Rainfall Prediction Study In Zhengzhou Area Based On Markov Chain

Rainfall is an important parameter of agricultural production,water conservancy construction and hydrological analysis,It is also one of the main causes of floods.Therefore,accurate prediction of rainfall can provide decision-making basis for agriculture,water conservancy and other relevant departments,in order to reduce unnecessary losses,has important guiding significance.However,the influencing factors of rainfall have the characteristics of fuzziness and uncertainty,which leads to the complexity and randomness of rainfall in time series and makes it difficult to accurately predict rainfall.At present,there are many rainfall prediction methods,such as ARIMA prediction method,classical correlation analysis method and multiple linear regression method and so on.Among them,markov chain is the most widely used,mainly because markov chain is an important random process,which has no after-effect,so it is suitable for rainfall prediction.However,when markov chain is used to predict rainfall,the prediction effect of extreme value is not ideal.Aiming at this problem,this paper constructs a rainfall prediction model based on markov chain,fuzzy set and gray prediction,and realizes the prediction of annual rainfall and flood season rainfall in zhengzhou.The main research work includes the following aspects:(1)A rainfall prediction model based on markov chain and fuzzy set is constructed.Firstly,the mean-variance grouping method was used to group the rainfall data series and determine the corresponding state of the rainfall data.Then,the markov test is carried out on the rainfall state sequence,and the rainfall prediction interval is calculated by the transfer matrix and weight.Finally,the level eigenvalues in fuzzy set theory are used to obtain the predicted values.(2)A rainfall prediction model based on markov chain,grey prediction and fuzzy set is constructed.First,the sliding average method was used to process the rainfall data,eliminate the outliers,and accumulate the results to obtain a new data series.Then,the new data sequence is used to set up the differential equation,and the parameter value is obtained through the operation,so the prediction function can be obtained and the predicted value can be obtained.Finally,the relative error sequence is obtained by comparing the predicted value with the actual value,and the model based on markov chain and fuzzy set is adopted to deal with it,and the final predicted value is obtained.Based on the data of rainfall in zhengzhou from 1951 to 2007,the rainfall forecast model was used to predict the rainfall in 2008,2009 and 2010.The results show that the model is effective and feasible in predicting annual rainfall and rainfall during flood season in zhengzhou.