Resilient Flight Control: An Architecture for Human Supervision of Automation

Shared Control Between Pilots and Autopilots: An Illustration of a Cyberphysical Human System

“…The details of the training are explained in the "Validation with Human-in-the-Loop Simulations" section. It is shown that SCA1 in [11] and SCA2 in [13] lead to better performance, as conjectured therein, when the human expert completes his or her assigned roles in the respective architectures. The resulting solution is an embodiment of an efficient cyberphysical human system, a topic that is of significant interest of late.…”

“…The foundations for SCA1 and SCA2 come from the results of [11]- [13], with [11] and [12] corresponding to SCA1 and [13] corresponding to SCA2. In [11], the assumption is that the autopilot is designed to accommodate satisfactory operation under nominal conditions, using the requisite tracking performance.…”

“…The SCA2 architecture utilized in [13] differs from that of SCA1, as mentioned previously, as the trading action from the autopilot to the pilot is replaced by a supervisory action by the pilot. In addition to using CfM, this architecture utilizes a second principle from cognitive engineering denoted as graceful command degradation (GCD).…”

“…A GCD can then be viewed as a control variable tuned to permit a system to reach its targeted CfM. The role of tuning this variable so that the desired CfM is retained by the system is relegated to the pilot in [13]. Specifically, a parameter n is transferred to the pilot from the autopilot, which is shown to result in an ideal tradeoff between the CfM and GCD by the overall closed-loop system with SCA2.…”

Abnormal Noise Suppression and Detail Protection for High-Resolution Range Profile of GM-APD Lidar

“…Geiger-mode of avalanche photodiodes (Gm-APD) Lidar range profile has rich range information and a high-speed data acquisition ability, which can greatly improve the system's long-distance detection capability, and has a wide range of applications in the fields of autopilot [1], active detection [2], and precision guidance [3]. Abnormal noise is the main factor that needs to be filtered out in HRRP.…”