会议专题

Application of the RAGA-BP ANN based on MCMC in Water Quality Evaluation

-The quality evaluation is a main matter in evaluation and protection of water resources; it can provide scientific advice for protection and cure of water resources deterioration and formulation of water resources management plan. So the water quality evaluation can also give guidance to sustainable water utilization. Back Propagation artificial neural net based on Real coded Accelerating Genetic Algorithm and Markov Chain Monte Carlo (RAGA-BP based on MCMC) was used to evaluation water quality. RAGA was used to optimize topology, initial weights and bias of BP; MCMC was introduced to produce enough training samples of BP and to consider fuzziness between grades of water quality. Results of study cases show RAGA-BP based on MCMC improve convergence velocity more 20% than traditional BP.

BP RAGA MCMC water resources Evaluation

Xing Zhenxiang Fu Qiang

College of Water Conservancy & Architecture Northeast Agricultural University Harbin, China, 150030

国际会议

2007 Conference on Systems Science, Management Science and System Dynamics(第二届系统科学、管理科学与系统动力学国际会议)

上海

英文

2459-2464

2007-10-19(万方平台首次上网日期,不代表论文的发表时间)