An Intelligent Photoelectric Detection Approach Based on Neural Network

Photoelectric detection system will be affected by a variety of noise and environmental temperature changes and produce certain errors. To realize the function of error compensation and improve the detection accuracy, an adaptive neural network detection algorithm based on photoelectric detection is established. Aiming at the defects of neural network, such as easily falling into local minimum, slow training speed and poor generalization ability, the local optimization ability of PSO is applied to improve and optimize the learning factor of neural network. The simulation results show that compared with the traditional algorithm, the improved scheme is easier to be implemented and it has faster convergence speed. In the practical application of photoelectric detection principle test box, the nonlinear error caused by temperature change in the illumination detection system can be effectively compensated, and the prediction accuracy can be also improved.