Search Method And Lightning Location System Design

Posted on:2019-10-17Degree:MasterType:Thesis
Country:ChinaCandidate:H LiuFull Text:PDF
GTID:2428330551456794Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Lightning is a serious natural calamity.Monitoring the occurrence of lightning and studying its activity routines are of great significance to the analysis of lightning activity characteristics,as well as lightning protection and disaster mitigation.At present,the multi station time difference location is the main method used in lightning detection at home and abroad.Traditional method that based on solving multi station nonlinear time difference equations has problems such as uncertain initial value,unstable iteration and low detection precision.In recent years,there appear new methods based on time difference region search algorithm that can improve the detection precision.However,there remain many problems such as high time and space complexity,and low algorithm efficiency.This thesis proposed a multi station lightning location optimization algorithm based on grid search,which was built on the existing grid search lightning location algorithm.Using this algorithm as the core processing method,designed and implemented a lightning location and information query&release system.The main contents of the paper are as follows:1.Proposed a multi station lightning location optimization algorithm based on grid search.It combines artificial fish swarm algorithm optimizing SVM(AFSA-SVM)and multi-dimensional hybrid spatial index for fast lightning location.The algorithm first divides the detection area into grids,uses the OKD tree to store the time difference vector of each grid to the detection stations which greatly shorten the query time and reduce the space to build the index.Secondly,the AFSA_SVM classifier is trained for every grid in the detection area to improve the classification accuracy.Finally,when there is a need to do the lightning location,the nearest neighbor search of the OKD tree is used to find the alternative grid,and the AFSA SVM classifier on the grid is used to identify whether it is the point.This algorithm can effectively improve the accuracy and real-time performance of multi station lightning location.Good results have been achieved when using this algorithm as the core algorithm in the lightning location module of proposed system.2.Designed and implemented a lightning location and information query&release system based on WEB-Spring architecture(Lightning Location And Information Public System,LL-IPS).The system mainly includes the data receiving communication module,the lightning location module and lightning information release module integrated by user authority management,lightning data real-time display,historical lightning data statistics and analysis,data backup and recovery,site monitoring and log management.Among them,the data receiving communication module is based on the Netty technology,which receive the lightning signal sent by each detection point server in real time.The lightning location module is responsible for generating the lightning location record and stored them in the system database.It uses the lightning information within certain threshold from the stations and call AFSA-SVM-GLLM algorithm to do lightning location.The module of data query and display is based on the Autonavi GIS,displaying the lightning data set of specified area at a time interval.3.Used Monte Carlo simulation of lightning location to verify the effect of AFSA optimizing SVM on lightning location.As well as verifying the practicability of the proposed system.
Keywords/Search Tags:Multi-station lightning location, grid search, mixed spatial index, AFSA-SVM
PDF Full Text Request
Related items