Design And Simulation Of Varnish Production Distributed Control System

Vanish is one of the most important kinds of paint in the family decoration. It is widely brushed on the surface of furniture, floors, doors, windows and cars. Varnish production belongs to the scope of fine chemical production and the process is extremely complex. At present, varnish production is dominated by the batch production. As the main equipment of varnish production, the batch reactor includes all the difficulties of batch process control. It covers the concepts of sequence control, logic control and loop control.For efficient production, we must precisely control the reactor temperature, which is the most important parameter of the batch production. The effect of temperature control has a great influence on the quality and output of the varnish products. In the reactor, the mechanism of polymerization is very complex, which makes the batch production has the characteristics of time-delay, non-linear and time-varying. Therefore, it is the key and difficult point to achieve the precise control of the reactor temperature.Based on the automatic control system of varnish production in a chemical plant, we designed a production process control system consisting of operation stations, control stations and process control networks. AI-T8000Distributed Control System (DCS) acquires the field data and outputs the control signal in the varnish production process; the operation stations running Wizcon Supervisor enable operators to monitor and control the production process in real time; the process control networks transmit the information and control signal between operation stations and control stations.In this paper, we wrote the conventional control program by Ladder (LD) language and the function blocks by Structured Text (ST) language. The design of conventional control program is based on the production system. This paper introduced in detail the control principle of the1st and8th varnish production subsystems and the solvent pump system. In addition, the paper also introduced the function blocks to be called by the conventional control program, including input signal processing, flow accumulating, temperature curve setting, split range control and pump on-off control. Finally, in order to research the reactor control method, we used the experimental data to establish a BP Neural Network Model of the reactor. According to the model’s characteristic of non-linear, we chose the Gain Scheduling PID to simulate the reactor temperature control process, and compared the result with the one that was controlled by the conventional PID. The simulation result showed that the BP Neural Network Model can accurately reflect the relationship between the input and the reactor temperature and that Gain Scheduling PID control method can precisely control the reactor temperature.