Pressure sensing recognition of FBG array based on random forest algorithm

In order to improve the precision of static load pressure recognition and identify the position of the applied force accurately, a fiber Bragg grating (FBG) flexible sensor array is proposed in this work. Numerical analysis for the package thickness (4 mm) and package position (2 mm from the bottom) of the FBG flexible sensor is performed using COMSOL, and optimal package thickness (4 mm) and package position (2 mm from the bottom) are selected in the analysis. By using 12-FBGs layout method and random forest algorithm, the position and load prediction model is established. The results show that the average error of the distance between the prediction points of coordinates X-Y and static load F and the real sample points is 0.092. Finally, to verify the proposed models, the pressure sensing experiments of the flexible FBG array are carried out on this basis. The weights of 100 g to 1 000 g are applied to different regions of the flexible sensor array one by one in accordance with a certain trajectory. The variation of each FBG wavelength was taken as the input of the stochastic forest prediction model, and the coordinate position and the static load size F were taken as the output to establish the prediction model. The minimum distance error between the actual point and the predicted point was calculated by experiment as 0.034 91. The maximum is 0.248 1, and the mean error is 0.151 5. It is concluded that the random forest prediction model has a good prediction effect on the pressure sensing of the flexible FBG sensing array.