To solve the problems of basic gravity aided matching algorithm, i.e., time consuming, low precision and inefficiency when the inertial navigation system (INS) has a large initial error, a real algorithm on adaptive chaotic ant colony optimization-RD is proposed by using modified ant colony algorithm in continuous space to approach an optimizing model. The search efficiency, noise immunity and matching probability are improved by adaptively adjusting pheromone, chaos adaptive manage of ant colony algorithm search strategy, parameter and local pheromone. Experiment shows that it is more robust to INS initial error, matching efficient and good real-time calculation.