A Decision Support System For Water Resource Programming

Water programming consists of that how to distribute limited water resource among different users and satisfy the water requirement of the economy and society. It can help to decides how to produce the largest profit with the limited water resource.The main contribution of this paper is to develop a system which will provide information management and decision support to the water programming in Dongying according to the actual water resource condition. The decision support system (DSS) in this paper consists of a database system and a model system. But it can provide information management, prediction and model solution. A database is built with ACCESS 2000 in this paper and a database management system (DBMS) is developed with the ADO controls, Tee Chart package and the OLE of the Delphi. The DBMS can manage the information well and provide friendly interface. In the model library, class of prediction models, class of water model are developed, and component package of programming methods is completed. A parent class of the time series prediction is developed with the object oriented programming (OOP) method. Then two children classes, the least squares support vector machine (LSSVM) and the grey model (GM) class, are produced to predict different series. A water model class is build to store and transfer data. At the same time, programming methods are encapsulated as components, such as simplex component, hybrid GA component, GA component. In the end, the system makes a water programming decision for 2007.Water programming models are studied in this paper and programming methods are researched a lot. That allows this paper to propose a hybrid genetic algorithm (GA) which can be used to solve the nonlinear water programming problem. This hybrid algorithm refers to the position displacement strategy of the particle swarm optimization (PSO) and makes use of the gauss PSO to improve the mutation operation of the GA. The hybrid GA is tested with three benchmark functions and applied to solve a portfolio selection problem. The result of the test and application shows that this hybrid GA is efficient to solve the nonlinear programming problems. A component which encapsulates the hybrid GA is programmed in the model library and it is used to solve the nonlinear water programming.