净水厂最佳投药量的神经网络控制系统

混凝剂的投加是净水厂水处理工艺的重要环节，本文通过对目前水厂实际运行过程中混凝剂投加量的影响因素的分析，利用神经网络预测理论，建立了净水厂混凝剂投加量的预测模型，并利用该模型对某水厂的实际运行情况进行了预测，对网络预测模型的性能进行了验证。结果表明，该模型具有很强的自学习性，自适应性和容错性，依靠网络的在线自学习，可使预测结果的准确度明显提高。

A neural network control system for the optimum coagulant dose rate in a water purification plant

Coagulation is an important process in water purification plants.By analyzing the coagulant dose rate and its influencing factors,a prediction model is established by means of the neural networks predicting theory so as to predict the dose rate of a purification plant.The model performance has been tested.The results indicate that this model has high self learning performance,adaptability and fault tolerance.The accuracy of prediction results can be improved obviously by the on line self learning of the networks.