Research On Temperature Modeling And Control Of Tunnel Kiln Based On Fuzzy Neural Network

Tunnel kiln is an important equipment in the production of refractory materials,and its control level affects the quality of the material directly.Therefore,the temperature control strategy plays an important role in the tunnel kiln control system.However,how to design a good temperature control strategy for tunnel kiln which can satisfy the produce requirements,enhance the quality of the product,reduce the energy consumption and reduce the pollution,becomes an urgent problem need to be solved.The control of tunnel kiln system mainly includes three aspects:Firstly,according to the given temperature curve to control the temperature of the tunnel kiln and ensure the quality of baked products.Secondly,maintain optimum air-fuel ratio.Because only when the air-fuel ratio is the best,the fuel can be fully burning.Otherwise,not only waste fuel but also pollute the environment.Thirdly,adjust the wind withdrawals and the air supply to maintain pressure balance in tunnel kiln and to guarantee the security of industrial operation.Because the tunnel kiln is a nonlinear time-varying system with the characteristics of large inertia,delay and strong coupling,it is not easy to establish the mathematical model precisely.According to the characteristics of the tunnel kiln,the second-order plus lag time model of the tunnel kiln is identified by using improved particle swarm optimization algorithm based on the data from the field,and the temperature model is proved to be effective.Various interference factors exist in the actual production process,and the change of model parameters will lead to poor control performance.The controller parameters need to be adjusted to get good control effect,which is not conductive to the automatic control of the system.For the model mismatch phenomenon mentioned above,the fuzzy neural network controller based on fuzzy control and neural network is used in this thesis.As an intelligent control,fuzzy neural network integrated both the advantages of fuzzy reasoning and neural network.It does not require accurate mathematical model of the controlled object,and only need data collected in the field to train the network can achieve good control results.Because the global search ability of particle swarm optimization algorithm is powerful in the early time and in the late its local search ability is bad;BP algorithm is badly influenced by the initial parameters and the global search ability is bad in the early time,but its local search ability is great in the late.This thesis integrated the advantages of particle swarm optimization algorithm and BP algorithm to use PSO-BP algorithm to optimize the parameters of fuzzy neural network off-line.Then use BP algorithm to optimize the parameters of fuzzy neural network on-line.Simulation results show that,in the case of model mismatch,the fuzzy neural network controller can effectively overcome the negative influence of model mismatch.The system has higher control precision,faster tracking speed,smaller overshoot,stronger anti-interference performance and good control performance,so it can meet the requirements of the tunnel kiln temperature control.