Application Research Of Moth-Flame Optimization Algorithm In Safety Monitoring Model Of Water Diversion Project

Water diversion project is an important project to mobilize water resources to solve regional water shortage,and plays a strategic,basic and overall role in the process of water resources allocation.Due to the impact of the space and environment,the water diversion project has many potential safety hazards within the channel.When the channel water body leaks through the permeable rock belt,the channel slope will be damaged by the infiltration,causing the collapse and the collapse,which will cause the slope instability to affect the water supply efficiency of the water diversion project.Therefore,it is of great practical significance to carry out safety monitoring of the water diversion project and timely and accurately grasp the working state of the diversion works.To monitor and predict the seepage pressure of the diversion project,it is necessary to consider the factors affecting the seepage pressure according to the different environments of the diversion project and its own characteristics,and establish a seepage prediction model that can reflect the actual seepage pressure to ensure the safety and stability of the diversion project..The moth-trimming optimization(MFO)algorithm is a new intelligent optimization algorithm with great development potential,which has the characteristics of simple implementation and high convergence precision.In this paper,the MFO algorithm is combined with the cross and cross algorithm and chaotic operator to improve the algorithm to form the cross-crawling moth-carrying(CCMFO)algorithm.According to the basic principle of seepage flow in the diversion project,the four influencing factors of water level,temperature,aging and rainfall are determined.Based on the multiple regression model and BP neural network model of osmotic pressure monitoring,a multivariate regression and BP neural network osmotic pressure prediction model was constructed.The CCMFO algorithm was used to optimize the regression model coefficients and BP neural network model weight threshold updating methods.The CCMFO algorithm optimized osmotic pressure prediction model is compared with the original model.The results show that compared with the traditional regression model,the CCMFO-regressive osmotic pressure monitoring model improves the seepage pressure fitting and prediction accuracy,and achieves better osmotic pressure prediction.Compared with the BP neural network model,the CCMFO-BP osmotic monitoring model achieves better results in terms of iterative velocity and convergence accuracy,and achieves the goal of improving the model fitting accuracy and enhancing the model prediction ability.The CCMFO algorithm improves the fitting effect and prediction accuracy of two kinds of osmotic monitoring models,and provides a feasible method for monitoring data analysis of similar projects.