Design Of Compound Controller For MDF Glue Applying System Based On SVM

China is lack of forest resource,and the per capita forest saving amount of our country is only 0.01m~3. With the rapid development of the public economy and rising of People's living standards , the demand of wood increases year by year,and the contradiction of Timber's supply and demand has become increasingly outstanding. The wood-based panel can use wood resource efficiency, it is a main industry in the wood domain because it has high increment and high-tech content. Medium density fiberboard's development is the fastest in the wood-based panel industry, and its advantages are lightweight and high strength,close in structure,uniform texture,surface finish,not easy to cause deformation and so on. China's MDF production has developed quite rapidly,and our country has been the biggest country of production of medium density fiberboard in the world.In the process of MDF production, glue applying is an important working procedure, since the physic mechanic function,the quality and production cost of MDF are all influenced by it greatly. The task of applying glue is to apply glue to fiber according to the changes of the amount of input fiber with a reasonable ratio of glue. In the process of applying glue,the glue quantity and fiber quantity should keep a accurate proportion and high dynamic balance. During actual production, the glue applying process is often interfered by a variety of factors,so glue applying system is a non-linear,time-varying and coupling system. For this system, it's difficult to build an exact model using classical identification method, and the traditional PID control can not meet the control requirement of system.According to the problem that it is hard to set accurate inverse model of glue applying system,the method that using least squares support vector machine for identifying the inverse model is proposed in this article. Support vector machine is a small sample study machine with the advantages that global optimum,simple structure, strong generalization ability and so on,which overcomes over learning of the neural network. In this paper, the inverse model of glue applying system is set up with neural network and support vector machine respectively,and the simulation results show that the model set by least squares support vector machine has higher precision.In this paper,according to the characteristics of the glue applying system that non-linear and control accuracy is not high, a complex control method which combine reverse control and PID control is proposed,and in the complex control method, the inverse model of the glue applying system is used as the feedforward controller,and the feedback control is realized by PID controller. The controller takes advantages of the inverse control and PID control respectively, has the characteristic of simple structure ,and its control effect is proved to be very well through simulation.