Data-driven Networked Control Of Ozone-based Water Treatment Processes

In recent years,with the rapid development of national industrialization,the amount of the factories built in cities increases,which leads to the water resource becoming deteriorate.The disinfection process of the sewage treatment receives the attention from the government.Nowadays,oxychlorination is the main method in most of the sewage treatment factories.However,there are many drawbacks during the process of chlorine oxidation.Ozone,as an oxidizing agent,will not generate any harmful substance during the sewage treatment so that it can avoid the secondary pollution,and it has been widely used in many sewage treatment factories.However,the ozone-based sewage treatment technique is more expensive than the chlorine-based sewage treatment technique.Therefor if we can control the usage amount of ozone accurately,the process of sewage treatment can be much better while the cost can be saved.The main contents of this paper are as follow:The technological process of the ozone sewage treatment and dielectric barrier discharge(DBD)ozone generator are studied.Then,according to the input and output data,the simulation of control process for an ozone generator is performed by neural network training model.A data-driven control simulation study on ozone generator is carried out,which is based on the characteristics of multi-input and multi-output and nonlinear and time-varying.An ozone sewage treatment model was established,we designed the ordinary PID control system and fuzzy adaptive PID control system for the model.We conducted simulation studies on whether the model is time-varying or not.Finally,we propose the data-driven networked predictive control method for the ozone generator so as to actively compensate for random delay packet dropout and packet disorder during the network transmission.Thus the long-distance control can come true,and the amount of ozone generated can be adjusted in time,which means that the energy consumption can be reduced.At the same time,the closed-loop cascade networked predictive control simulation for the sewage treatment is implemented,which shows that the amount of ozone generation can be changed by the condition of the cesspit,and the energy consumption can be further reduced.