Automatic Collimation of Optical Path Based on BP-PID Control

In the process of detecting the fiber quality, the effect of optical collimation is directly related to the measurement results. The collimation of optical path can be realized by manual adjustment, but it takes one and a half hours to complete collimation and measurement, and the average standard deviation of the measured divergence angle is 0.0712. In order to solve the problem of manual adjustment, a method based BP-PID algorithm is adopted to realize automatic collimation of optical path. The simulation results of BP-PID (PID control based on BP neural network) algorithm show that it has the advantages of fast response, good robustness and small overshoot. The experimental data show that the alignment time of the optical path is reduced by about 5 times and the average standard deviation of the measured divergence angle is 0.0617. This method not only improves the collimation efficiency, but also keeps the data more stable.