Artificial Intelligence Deep Learning-Based Life Prediction and Fatigue Durability FEM Analysis Based on Control Arm Fatigue Accumulation

In this study, the life prediction and fatigue durability analysis based on artificial intelligence deep learning were mainly based on finite element analysis (FEA) and fatigue theory.The data-based artificial intelligence PRUL prediction was simulated based on Run to Failure data, that is, the sensing data of the conditions that occurred in the control arm during the execution from the last maintenance activity to the next activity.Based on the Pitch Angle signal calculated from the vehicle dynamics model during driving, random noises such as Random Road Profile were added to create a virtual sensor input signal, and then the previously designed algorithm was applied to calculate the actuator driving angle, and the simulation was performed to obtain the results.The basic characteristics of the control arm were FEM simulated and the displacement variation analysis was analyzed according to the failure.The Von Mises Stress analysis was performed for the structural destruction prediction and torsional analysis for load distribution.The PRUL prediction can be considered a special form of survival analysis, and basically, two types of technologies were considered for predicting the PRUL: model-based technology and data-based technology.Data-model-based techniques use physical models to predict the occurrence of component failures over time, while data-driven techniques predict PRUL based solely on past observations without making assumptions about how components may fail over time, which is why this study was conducted to derive the results