The Application And Research Of Water Optimize Dispatching System Of The Ganjiang River During Non-flood Season

With the rapid development of the economy, demand on water resources has been increasing. But water pollution and environmental changes have lessened available water resources. Gan River Basin is located in the wet areas in China, but water is unevenly distributed in terms of time and space. There is also obvious contradiction between water supply and demand in non-flood season. Water dispatching is an important method to optimal allocation of water resources, and the scientific management of the mutil-reservoirs group is of great significance to solve water shortages of Gan River Basin in the non-flood season.This paper describes the non-flood season Gan River Basin water optimal dispatching model, intelligent algorithm reseach and application platform development. First, combined with the specific circumstances of the Gan River Basin, I proposed4matrixes and abstracted the basin into a pure matrix expression. Then I built water dispatching optimal model of Gan River Basin in the non-flood, and I applied penalty function to deal with complex constraints. Eventually I built a generic, standardized water dispatching optimal model based on the relations-matrix and the flag-matrix. Second, in order to solve the water dispatching optimal model, I presented a double-layer multi-swarm particle swann optimization algorithm with differential evolution (DE-TMPSO). Experiments confirmed that this improved algorithm has some advantages compared with the basic particle swarm algorithm, such as fast convergence, high accuracy, avoiding local optimal value and so on. Then I applied this new algorithm to the water dispatching optimal model, taking the discharged volume of each reservoir in each dispatching period as decision variables, and got better scheduling results. Finally, I developed a B/S architecture decision support platform of the non-flood water dispatching using C#language and SQL2005database. The platform realizes the water resource information management and the water data sharing, it provides reference and guidance for water scheduling operation. The establishment of matrix model of Non-flood Gan River Basin water optimal dispatching, and the proposed of double-layer multi-swarm particle swarm algorithm has enriched the research of water dispatching problem. Decision support platform realizes water resource information management and intelligent reservoir group scheduling, it makes the water dispatching of Ganjiang River Basin more scientific and reasonable.