Pareto蚁群算法与遥感技术耦合的水资源优化配置

Combination of Pareto ant colony algorithm with remote sensing for optimal allocation of water resources

中文关键词  优化配置  水资源  多目标  遥感  蚁群算法

英文关键词  optimal allocation  water resources  multiple objective  remote sense  Pareto ant colony algorithm

中文摘要

      为了尝试用Pareto蚁群算法(PACA)和遥感技术(RS)来求解复杂的水资源优化配置问题, 建立了以经济、社会和生态环境综合效益最大为目标, 以供水、需水、水质等为约束条件的基于像元的水资源优化配置模型. 通过局部信息素强度限制、全局信息素动态更新、Pareto解集过滤器构建等策略, 使蚂蚁向信息素浓度大的优化边界移动,以提高PACA的全局搜索能力和收敛速度. 以中原地区某县为仿真对象, 借助RS获取其土地利用类型, 利用PACA在栅格地图上求解水资源优化配置模型, 并得到水资源最优配置方案. 最后PACA与遗传算法(GA)和BP神经网络算法(BP-ANN)进行了比较. 结果表明, PACA能有效地求解大范围、多目标水资源优化配置模型, 并提高了算法的全局搜索能力、收敛速度和计算结果的精度.

英文摘要

      To solve the optimal allocation problem of water resources with Pareto ant colony algorithm (PACA) and remote sensing (RS), we develop an optimization model in pixel scales. This model produces the largest social, economic, and environmental benefits under constraints on water supply, water demand and water quality. By limiting the local pheromone scope, dynamically updating the global pheromone and filtering the Pareto solution set, we improve the PACA to make ants move towards the optimal border with higher pheromone density, and enhance the global search capability and raise the convergence rate. To validate the feasibility and effectiveness of the PACA, a county in central China is selected as the simulation object, from which the data of the land-use pattern is obtained by using the RS technology. By solving the multi-objective model, we obtain the optimal allocation scheme for water resources with the aid of PACA on a raster map. Performance and convergence of the PACA are compared with those of the genetic algorithm (GA) and BP neural network algorithm (BP-ANN); results show that PACA can effectively solve the large-scale, multi-objective optimization model of water resources with stronger global search capability and higher convergence rate and precision.