Research And Application Of Large Nitrile Latex Reactor's Automatic Temperature Control System

Reactor is a common vessel for production in chemical industry,and the precision of its temperature control is very important to the quality of production.With the continuous research of automatic theory,the automatic temperature control system has been widely applied in the reactors' production.As the large volume and complicated parameters of the reactors,the result of automatic temperature control has always been unsatisfied.Thus,the research of automatic temperature control system and the temperature control strategies should be further studied in order to build a automatic temperature control system which is effective and stability.The whole process of nitrile latex reactor's automatic temperature control system is introduced in this thesis from the following two parts: one is the research of PID controller based on BP neural network algorithm in reactor's temperature control,the other is the design of PLC's control system.And the past and the present situation of the reactor's automatic temperature control system is also introduced in it.After the analysis of the control object and the logical relations between the processes,the distributed software redundant structure of master and slave station was chosen to control the whole system.The system achieved to transfer data between PLC and modules based on the Profibus-DP communication protocol,and the MPI network was built in order to achieve the communication between upper computer and PLC.The influence of the neural network-PID and the traditional PID in the reactor's temperature control process is analyzed in this thesis,and a multi-mode PID controller based on the combination of the traditional PID and the neural network-PID is choosen to control the temperature.This controller can effectively improve the heating speed,reduce the overshoot,and increase the robustness of the system.Besides,the design of PLC's hardware and software,the design of the upper computer,the monitor system is also introduced detailedly in this thesis.It has been almost one year since the automatic temperature control system of large nitrile latex reactor put into use.During this periods,the system ran stably and could meet the enterprise's temperature control requirement.The whole system has improved the quality and efficiency of the production,and has created much more profit for the enterprise.However,there is still some deficiencies in the structural design of the neural network,which needs to be further optimized.