The invention discloses a self-adaptive moon soft landing orbit rapid optimization controller. The orbit rapid optimization controller comprises a simulation module, an optimization module and an execution module. In the simulation module, a moon soft landing orbit is separated into a plurality of small segments, a thrust direction angle of each node of segmentation parts and a corresponding moment are utilized to obtain a thrust direction angle changing curve through fitting, and a whole soft landing orbit is obtained through soft landing kinetic model simulation. The optimal thrust direction angle and the optimal soft landing terminal moment of each node are obtained through optimization by the optimization module. The optimization module adopts an improved intelligent algorithm, an improved intelligent optimization method controls the size of a subgroup, adds disturbance factors, increases diversity of search, and improves convergence performance of the algorithm according to an inertia weight which is self-adaptively changes according to an evolution state The controller can rapidly obtain a moon soft landing orbit which enables fuel consumption to be minimum through optimization, and finally realize optimal control of the orbit.