The Study And Application Of The Decoupling Control System Of Basis Weight, Moisture And Ash Of Paper

The paper has become an important part in people's lives, and the higher quality of thepaper is required. People are constantly to use the new technology and new equipment in paperproduction process,to improve the quality and yield of the paper. Computer automatic controlis in the requirement of the development. In order to pursue higher economic benefit, the betterand more advanced automation control is used in papermaking industry.The basis weight, moisture and ash of paper are the most important quality indexes of thepaper. The strict control of the three indexes can improve the economic benefits of the paperproduction. At the conclusion of the current situation of the development of both at home andabroad,this paper briefly introduces the control characteirstics of basis weight,moisture andash of paper and the whole process. It also tells us some control algorithms of quality controlsystem and the advantages and disadvantages of each control algorithm. According to theresearch of the paper multivairable system and the data rfom the mill, the paper builds up themathematical model of coupled control system.According to the characteirstics of the mathematical model of the control system, the paperdesigns the two controllers, the conventional PID controller and diagonal recursive neuralnetwork controller. The DRNN controller is based on the diagonal recurrent neural network. Itcan perfectly achieve to decouple multi-vairable control system by using diagonal recurrentneural network to identiyf the system model and adjust the PID parameters. Under the sameconditions, neural network control compared with PID control has saved time by50%or so, theovershoots of the control is decreased by50%, the response time is shorter and the system hasgood performance. The results declare that the PID control by unpacking coupling has powerfuladaptive ability, short response time, fast anti-interference, good decoupling effect.In order to adapt to the demand of actual production process, the paper design theexperiment based on the THJS-3advanced process control device with the diagonal recursiveneural network PID decoupling control algorithm. The expeirmental curve shows theinterference of the two tanks is just about7%.The results declare that the decoupling controlsystem has good steady state and dynamic performance. Finally, some summarization andsuggestions are given in the last chapter,which tells us the feasibility and effectiveness of thesystem, the problems in the study and the good foundation for the further study on this subject.