Method Of Excitation Parameters Optimization Of Eddy Current Thermal Imaging Detection System Based On Genetic Algorithm

There are many motivating factors that affect the eddy current effect of thermal imaging detection, and the combination of different incentive parameters can lead to different detection effect. In order to give full play to the advantages of eddy current thermal imaging detection technology, it is necessary to study the incentives of excitation parameters and optimization of excitation parameters of the incentive level.Based on 45 # steel plate as the object, the electromagnetic- thermal coupling field finite element simulation model is build based on steel plate contains internal defects, by factors and levels to carry on the design of excitation parameters, combined with the experimental analysis of the reliability of the simulation model, based on the design of genetic algorithm to optimize the simulation model of incentive conditions, make steel plate defect area and normal area characteristic difference degree maximization, make steel plate are more likely to detect defects.First of all, the orthogonal experiments are designed by the coil number of turns, actuation time and lift-off distance. the three-dimensional induction heating model is established based on the orthogonal experimental data. Through experiment contrast maximum temperature, minimum temperature and average temperature, the error is within 6.60%. Using the method of partial least-squares regression parameters is constructed and motivate the mathematical model of thermal effect indexes, and quantitative analysis of the influence the electromagnetic heat effect of the three incentives parameters on the incentive degree of the influence of thermal effect, model is obtained by the partial least squares method between observation and prediction error is within 8.91%, indicates that the building of mathematical model can be more accurate characterization of the relationship between the electromagnetic parameters and the incentives thermal effect.Then to prepare for subsequent optimization, the three-dimensional model of induction heating for dimension reduction, containing defects of two-dimensional induction heating model is established, in order to save time cost optimization time step, shorten the calculation time of the simulation, to join the new excitation parameters of the coil position design of four factors and four levels after 16 set of orthogonal experiment, the simulation and experiment of orthogonal experiment, the comparison of five different areas of the temperature and the average temperature of the steel plate surface, the error is within 7.23%, and suggests to establish the two-dimensional induction heating model containing defects can be good for subsequent optimization algorithm of the model.At last, the GA optimization model is set up. We use the finite element simulation instead of the objective function, and determine the constraint condition of parameter optimization. The GA program stops after 30 times. The optimal excitation parameters is 8 turns, location of 0.005 miles, liftoff of 0 mile, time of 70 seconds. According to eight sets of validation experiments, the result shows that the defect recognition ability under the optimal incentive is more better than other groups after treat the infrared heat picture.