Industial Microwave Drying System Performance Assessment, Modeling And Optimization Based On Intelligent Control

High power Microwave(MW) heating is an efficient and non-pollution heating with the advantage of average, quick and no chemical energy exchange. MW heating is widely applied in industry, especially in material drying, such as ceramic drying process described in the article. At present, the research on MW drying mostly focus on drying process, while the study on drying automation, energy saving and pollution reduction is rare. Improper control of MW will lower the product selection. Meanwhile, the poor performance of controller or control strategy would cause a waste of energy, event incident like fire and radiation leak. As author’s practice, the selection of drying process in some ceramic factory is just 90%, and a fire incident might happen about every 2 month which will cause equipment damage and unscheduled break down. Increasing the select and reduce incident can greatly improve the enterprise’s efficiency. So to research and optimize the control strategy in MW drying process has a great practical significance.The main contents of this thesis are as following:Firstly, this paper analyzes the MW drying process and its control fundamental, and review the current control strategy. Based on the shortage in current control system, control performance assessment and intellectual control for MW drying control are proposed.Secondary, after introducing the theory of controller performance assessment, the method of controller performance assessment based on history data benchmark is introduced to MW drying control system. Combined with the characteristic of MW drying, the thesis construct a comprehensive performance index as the benchmark of performance assessment, subsequently carry on an experimental analysis.Finally, the application of Neural Network(NN) in aspect of system identification and prediction is introduced, as well as the theory of self-adaptive dynamic programming. MW drying is non-linear, coupling and large lag process, therefore an Elman NN model is developed to identify and simulate the control object. Meanwhile self-adaptive dynamic programming is used in optimizing MW drying control. Lastly, the MW drying NN model and optimizing control based on ADP is trained and tested by Matlab Simulink...