Learning-based cable force planning for Mars ascent vehicle cold launch test
摘要
The cold launch of the Mars ascent vehicle (MAV) is a new method for launching spacecraft from the surface of extraterrestrial planets. To verify its effectiveness of this method, a low-gravity simulation test on the Earth using the cable-driven mechanism is required. This paper analyzes the essence of the low-gravity environment simulation tests with cable-driven mechanisms, and proposes a new evaluation index based on the kinematic similarity for cold launch tests. Based on this index, this paper adopts a 5-cable mechanism configuration, and proposes a new learning-based cable force planning method, which fully considers the characteristics of cold launch test. The results show that compared to the classical method, the low gravity simulation test using the new method can achieve the test purpose more effectively by making the motion of MAV in the simulation test more similar to that in the real Mars gravity environment.