This paper presents an intelligent morphing decision method, utilizing neural networks, designed to address the successive autonomic decision challenge faced by hypersonic morphing flight vehicles (HMFV). Initially, a dynamic model of a HMFV with a successive adjustable sweep angle is developed. Then, considering the external disturbance and the heat flux density constraint of the flight process, by segmenting the reference track and add interference, a trajectory sample set is generated by the Legendre pseudo-spectral method. The neural network is trained to automatically adjust the sweep angle during the gliding phase by analyzing the flight state, making intelligent decisions. Finally, the developed intelligent decision-making algorithm is utilized for trajectory optimization. The simulation results demonstrate a significant enhancement in the distance achieved through intelligent decision morphing compared to program-based morphing. The trajectory can be adjusted to enhance the range based on the intelligent decision-making sweep angle of the current flight state.