Burning rate research of NEPE propellant containing CL-20 on the base of improved Canny edge detection and PSO-BP

A sealed high-pressure laser ignition experiment platform is established to investigate the burning rates of NEPE propellant containing CL-20 under nitrogen and air environments. The entire combustion process of the propellant is captured by a high-speed camera. Captured images during the steady burning period are processed to obtain the curves and positions of the propellant burning surface with the help of an improved Canny edge detection technique with excellent accuracy. Subsequently, the regression velocity of the propellant burning surface is calculated to determine the burning rate of the NEPE propellant containing CL-20. Furthermore, the Particle Swarm Optimization algorithm -Back Propagation neural network model (PSO-BP) is developed for predicting the burning rate, which is compared with the traditional BP algorithm. The PSO-BP model exhibits significantly lower errors and indicates more accurate predictions and better model performance. Improved Canny edge detection and PSO-BP neural network provide a novel approach for measuring and predicting the burning rate of high-energy solid propellants.