Damage Evaluation Method of CFRP Structures Based on PCA and Random Forest Algorithm

The accurate and timely assessment of carbon fiber reinforced plastics (CFRP) structure damage is very important for ensuring the safety of aviation and aerospace composite structures. This paper studied the CFRP damage evaluation method by using principal component analysis (PCA) and random forest algorithm. First, the structure was excited to obtain the dynamic response signal. Thus, the frequency characteristics were obtained. Then, PCA was used to extract the damage feature. On this basis, the structural damage evaluation model based on random forest was established. Finally, the damage assessment experiment system of composite structures was established and experimentally verified. The testing result showed that the proposed method by using principal component analysis and random forest algorithm is effective.