Pai Tzu-Yi, Wang S C, Chiang C F, Su H C, Yu L F, Sung P J, Lin C Y, Hu H C
Department of Environmental Engineering and Management, Chaoyang University of Technology, Wufeng, Taichung 41349, Taiwan.
Bioprocess Biosyst Eng. 2009 Oct;32(6):781-90. doi: 10.1007/s00449-009-0304-2. Epub 2009 Mar 1.
Three types of adaptive network-based fuzzy inference system (ANFIS) in which the online monitoring parameters served as the input variable were employed to predict suspended solids (SS(eff)), chemical oxygen demand (COD(eff)), and pH(eff) in the effluent from a biological wastewater treatment plant in industrial park. Artificial neural network (ANN) was also used for comparison. The results indicated that ANFIS statistically outperforms ANN in terms of effluent prediction. When predicting, the minimum mean absolute percentage errors of 2.90, 2.54 and 0.36% for SS(eff), COD(eff) and pH(eff) could be achieved using ANFIS. The maximum values of correlation coefficient for SS(eff), COD(eff), and pH(eff) were 0.97, 0.95, and 0.98, respectively. The minimum mean square errors of 0.21, 1.41 and 0.00, and the minimum root mean square errors of 0.46, 1.19 and 0.04 for SS(eff), COD(eff), and pH(eff) could also be achieved.
采用三种基于自适应网络的模糊推理系统（ANFIS），将在线监测参数作为输入变量，来预测工业园区生物废水处理厂出水的悬浮固体（SS(eff)）、化学需氧量（COD(eff)）和pH值（pH(eff)）。同时使用人工神经网络（ANN）进行对比。结果表明，在出水预测方面，ANFIS在统计学上优于ANN。预测时，使用ANFIS对SS(eff)、COD(eff)和pH(eff)的最小平均绝对百分比误差分别为2.90%、2.54%和0.36%。SS(eff)、COD(eff)和pH(eff)的相关系数最大值分别为0.97、0.95和0.98。SS(eff)、COD(eff)和pH(eff)的最小均方误差分别为0.21、1.41和0.00，最小均方根误差分别为0.46、1.19和0.04。
