摘要
为了提高地下水质量的预测精度,提出了一种由海豚群算法与BP神经网络结合的水体质量评价方法(DPA-BP)。算法的主要思想是将求解BP神经网络最优权值和阈值的过程转化为海豚群捕食寻求最优位置的过程,通过逐代寻优,确定BP网络评价水环境质量的最优权值、阈值。并对文献[9]中8个采样点地下水的数据与其他算法进行对比评价。结果表明,DPA-BP算法对地下水质评价的精确性高、稳定性好,具有较好的鲁棒性和实际工程实用价值。
In order to improve the prediction accuracy of groundwater quality,a new method combining the dolphins algorithm and BP neural network was put forward.In the method,procedure for solving the optimal BP neural network weights and thresholds is transforming into the process of dolphin species predator finding the optimal location,and it effectively combine the good generalization mapping capability of BP neural network neural and the global optimization algorithm and local search capability of DPA.The dairy eight sampling points of Panshi groundwater test data are used as the test samples to evaluate groundwater quality.The results show that the quality assessment values are accurate by using the proposed algorithm,and the algorithm has strong robustness and practical engineering value.
出处
《节水灌溉》
北大核心
2015年第9期66-69,73,共5页
Water Saving Irrigation
基金
国家自然科学基金青年项目(11301454)
