基于粒子群规则挖掘算法的洪灾风险评价模型

运用粒子群优化算法(particle swarm optimization, PSO)进行规则挖掘是一个新的研究热点. 提出了一种基于粒子群规则挖掘算法(PSO-Miner)的洪灾风险评价模型. 基于GIS技术利用该模型对北江流域洪灾风险等级进行了评判, 结果表明: PSO-Miner算法是一种无参数评判的智能方法, 具有较强的全局收敛能力和鲁棒性; 所挖掘的If-Then评判规则能更简单和准确地描述各评价指标与风险等级之间的复杂关系; 总体精度比BP神经网络模型的更高, 而且能客观地反映北江流域洪灾风险实际情况; 与GIS技术结合, 便于分析洪灾风险的空间 格局及内在规律, 具有较好的适用性.

Flood risk assessment model based on particle swarm optimization rule mining algorithm

Particle swarm optimization (PSO) as a novel intelligent optimization algorithm has been used successfully in many fields, but its application to flood hazard risk assessment is a new research topic. This paper introduces the theory and flow of application of particle swarm optimization rule mining (PSO-Miner) algorithm to flood damage risk assessment. This paper selected Beijiang River Basin, China, as study area for flood damage risk assessment based on PSO-Miner algorithm and BPANN method. The results of a case study indicate that the advantages of PSO-Miner algorithm can be summarized as follows: It does not assume an implicit assumption for processing dataset and has strong robustness; it can mine very simple assessment rules; it can have a better performance than BPANN model. So the PSO-Miner algorithm provides a new approach for flood risk assessment.