This paper proposes a novel aircraft morphing fuselage and its follow-up load device. The follow-up load unit model is derived, and a theoretical model of the follow-up load system is built by morphing disturbances. A back propagation (BP) neural network model is used to build the relationship between the PID and response parameters of the system. A follow-up load experimental platform is built, and experimental verification is conducted. Based on the experimental dataset and the BP neural network, an experimental proxy model of the follow-up load system is built. The Genetic (GA) and Nondominated Sorting Genetic II (NSGA-II) algorithms are used to optimize the PID of the experimental proxy model. The system response by the optimal PID outperformed that by manually tuned PID, with the response error remaining within 3%. The correctness of the theoretical model of the system and the feasibility of using the BP neural network to build an experimental proxy model and perform offline PID optimization are verified. The proposed method of optimized PID for the follow-up load system by the experimental proxy model addresses the lag issues in online PID optimization and reduces potential damage to the experimental system caused by repeated manual tuning.