Online learning based intelligent temperature control during polymer composites microwave curing process

Since 1980, microwave curing technology has been developed to process polymer matrix composites, and given great expectations to realize an extremely uniform and high efficient curing process. However, the uneven electromagnetic field in the microwave cavity results in an uneven temperature distribution on the surface of composite laminates, which has been seriously restricting the industrial application of the microwave curing technology. Existing methods attempted to solve the problem by the random superposition of uneven electromagnetic fields, but the results were far from meeting the high quality requirements of aerospace parts. In this paper, neural networks were used to learn the dynamic temperature behaviors of the composite under various microwave control strategies during curing, which continuously provided accurate solutions to the uneven temperature distribution monitored in real time by active compensation. Experimental results showed significant improvement in temperature uniformity compared with existing representative methods.