基于IGWO-KELM的消毒剂有效氯浓度软测量研究

Research on Soft Measurement of Effective Chlorine Concentration of Disinfectants Based on IGWO-KELM

摘要:

针对消毒过程中有效氯浓度依赖离线取样分析、实时性较差的问题，对基于多传感信息的有效氯浓度软测量方法进行了研究。以pH值和氧化还原电位（ORP）为输入变量，以DPD分光光度法测定的有效氯浓度为输出，构建了有效氯浓度非线性映射模型。采用改进灰狼优化算法对核极限学习机的惩罚因子和核宽参数进行全局寻优，并引入Tent混沌映射、Levy飞行机制和黄金正弦策略，以提高种群多样性、全局搜索能力和局部开发能力。实验结果表明，IGWO-KELM模型在测试集上的MAE、RMSE和R2分别为0.6497、0.8634和0.9654，预测精度和泛化性能优于KELM和GWO-KELM模型。所提方法实现了有效氯浓度的准确估计，可为消毒过程浓度监测与后续智能调控提供技术支撑。

Abstract:

Regarding the problem that the effective chlorine concentration during disinfection requires offline sampling analysis and has poor real-time performance, a study was conducted on the soft measurement method of effective chlorine concentration based on multi-sensor information. Taking pH value and oxidation-reduction potential (ORP) as input variables and the measured effective chlorine concentration by DPD spectrophotometry as the output, a nonlinear mapping model of effective chlorine concentration was constructed. The improved grey wolf optimization algorithm was used to globally optimize the penalty factor and kernel width parameters of the kernel extreme learning machine, and Tent chaotic mapping, Levy flight mechanism and golden sine strategy were introduced to improve the population diversity, global search ability and local exploration ability. Experimental results show that the MAE, RMSE and R2 of the IGWO-KELM model on the test set are 0.6497, 0.8634 and 0.9654 respectively, and the prediction accuracy and generalization performance are superior to those of the KELM and GWO-KELM models. The proposed method achieves accurate estimation of the effective chlorine concentration and can provide technical support for concentration monitoring during the disinfection process and subsequent intelligent regulation.