|
| 基于PSO优化LSSVM的出水BOD5预测建模 |
| Predictive Modeling of Effluent BOD5 Based on PSO Optimized LSSVM |
| |
| DOI：10.16018/j.cnki.cn32-1650/n.202104003 |
| 中文关键词: 粒子群 LSSVM 出水BOD5 建模 |
| 英文关键词: particle swarm optimization LSSVM effluent BOD5 modeling |
| 基金项目: |
|
| 摘要点击次数: 6316 |
| 全文下载次数: 5226 |
| 中文摘要: |
| 针对城市污水处理中有机物污染度指标BOD5缺少运行状态信息难以做到实时检测，提出了一种基于粒子群算法（PSO）优化最小二乘法支持向量机（LSSVM）的出水BOD5预测控制策略。在保证出水水质参数稳定达标条件下，提取城市污水处理过程中输入输出参数数据，通过LSSVM对被控对象出水BOD5进行建模，同时利用PSO对LSSVM模型进行参数寻优，获得最佳正则化参数γ和核函数参数σ。仿真结果说明，该模型提高了对出水BOD5值的预测精度并具有良好的泛化能力，达到了实时性的效果。 |
| 英文摘要: |
| In view of the fact that BOD5 is lack of running state information in urban sewage treatment, and it is difficult to achieve real-time detection, a predictive control strategy of effluent BOD5 based on particle swarm optimization(PSO) and least support vector machine(LSSVM) is proposed. Under the condition that the effluent quality parameters are stable and up to the standard,the input and output parameter data in the process of urban sewage treatment are extracted, the effluent BOD5 of the controlled object is modeled by LSSVM, and the parameters of LSSVM model are optimized by PSO to obtain the optimal regularization parameters and kernel parameters. The simulation results show that the model improves the prediction accuracy of effluent BOD5 value,and has good generalization ability, achieving the real-time effect. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |