| Dimethyl carbonate(DMC)is a common chemical raw material and chemical intermediate.In this paper,the 100,000 tons/year transesterification production plant of a chemical enterprise was used as the object,and the following research was carried out.The transesterification process for the production of dimethyl carbonate equipment mathematical modeling,the unit module of the material balance,energy balance,gas and liquid balance,etc.A dynamic simulation structure based on the steady-state process was established,and the sequential module method was used to solve and calculate the structure,and the parameters of the process parameters were debugged according to PFD and P&ID drawings.Finally,the configuration diagram of the whole process is drawn according to the actual field conditions,and functions such as accident handling,chain control and system scoring can be developed to guide the start and stop work.Finally,a set of simulation system with simulation training function is formed.Based on the dynamic simulation of the whole process of the dimethyl carbonate plant,the soft sensing modeling of its important equipment is studied in this paper.Support vector machine(SVM)analysis regression method was applied in this measurement modeling,based on dynamic simulation data.The feed flow fluctuation of the reactive distillation column is predicted and the predicted anastomoses are more than 99%,the extracted components of the reactive distillation column are predicted and the predicted anastomoses are more than 98%,and the components of the DMC at the top of the reactive distillation column are predicted and the predicted anastomoses are more than 96%.Through the soft sensing modeling of reactive distillation column,the prediction and analysis of each index of the column are realized.Aiming at the problems of flow fluctuation and intense control response in the dimethyl carbonate unit,the advanced process control(APC)of part of the unit is studied.In this paper,dynamic matrix control(DMC)is used as the design framework,and model identification,feedback correction and rolling optimization are carried out according to the process characteristics.The dynamic matrix control weakens the fluctuation caused by the disturbance of liquid level on the flow rate and greatly improves the stability of the process. |
