论文部分内容阅读
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.
It is a new research hotspot to use rule-based particle swarm optimization (PSO) for rule mining.A model of flood risk assessment based on Particle Swarm Optimization (PSO-Miner) is proposed.Based on GIS, The results show that the PSO-Miner algorithm is a kind of intelligent method without parameter evaluation, which has strong global convergence ability and robustness. The mining If-Then rule can be simpler and Accurate description of the complex relationship between each index and risk level; the overall accuracy is higher than the BP neural network model, and objectively reflect the actual situation of flood risk in the Beijiang River Basin; combined with GIS technology to facilitate the analysis of the spatial pattern of flood risk And the inherent law, has good applicability.