Online adaptation for autonomous unmanned systems driven by requirements satisfaction model

Luo, Yixing; Zhou, Yuan; Zhao, Haiyan; Jin, Zhi; Zhang, Tianwei; Liu, Yang; Barthaud, Danny and Yu, Yijun (2022). Online adaptation for autonomous unmanned systems driven by requirements satisfaction model. Software and Systems Modeling (Early Access).

DOI: https://doi.org/10.1007/s10270-022-00981-7

Abstract

Autonomous unmanned systems (AUSs) emerge to replace human operators for achieving better safety, efficiency, and effectiveness in harsh and difficult missions. They usually run in a highly open and dynamic operating environment, in which some unexpected situations may occur, leading to violations of predefined requirements. In order to maintain stable performance, the AUS control software needs to predict in advance whether the requirements will be violated and then make adaptations to maximize requirements satisfaction. We propose Captain, a model-driven and control-based online adaptation approach, for the AUS control software. At the modeling phase, apart from the system behavior model and the operating environment model, we construct a requirements satisfaction model. At runtime, based on the requirements satisfaction model, Captain first predicts whether the requirements will be violated in the upcoming situation; then identifies the unsatisfiable requirements that need to be accommodated; and finally, finds an optimal adaptation for the upcoming situation. We evaluate Captain in both simulated scenarios and the real world. For the former, we use two cases of UAV Delivery and UUV Ocean Surveillance, whose results demonstrate the Captain ’s robustness, scalability, and real-time performance. For the latter, we have successfully implemented Captain in the DJI Matrice 100 UAV with real-world workloads.

Viewing alternatives

Metrics

Public Attention

Altmetrics from Altmetric

Number of Citations

Citations from Dimensions

Item Actions

Export

About

Recommendations