Optimization Of Crop Irrigation Schedule Based On Years Of Rainfall Data

In our country, water resources are not only unevenly distributed but also very poor.As an agricultural country, how to improve the utilization rate of water in agriculture, toachieve the purpose of water-saving and high yield, has strategic significance foralleviating the contradiction between supply and demand of water resources in China. Inthe irrigation area water use planning, how to formulate a reasonable irrigation system isthe key problem.When discussing problems of water-saving irrigation, we need conside the influenceof rainfall on irrigation system, and make full use of rainfall to reduce irrigation water,improve the utilization of water resources. While formulate irrigation system, we candepend on the field moisture dynamic variation, and the field and the dynamic change ofwater content, which associated with rainfall and irrigation water control measures such asmeteorological conditions. However the weather conditions are random, in different years,whether it is in the period of crop growth in total rainfall and rainfall distributions of timesometimes vary greatly. Therefore it is difficult to implement in practice to optimizatde thecrop irrigation schedule based on rainfall randomness.Precipitation of crop growth period and its temporal distribution of direct effects ofirrigation schedule, and the long-term prediction of rainfall is one has so far failed to solvethe problem very well, so how to consider the randomness of the rainfall is the key to solvethis problem. This article discusses the three different methods to concedering tne rainfallrandomness: typical year method, the method based on years of rainfall data, and BPneural network based on the known data to predict rainfall method.Under the condition of insufficient irrigation schedule, based on the field waterbalance model and crop yield calculation model, and considering three different methodsof dealing with the rainfall randomness, a multi-objective optimization model was putforward to optimize irrigation date and corresponding quantity of irrigation simultaneously.In the model, irrigation date and the quantity of irrigation water were regarded asoptimization variables. The model and algorithm were verified by irrigation scheduleoptimization of corn.This paper discusses the principle of non-dominated sorting genetic algorithm and itsapplication in practice. The model and algorithm were verified by irrigation scheduleoptimization of corn. We can get different irrigation schedule optimization results based on the three different methods of dealing with the rainfall randomness. Analysis of the results,summary the advantages and disadvantages of three different methods, decision makerscan be available according to the time of water irrigation system selection and expectationsof production.