With the rapid development of navigation technology, the application fields of unmanned technology are expanding constantly, including the low-dynamic scenarios, such as unmanned agricultural machinery. In a low-dynamics scenario, the classic dual-antenna BeiDou Navigation Satellite System (BDS)/micro-electro-mechanical system inertial navigation system (MEMS INS) integrated navigation algorithm has two limitations. First, the reliability of a priori system noise is significantly reduced because of the large number of turning and bumping maneuvers. Second, the accuracy of the dual-antenna BDS heading angle and a priori measurement noise cannot satisfy the necessary requirements in a complex environment. To solve these problems, an adaptive BDS/MEMS INS navigation algorithm based on low-dynamic carrier-motion characteristics and dual-antenna constraint information is proposed. This new method can identify turning and bumping states and appropriately adjust the MEMS INS system noise matrix. To avoid the influence of a low accuracy heading angle, the heading angle measurement information was replaced by position and velocity measurement information. Simultaneously, the velocity and distance vectors between the dual antennas are added as prior constraints that can adaptively adjust the BDS measurement noise matrix. The experimental results show that the positioning and heading angle accuracies of the proposed method are controlled within centimeters and 1°, respectively, thus significantly improving the navigation performance in the carrier turning, bumping, and occlusion environments. This research will be helpful in low-dynamic applications such as unmanned agricultural machinery.