Study On Wood Drying Model Optimization And Intelligent Control Methods

Wood drying is an important process for wood industry.Efficient wood drying process is crucial to improve wood product quality and reduce the waste of raw wood material.Improvement of wood drying equipment can reduce manual workload and the difficulties of quality control.However,wood drying equipment manufacturing enterprises are facing the problem of insufficient high-tech application and innovation capabilities,resulting in large errors in sensing system of wood drying equipment,low automation levels,and poor control accuracy.To overcome these obstacles of wood drying equipments,this study established a wood drying model which is suitable for wood drying controller design.On this basis,high-performance,low-energy wood drying controllers that can meet the actual production requirements were designed.Wood drying moisture content prediction model was established to provide a basis for adjusting the temperature and equilibrium moisture content.Auto regressive with exogenous(ARX)was used to build the wood dry moisture content prediction model.Recursive least square method(RLS)and forgetting factor RLS method were used to identify system parameters.Results show that the prediction accuracy of RLS is higher than that of FF-RLS,but the convergence speed of FF-RLS is shorter.Wood drying temperature and equilibrium moisture content control models were built.Non-linear Hammerstein-Wiener method was used to establish wood drying process temperature and equilibrium moisture content control model.Relationship between opening degree of steam injection valve,heating valve,moisture exhaust valve and the dry bulb temperature,equilibrium moisture content was described.Piecewise Linear,Sigmoid network,Saturation function and Wavelet network were used to identify the non-linear part of the input and output of the Hammerstein-Wiener model.Results show that the wood drying temperature prediction model based on Piecewise Linear function has the highest accuracy,and the wood drying equilibrium moisture content model based on Sigmoid network prediction accuracy is the highest.Traditional proportional-integral-derivative(PID)controller for wood drying was optimized.In order to solve the problems that the traditional PID controller for wood drying,back propagation neural network(BP neural network)and radial basis function neural network(RBF neural network)were utilized to optimize the wood drying PID controller.Simulation and drying test results show that compared with the traditional PID method,BP-PID and RBF-PID controllers have significantly improved the control performance of the wood drying system,while shortening the drying time and reducing energy consumption.Wood drying sliding mode controllers(SMC)were designed and optimized.In order to further improve the precision and efficiency of the wood drying control system and the quality of wood drying,sliding mode wood drying controllers based on exponential reaching law and variable rate reaching law were proposed.To reduce the chattering and softening control signal of the SMC wood drying controller,RBF neural network and fuzzy algorithm were used to adjust the gain of the SMC controller switching function.Simulation and drying test results show that:RBF-SMC wood drying temperature controller and equilibrium moisture content controller have the most superior control performance.In the drying test,compared with other SMC controllers,RBF-SMC controller has the shortest drying time and minimal energy consumption.Compared with BP-PID and RBF-PID controllers,SMC and its optimized controllers significantly improve the performance of the wood drying control system,reducing energy consumption and improving wood drying quality.This study establishes a wood drying model suitable for wood drying controller design.On this basis,wood drying PID controller was optimized with BP and RBF neural network.In order to further improve the control performance of the system,wood drying SMC controoler were proposed.Simulation and wood drying test results show the feasibility and effectiveness of the presented methods.The study provides valid solutions for improving wood drying efficiency and quality,which has significant meanings for improving the intelligence and precision of wood drying equipment.