Occupational Health Prediction Based On Grey GM(1,1)-Markov Chain Model

Due to the rapid development of industrial economy and workers’ less awareness of occupational disease prevention,occupational health problems become increasingly prominent in China.At present,many studies have pointed out that the incidence of occupational diseases is increasing,but there is no quantitative study on the trend of occupational diseases and the prediction of the concentration of occupational hazards in the enterprise in China.This paper mainly uses grey prediction theory,Markov chain theory and metabolic theory.Firstly,the grey GM(1,1)prediction model with different dimensions is constructed,and the fitting effect of the model is compared with the average of the relative error,the posterior difference ratio and the small error probability.Based on the selected grey GM(1,1)model,the metabolic theory was introduced to construct a gradually progressive metabolic grey GM(1,1)model with equal dimensions.and the fitting effect was compared.As the grey GM(1,1)model can only reflect the long-term growth or decline of the development trend,it can not reflect the drawbacks of its volatility characteristics,so the introduction of Markov chain model theory knowledge.The grey GM(1,1)-Markov chain model is used to predict the recent time of the study object,and then the metabolic theory is used to predict the incidence of occupational diseases and the trend of carbon monoxide concentration in the future.Based on the eight-dimensional grey GM(1,1)prediction model and the metabolic principle,the eight-dimensional metabolic grey GM(1,1)prediction model with the same dimension was established.The prediction of the incidence of occupational diseases in China in 2013-2018 is:{36411,42404,49461,56213,63635,74022},among which the predicted results in 2013 and 2014 are larger than those of actual occupational diseases,which were 37.96%and 41.48%.The relative error of the occupational disease incidence was 32273 and 41280 in 2013,and the relative error of the actual incidence of occupational diseases was about 22.28%and 56.41%respectively.The predicted results shows that the error is bigger,2014-2018 years of occupational disease prediction error will be greater.The ten-dimensional grey GM(1,1)-Markov chain model was established by the data of occupational diseases from 2005 to 2014.The relative error of occupational disease prediction in 2013 and 2014 decreased to 3.85%and 6.12%respectively,and the prediction accuracy was very good.And the forecast level is 1.the number of occupational disease forecast results in 2015-2018 is:{31196,36284.37724.39147}.At the same time,the grey GM(1,1)-Markov chain model is applied to the prediction of carbon monoxide concentration,and the prediction accuracy is basically qualified(Grade two).