Multilayer Ultrasonic Echo Structured Sparse Model and its Application

The effective separation of the multilayer ultrasonic echo is the key for layered defective imaging and feature information identification; it is an urgent bottleneck problem to be addressed in the multilayer ultrasonic echo processing field. In this paper, starting from sound-electric equivalent model of ultrasonic echo, combining with multilayer signal features, a structured sparse model of multilayer ultrasonic is built. Studies have shown that multilayer ultrasonic echo could be characterized as Hankle matrix and Toeplitz matrix separation and recognition model. This can be applied to multilayer ultrasonic stress detection field for metal matrix composites structure.