机构地区: 嘉应学院电子信息工程学院
出 处: 《水科学与工程技术》 2011年第1期14-16,共3页
摘 要: 作为水质净化重要环节的混凝投药是一个非线性系统,目前还很难对其建立准确的数学模型。该文提出了基于改进BP神经网络的解决方法,根据水源参数的具体特征提取特征值并建立相应的神经网络,通过训练,网络具有较强的适应和学习功能,通过仿真和实验达到了很好的混凝投药控制效果,使混凝投药系统的控制迈向智能化。 The Coagulant Dosing is a nonlinear system. It is very important in the process of water purification. It is difficult to establish a precise mathematical model. The paper puts forward a solution based on improved BP algorithm. According to the water source parameters, determines characteristic values and establishs neural network which is very excellent in learn speed and adaptability. The simulation and experiment of the coagulant dosing system show that the demand of control precision is satisfied. The paper put forward a improved BP algorithm which makes the coagulant dosing system control more intelligentize .
领 域: [环境科学与工程]