基于ANFIS的水质净化预测控制  

Forecast Control of Water Purification based on ANFIS

作　　者：张春美[1] 石川[2] 曹俊琴[1] 

机构地区：[1]太原科技大学,太原030024 [2]山西广播电视网络集团有限公司,太原030001

出　　处：《电脑开发与应用》2008年第4期65-66,72,共3页Computer Development & Applications

摘　　要：水资源是基础性自然资源,也是战略性经济资源,正确对自来水的净化进行预测控制对于社会经济和环境的协调发展具有重要的意义。传统的预测控制方法不能及时适应水质的不确定动态变化,净化效果不佳。自适应神经模糊推理系统(ANFIS)是模糊逻辑和神经网络最完全的融合,ANFIS的自适应性质使得它可以替代控制系统的任意神经元网络并执行同样的功能。针对水净化过程的不确定性,建立了以ANFIS为基础的自来水质净化的预测控制,仿真结果显示:该自适应神经模糊控制器具有较高的控制精度,控制效果较好。Water is not only the basic natural resource but also the strategic economic resource. The traditional method of forecast control can not meet uncertain and dynamic changes of water quality and the result of purification is not satisfactory. ANFIS （Adaptive Neuro-Fuzzy Inference System） is the most complete system which is the combination of fuzzy logic and neuro network, the adaptability of ANFIS make it replace any neuro network and conduct same functions. Considering the uncertainty in water purifying process, an ANFIS-based system model is established to perform forecast control of purifying water. The simulation results demonstrate that the adaptive neuro-fuzzy controller has high precision and good control performance.

关 键 词：自适应神经模糊系统 SUGENO模糊模型 预测控制 混合学习算法 

分 类 号：TP391.76[自动化与计算机技术—计算机应用技术]