Aluminum alloy thick plate stress detection method based on PSO-GSA-GRNN model

The invention discloses an aluminum alloy thick plate stress detection method based on a PSO-GSA-GRNN model. The method comprises the following steps: S1, selecting a material, selecting a sample blank, processing the sample blank into a tensile sample, treating the sample blank, putting the treated sample blank into a material testing machine, carrying out a tensile experiment, respectively standing in a thermostatic bath, measuring sound time differences at different temperatures by using ultrasonic waves, and calculating to obtain stress; s2, establishing a stress detection model, integrating the dendritic neural network (DD) and sound time difference parameters of traditional ultrasonic detection into a deep learning model, and establishing a dendritic neural network expression by taking the temperature, the elongation rate and the sound time difference as input and taking the real stress as output; according to the aluminum alloy thick plate stress detection method based on the PSO-GSA-GRNN model, the ch