Propagation Characteristics of Measurement Error of Low-Cost Sensors in Solar-Powered Unmanned Aerial Vehicle (UAV) Flight Control

The large measurement bias and noise fluctuations of low-cost sensors hinder their use in industrial-grade UAVs. However, for large-scale, low-speed fixed-wing UAVs, attitude stabilization throughout the entire flight is more critical than trajectory accuracy. This attribute enables the practical application of low-cost sensors, with solar-powered UAVs exemplifying this category. In light of this, this paper proposes a state estimation method for low-cost flight controllers that incorporates attitude, heading, and navigation hierarchies, and analyzes the error propagation characterization in estimation, control law solving, and control mode switching. The results show that the cross-axis error and root-mean-square noise of the sensor can be relaxed to 1.9 and 1.3 times of the nominal, which can be used as relaxed stability margins for sensor applications; estimation error stabilization guarantees the stability of error propagation through control laws and state switches. This paper demonstrates that measurement error bounds for low-cost sensors can be relaxed and that adaptations at the estimation, control, and navigation levels can facilitate the use of low-cost sensors and support the cost-effective development of navigation systems.