Abstract: The shape of aircraft with ‌telescopic deformation structures‌ is complex, and the distribution of ‌aerothermal‌ data varies ‌significantly‌. It is challenging for traditional surrogate models to capture the ‌aerothermal‌ data distribution of ‌telescopic structures‌, making effective prediction of ‌aerothermal heating‌ on structural surfaces difficult. Based on ‌a‌ conditional diffusion model, ‌a Heating-MLP Diffusion (HMD) method for deformable structures‌ is proposed, comprising two processes: forward diffusion and reverse denoising. In the forward diffusion process, the original ‌aerothermal‌ data is gradually ‌corrupted‌ until it becomes pure Gaussian noise. In the reverse denoising process, ‌using‌ the shape and operating condition parameters of the deformed structure as ‌conditional inputs‌, a fully connected neural network predicts the noise added at each diffusion step, thereby learning the implicit ‌aerothermal‌ data distribution characteristics and enabling ‌aerothermal heating‌ prediction on the surface of ‌aircraft telescopic wings‌. Numerical simulation data validate the proposed model. Experimental results demonstrate that compared with Gaussian processes, neural processes, and neural networks, the ‌conditional diffusion model-based‌ aerothermal prediction method achieves higher accuracy, with a mean absolute percentage error of approximately 10%. This proves its effectiveness in predicting ‌aerothermal heating‌ on ‌high-speed aircraft‌ wings with telescopic deformation structures.