| Abstract |
: |
To ensure a failsafe flight of a drone, adequate mitigation is conducted for accurate detection of the battery discharge level and problematic situations that may cause the drone to fall below a certain level. On one hand, if the mitigation starting time is too late, the drone would fall down before finishing failsafe landing, which is against the safety requirements. On the other hand, if the mitigation starting time is too early, the drone will still have some battery power left when it finishes landing, which is in contradiction to the efficient use of the battery resources. Therefore, in this study, we analyze how the mitigation starting time changes according to the frequency and aim to experimentally discover an adequate level of data sensing frequency. The experiment targets a 3DRobotics product, Pixhawk firmware, which controls a quadcopter, and utilizes a battery model developed mathematically. Experimental results revealed that the mitigation starting time was almost identically maintained for the frequency of certain domains and delayed rapidly for the low-frequency domain. |
