Bionic navigation based on geomagnetic feature search does not require prior geomagnetic maps. It is one of the effective navigation methods for autonomous underwater vehicle (AUV) underwater. It has the characteristics of high concealment and low cost. The main problem is the low efficiency of search and navigation. This paper proposes a new bionic navigation algorithm based on Differential Evolution Algorithm (DE), which can make rapid magneto tropic heading decisions based on geomagnetic trends. By improving the mutation mechanism of the evolution strategy, the algorithm obtains better individuals in the population, thereby improving the efficiency of search and navigation. Simulation experiment results show that the algorithm can enable AUV to complete navigation tasks, and is superior to Evolutionary Algorithms (EAs) in terms of navigation search distance, time-consuming, and navigation accuracy.