OPTIMIZATION METHOD OF TOTAL NITROGEN REMOVAL PROCESS BY NEURAL NETWORK AND ADVANCED MODEL PREDICTIVE CONTROL

摘要

The present invention developed a DA-901 (ammonolysis), and DA-902 (urease) prediction model for the total nitrogen in the waste water to the optimum processing for the treated effluent, operation optimization model and multi-variable predictive control system in the TN process and real-time optimal control by the model applied to automatic control system using a software design apply, with the stabilization, energy saving, TN process predicted control the automatic control system of a new type that can be achieved, such as advanced wastewater treatment process It relates to an optimal method for treating wastewater.; Optimal method for treating wastewater provided by the present invention and the TN component prediction system for managing a total nitrogen component values ​​to predict the total nitrogen change in real time and analyzing the correlation between the operating parameters of the total nitrogen in the waste water, wherein consider the TN component prediction system to set the process output prediction value in real time to the future control based operation optimization system and, based on a process model for providing them after calculating optimum operating conditions of the respective operation parameters to the target value in the subsequent control stage section that it is determined whether or not to proceed along the trajectory, and comprises a multi-variable predictive control system to compensate for calculating a set of control output values ​​in the control section in order to minimize the error is characterized.