Optimal Design Of Traveling Wave Induction Heating Device

Heat Treatment, HT technology directly affects the performance of the steel wire products.If using the traditional HT technology will lead to different heated speed on each parts of theheated strip, and then lead to different oxidation degree each parts of the heated strip, even causethat the strip is damaged and other serious problems because of local temperature of the strip toocold or overheat. In order to solve these problems, a novel HT technology arises at the moment.Traveling Wave Induction Heating, TWIH emerge as the times require. TWIH technologyhas all advantages of traditional induction heating technology. TWIH technology not only canproduce more uniform temperature distribution than traditional induction heating technology, butalso has other significant advantages such as small vibration and low industrial noise. AlthoughTWIH technology has so many advantages, it is still in the phase of theoretical research. Thereare lots of issues need to be further studied. After consulting a large number of relevant literaturedata and summarizing the previous research results, this paper establishes TWIH device3Dmodel, and then analysis TWIH device eddy current field, temperature field and their coupledfield. In order to obtain more uniform temperature distribution at strip outlet side, this paperestablishes TWIH device optimization model. The data interface problem between ImprovedArtificial Searching Swarm Algorithm, IASSA and ANSYS has been solved in this paper. Then,this article optimizes TWIH device with IASSA which is proposed firstly in this article.Artificial Searching Swarm Algorithm, ASSA is a novel bionic intelligence algorithm.ASSA simulates soldiers perform a specific search tasks and find search target throughperforming appropriate rules. Thus, the optimization design problem is solved correspondingly.This paper summarizes the basic principle and implementation steps of ASSA; test theperformance of ASSA with a series of complex function; analysis detailed the influence ofvarious parameters for convergence speed and optimization ability. According to the analysisresults about the influence of various parameters for ASSA performance and consulting relatedliterature, this paper improves the step parameter, introduces dynamic step and puts forwardIASSA. Then this paper tests the performance of IASSA with a series of complex function. Theeffectiveness of IASSA is verified by test results.The computation time is too long because of the vast computation in optimization process.This paper explores the prediction ability of Artificial Neural Network, ANN to solve this problem.Conducting uniform experiment, the finite element calculated results by ANSYS is the trainingsample data to train ANN and to predict the average relative error of element joule heat on thestrip surface. The trained ANN replaces ANSYS calculation. Then, trained ANN combine withIASSA to optimize TWIH device. This way can reduce computation in optimization process,furthermore reduce the optimization time. This way is a beneficial attempt for solving the vastcomputation of electromagnetic inverse problem.