Research On Quota Compilation Of Water Conservancy Enterprises Based On BP Neural Network And Its Optimization Model

In the context of China's economic integration into the world economy and the domestic bill of quantities pricing system,the rapid development of the water conservancy industry has put forward the real-time and efficient requirements for the establish enterprise quota.in practice,however,most of the domestic water conservancy construction enterprises lack of fixed ideas and effective establishment method,It is mainly related to the historical development of the national planned economic system and the lack of quota system and the lack of quota system,It lead to the existing water conservancy industry concerning the meaning of enterprise construction quota consciousness weak,do not think or nor ability to establish quota belong to own characteristics of water conservancy,It is still used in practical engineering cost budget quota standards of national or local.In view of this,this paper takes the construction of water conservancy industry norm as the research object to carry out research work of quota establishment,points out that traditional and current quota establishment method is inefficient tedious,time-consuming and difficult to make full use of historical information and the disadvantages of renewable capacity.As the same time,analyzes information technology the development such as BIM?cloud computing and big data,Paper is designed to solve the problem of quota establishment from data information and intelligent prediction of large data perspective,through the neural network theory and intelligent optimization method,It is put forward based on BP neural network prediction model of quota establishment and its optimization algorithm,provide suppor that the water conservancy industry for the future engineering cost method to the digital information t,paper hope to be able to make the water conservancy industry and other industry enterprise quota compilation can satisfy social development needs of the real-time and efficient,updated dynamically,predictable.This article main research content,points out the enterprise quota original outlier data processing,and puts forward the improved Mrs Grubbs criteria based on normal distribution for outlier test way of improvement,this step is important for late quota establishment modeling.Quota establishment phase establish pure BP model and the optimization of GA-BP model and PSO-BP model,with the help of collecting post-processing of the data model for the training and testing,and three models of effect is analyzed and compared.The research results show that the most relative error of BP network model is developed for quota within 10%,it is feasible to solve the problem of actual quota establishment,but the BP neural network fall into the local optimum easily and the accuracy and speed to be optimized.In view of this,paper further put forward BP optimization model based on genetic algorithm and particle swarm algorithm,the content is to optimize the initial weights and threshold of BP neural network,the optimization model combined with BP neural network of nonlinear approximation,local optimization ability and global search of genetic algorithm.The optimization results show that the optimization model is effectively improved in the optimization speed and accuracy,in which the accuracy of PSO-BP model is slightly better than that of GA-BP,and the PSO-BP coding is simpler and faster.It is concluded that the BP model and the BP model optimized by intelligent algorithm can effectively solve the problem of quota system in the water conservancy industry,The model meet the requirements of real-time efficiency and the development of the times.