随着公用事业向过程自动化和无人值守运行的方向发展，饮用水处理行业必须积极探索新技术，以优化过程控制策略。人工神经网络（ANN）用于协助控制水处理过程时，有可能显著降低操作和维护成本，改善客户服务和水质。本文介绍了人工神经网络（ANN）模型的结果，该模型用于预测颗粒物在关闭状态下的去除。E.Weymouth过滤厂（Weymouth）和氧化示范项目（ODP）工厂，由加利福尼亚州拉凡尔纳市南加州大都会水区（MWD）所有和运营。 从美国水务协会研究基金会（AWWARF）资助的这项研究得出的模型可以被纳入MWD设施的间接过程控制方案。间接过程控制应用程序，包括虚拟实验室和操作员培训模块，可以帮助MWD设施的操作员优化颗粒去除。包括13个参考文献、表格、图表。

As utilities move towards process automation and unattended operations, the drinkingwater treatment industry must actively search out new technologies that allow for theoptimization of process control strategies. When used to assist in the control of water treatmentprocesses, artificial neural networks (ANNs) have the potential to significantly reduce operationsand maintenance costs, to improve customer service and water quality. This paperpresents the results of ANN models developed to predict the removal of particulate matter at theF.E. Weymouth Filtration Plant (Weymouth) and the Oxidation Demonstration Project (ODP)Plant, owned and operated by the Metropolitan Water District of Southern California (MWD) of La Verne, California.The models derived from this research, which is being funded by the American Water WorksAssociation Research Foundation (AWWARF), can be incorporated into indirect process controlschemes at the MWD facilities. The indirect process control applications, including virtuallaboratories and operator training modules, can assist operators at MWD facilities in optimizingparticle removal. Includes 13 references, tables, figures.