As the key component of the crankshaft, the performance of the bearing bush directly affects whether it can operate normally or not. Although the deformations of the key components belong to the micrometer level, some detection methods use sensors with low sensitivity, low accuracy and large volume. Therefore, the purpose of this paper is to propose a deformation reconstruction and compensation model based on the combination of fiber Bragg grating and multilayer perceptron (FBG-MLP).
At first, according to the deformation characteristics of the bearing bush in the finite element simulation, a FBG-based sensing distribution network is designed. Then, the study compensates for the systematic errors caused by FBG pasting angles by optimizing the algorithm. Finally, the FBG-based reconstruction model is developed to depict and reconstruct the deformation of the bearing bush.
The experimental results demonstrate that the FBG-based reconstruction model using the compensated data significantly reduces deformation errors and improves the detection precision.
The model provides a new strategy for online monitoring of machining deformation for the crankshaft and its critical components, which has application prospects in ensuring the machining quality and service life of the crankshaft.
