基于蚁狮算法优化深度极限学习机的水质评价研究

Research on water quality evaluation based on ALO-DELM

为提高水质评价的准确性,提出一种基于蚁狮算法优化深度极限学习机的水质评价模型.由于深度极限学习机采用随机方式对权重和阈值进行初始化,使权重和阈值存在随机性和不确定性,针对这一问题,选用蚁狮算法对深度极限学习机的权重和阈值进行初始化,然后用训练样本集数据对深度极限学习机进行训练,用训练好的模型对测试样本进行水质评价预测,并与深度极限学习机及其他智能算法优化深度极限学习机模型进行对比.对比实验结果表明,蚁狮算法优化的深度极限学习机模型的水质评价预测结果明显优于深度极限学习机,也优于其他智能算法优化的深度极限学习机模型,验证了该方法在水质评价预测中的有效性.

To enhance the accuracy of water quality evaluation,a novel evaluation model based on ant lion algorithm optimized deep extreme learning machine is proposed.Given the random initialization of weights and thresholds by deep extreme learning machine,the weights and thresholds exhibit randomness and uncertainty.To mitigate this issue,ant lion algorithm is employed to initialize the weights and thresholds of deep extreme learning machine,and then the extreme deep learning machine is trained a dataset comprising training samples.The trained model is subsequently utilized to predict water quality evaluation of test samples,and compared with deep extreme learning machine and other intelligent algorithm optimized deep extreme learning machine models.The comparative experimental results show that the water quality evaluation prediction results of the deep extreme learning machine model optimized by ant lion algorithm are significantly better than those of deep extreme learning machine,It is also superior to other intelligent algorithm optimized deep extreme learning machine models,verifying the effectiveness of this method in water quality evaluation and prediction.