UAV height measurement algorithm based on covariance intersection and three-step delay time fusion

With the rapid expansion of unmanned aerial vehicles (UAVs) technology, these UAVs are now extensively utilized in both military and civilian sectors. UAV terrain-following, a subset of flight control technology, has found extensive use in several fields, including topographical mapping, agricultural planting, and demining surveys. A good height-measuring technique is vital in complicated application environments with high requirements for precision in height measurement. This work provides a data fusion algorithm that combines many sensors at different heights to address the limitations of using a single height sensor in various application scenarios. Firstly, the data collected from the sensors is initially processed using a three-step delayed time fusion algorithm, which considers a larger amount of historical data. Then, a covariance intersection (CI) fusion algorithm is used to fuse data from different sources, and the data is further complementarily fused by using an accelerometer, improving the height measurement accuracy. Finally, an SG-HANTS filtering technique is utilized to efficiently handle data that contains both noise and missing values. The results indicate that the technique presented in this research when compared to a similar approach, decreases the root mean square error by 15.21% and reduces the maximum error by 0.0336m. The UAV height measurement algorithm is cost-effective, highly accurate, and reliable.