Soft Measurement Technique of Sewage Treatment Parameters Based on Wavelet Neural Networks

Abstract

There are certain nonlinearity, time variability, randomness and uncertainty in the process of using sequencing batch sludge method in sewage treatment. Therefore, propose a soft measurement technique for sewage treatment parameters basing on the model of Kernel Principal Component Analysis and Wavelet Neural Network. Use Kernel Principal Component Analysis as concise as possible in the case of the input variables can ensure that a smaller amount of loss and combine WNN soft-measurement model and on-line measuring instruments together, do real-time detection for redox potential, dissolved oxygen, PH, COD and so on parameter control information .PLC controller outputs control signals to control the entire system equipment operation. The simulation results show that compared with traditional methods, there is good dynamic performance, fewer error, which is with good robustness and stability.