论文部分内容阅读
针对水处理过程中混凝剂的准确投加,以及投药过程中的时滞、网络延迟等问题,采用基于网络学习控制的智能控制算法来改进投药控制系统。用远程专家系统和自学习BP神经网络复合算法的优点,即专家系统的前馈补偿能力解决流量、浊度突变、延迟等干扰因素;神经网络的非线性映射能力解决水处理非线性影响;滚动学习和反馈学习来解决时变、时滞等问题。该算法较好地解决了网络延迟造成的系统性能下降问题,加快了神经网络的训练速度。将该算法应用于水厂自动化系统,可以实现最佳投药量控制,水质符合标准,并取得良好的经济效益。
In view of the accurate dosing of coagulant in the process of water treatment, as well as the delay in the process of drug delivery and network delay, the intelligent control algorithm based on network learning control is used to improve the dosing control system. Using the advantage of the remote expert system and the self-learning BP neural network compound algorithm, that is, the feedforward compensation ability of expert system can solve the interference factors such as flow rate, turbidity mutation and delay; the nonlinear mapping ability of neural network can solve the nonlinear influence of water treatment; Learning and feedback learning to solve time-varying, time-lag and other issues. The algorithm solves the problem of system performance degradation caused by network delay and accelerates the training speed of neural network. Applying this algorithm to the waterworks automation system can realize the best dosage control, the water quality conforms to the standard, and obtains the good economic benefits.