Accident Prediction Based On Improved Grey System And GABP Network Combination Model

In recent years,the situation of safety production management has been improved,but in general,the total amount of casualties in production safety is still high,which has brought serious negative impact on the economy and society.Through the accident prediction,we can grasp the future trend of the accident,so as to provide scientific basis for the establishment of scientific and reasonable technical measures to prevent accidents.In this paper,two single models of improved Grey prediction model and improved GABP prediction model are established respectively,and the combination model is established by using the optimal combination forecasting method to improve the model accuracy.The details are as follows:(1)Establish an improved Grey forecasting model.The weight vector calculation method of weighted geometric mean weakening operator is constructed,and the equal dimension metabolic model is introduced to establish the grey prediction model.Then the model precision is obtained by the corresponding precision calibration standard.Finally,the accuracy indexes of different models are calculated by examples,and the results show that the improved model is superior to the other two models in the accuracy of the model.(2)Establishment of improved GABP prediction model.Based on the data of2004-2015 years' casualty accidents in a city,the correlation coefficient between the accident and GDP,the average wage of workers and the proportion of the third industry and the number of accidents are analyzed,and the prediction index of BP neural network is determined.The genetic algorithm is used to select the initial weights and thresholds of BP neural network scientifically and reasonably,and the improved selection method of genetic algorithm is also improved.In this paper,the genetic algorithm is optimized by the combination of block replication and roulette wheel selection method and adaptive crossover and mutation operator.An example is given to verify that the improved GABP prediction model has higher predictionaccuracy than the BP prediction model and the GABP prediction model.(3)Using the single model constructed in the first two chapters,the combined model is calculated.Firstly,the weight coefficients of each single model are obtained by calculating,so that the mathematical formula of the combined model can be obtained.Then,an example is used to test the prediction effect of the combined model on the number of accidents.Finally,the comparison shows that the prediction accuracy of the combination model established in this paper is better than the other two models.