Risk Assessment And Prediction Research Of Lightning Disaster In Ningxia

Lightning disaster will bring serious economic losses to the country,and cause great danger to people’s life,production and life.However,due to the different environment in different regions,the impact factors leading to the occurrence of lightning will be different.At present,there is no method to accurately assess and predict the risk of lightning across the country.Therefore,this topic comprehensively considers three aspects of disaster causing factors,disaster environment and disaster resistance ability,uses SPSS principal component analysis method to establish a lightning risk assessment model for Ningxia,and divides lightning risk areas in Ningxia.At the same time,BP neural network is used to predict the occurrence of lightning disaster in Ningxia,and the accurate lightning prediction model is obtained,which can be used to accurately predict the lightning disaster area.This study has practical guiding significance and practical application value for reducing lightning disaster,emergency management and government decision-making in Ningxia and even the whole country.This paper mainly studies from the following three aspects,the specific content is as follows.(1)To assess the risk of lightning disaster in Ningxia,it is necessary to select the factors affecting the occurrence of lightning.Since there are many factors affecting the occurrence of lightning,analytic hierarchy process(AHP)is adopted to give priority to the overall goal of lightning disaster risk assessment,and the risk value of each region in Ningxia is comprehensively studied and assessed from three main aspects:disaster resistance,disaster environment and disaster causing factors.The analysis shows that these three aspects are composed of different impact factors,and the selection of these three aspects of factors will be different in different places.(2)The lightning disaster risk assessment model of Ningxia was established,and the impact factor data of various regions in Ningxia were obtained through statistical analysis.Firstly,the analytic hierarchy process is used to select seven factors:annual frequency of cloud flash,annual frequency of ground flash,average lightning current intensity of thousand amperes,average annual number of houses damaged by lightning,height of the terrain,air humidity,and proportion of houses without lightning disaster defense system.Secondly,the coefficient of each factor in the comprehensive score model was calculated using SPSS principal component analysis method.They are 0.004872131,0.051450329,0.066425375,0.205600074,0.249310058,-1.002372123 and 2.385839541,respectively.Finally,the product sum of the coefficients of each factor and the standardized data of the corresponding major impact factors of lightning was calculated to obtain the total lightning risk index of each risk area,and each area was divided according to the range of the number of risk indicators.(3)Accurately predict the impact of lightning,and use BP-CNN neural network for supervised learning.The back propagation algorithm is used to train the network,and the predicted value is compared with the actual value,and then the weight is adjusted to make the predicted value more close to the actual value.Mean square error(MSE),root mean square error(RMSE),mean absolute error(MAE),mean percentage error(MAPE)and correlation coefficient(R)were used to evaluate the accuracy of lightning prediction.The data of 200 days with the most frequent lightning occurrence were selected from the data set of Ningxia from 2019 to 2021 as the training data and prediction data of this model.In the process of model training,the results of MSE=0.00028066,RMSE=0.016753,MAE=0.014252,MAPE=0.093865%were obtained,and these results were close to 0,indicating that the prediction accuracy of the model was high.In addition,the correlation coefficient R=1 between the predicted value and the real value also shows the reliability and accuracy of the model.The results of this study have practical guiding significance and practical application value,and can provide practical help for regions like Ningxia.