Aiming at the problem of low accuracy of positioning and navigation system during the parking of agricultural machinery, an adaptive zero velocity update algorithm containing velocity constraints and heading angle constraints was proposed, and a single-antenna BDS/INS integrated system on a self-developed small agricultural machinery was built to verify the feasibility and advantages of the proposed method. In order to ensure that the agricultural machinery remained in a stationary state during zero velocity update, a zero-velocity detector based on the acceleration of x axis (right side) and the angular velocity variance of z axis (top side) was designed through the analysis of the measured data, and a large number of misdetected zero-velocity intervals that were discretely distributed but with short durations were eliminated based on the length of the zero-velocity duration window. Due to the inability of a single antenna combination system to provide a valid estimated value of heading angle at rest, the heading angle at the moment of parking would be used as the input value for the zero-velocity update heading angle measurement equation during the corresponding parking phase. And in order to reduce the influence of the abnormal BDS positioning observations on the integrated system, Sage-Husa adaptive filtering was introduced to estimate its positioning noise in real time. The field experiments showed that the zero-velocity detection method had a high accuracy rate, which was in line with the actual driving state of agricultural machinery. During periods of BDS availability, adaptive zero velocity update improved the accuracy of heading angle, velocity, and positioning by approximately 99.3%, 93.3%, and 50%, respectively. When BDS signals were disturbed, the positioning accuracy improvement can exceed 90%. During a 60-second BDS outage, pure INS navigation errors grew rapidly and led to failure, whereas zero velocity update maintained high INS accuracy for an extended period, meeting the precision requirements for unmanned agricultural machinery operations in hilly and mountainous areas. The research result demonstrated that the method can improve the accuracy and anti-interference of single antenna BDS/INS integrated navigation system under complex working conditions, and it was also helpful to provide references for other unmanned agricultural machinery research.