Background <p>Carbon fiber reinforced plastic (CFRP) has emerged as the predominant material for wing skins, which are susceptible to internal crack damage. If such damage is not detected in a timely manner, it may pose a threat to flight safety. The Lamb wave‑based positioning technology for detecting internal crack damage in CFRP wing skins encounters challenges such as noise interference arising from multi‑sensor signal modal aliasing, the difficulty in extracting complex signal features, redundancy in high‑dimensional time‑domain signals, and the high computational complexity related to hyperparameter optimization in deep learning models. This study presents a novel damage positioning method that leverages multimodal signal processing and Bayesian optimization within a deep stochastic configuration network.</p> Methods <p>Initially, a numerical simulation and data generation framework was meticulously designed. Secondly, the permutation entropy and Higuchi fractal dimension methods were employed to analyze sensor signals corresponding to damage at various locations. Thirdly, signal feature extraction was accomplished using continuous wavelet transforms and Markov transition fields. Subsequently, functional principal component analysis was employed to retain the most significant damage characteristics, thereby effectively reducing the data dimensionality. Then, the Bayesian optimization algorithm was applied to optimize the hyperparameters of the deep stochastic configuration network, effectively reducing computational complexity. Finally, the optimized deep stochastic configuration network was used to accurately accomplish the damage positioning task.</p> Results <p>This paper verified the effectiveness of the method through numerical simulation and real experiments. When locating internal cracks in the CFRP wing skin, the accuracy of BODSCN was significantly better than that of the comparison models CNN, ELM, RVFLN and DSCN.</p> Conclusions <p>This study thus provides a high‑precision and high‑efficiency approach for the health monitoring of CFRP aerial vehicle composite material structures.</p>