Abstract:Real-time and stable positioning data is essential for the UAV to perform various tasks. The traditional multi-sensor data fusion algorithm needs to know the measurement noise of sensor data, and even if there are corresponding adaptive methods to estimate the noise, most methods cannot deal with time-varying noise. In addition, traditional fusion algorithms usually are complicated, causing a large amount of calculation. In this paper, a multi-sliding window classification adaptive unscented Kalman filter (MWC… Show more
“…The impact of sensing information on UAV flight status in complex environments is also crucial. In the process of UAV path planning, many factors will cause a loss of sensor perception, and eliminating the impact of sensor noise is an issue that must be considered [31]. Literature [32] constructs a two-stage agent training method to reduce the interference of unknown noise levels through local noise observation and policy gradient optimization.…”
Unmanned Aerial Vehicles (UAVs) autonomous navigation based on reinforcement learning (RL) usually requires training agents in simulation scenarios and then transferring the trained agents to application scenarios. However, due to serious distribution mismatch between the idealized simulation scenario and the application environment and the inevitable uncertainty perception problem of airborne sensors in complex scenarios, the navigation performance of UAV under migration applications is not ideal. This work fully analyzes the factors that affect UAV navigation performance, including algorithm performance, training strategy, and state awareness. Based on the analysis results, this article proposes a framework to improve the autonomous navigation performance of UAVs in the migration process from training to application, which consists of three parts: "scenario-perception-algorithm". In addition, this paper proposes improvement strategies for each part from the perspectives of spatial features, temporal features, and perceptual denoising. We combine the proposed framework with navigation algorithms to improve the navigation decision-making performance of UAVs in migration applications under uncertainty perception. Many simulation experiments demonstrate the effectiveness of the proposed framework and its robustness to uncertainty perception.
“…The impact of sensing information on UAV flight status in complex environments is also crucial. In the process of UAV path planning, many factors will cause a loss of sensor perception, and eliminating the impact of sensor noise is an issue that must be considered [31]. Literature [32] constructs a two-stage agent training method to reduce the interference of unknown noise levels through local noise observation and policy gradient optimization.…”
Unmanned Aerial Vehicles (UAVs) autonomous navigation based on reinforcement learning (RL) usually requires training agents in simulation scenarios and then transferring the trained agents to application scenarios. However, due to serious distribution mismatch between the idealized simulation scenario and the application environment and the inevitable uncertainty perception problem of airborne sensors in complex scenarios, the navigation performance of UAV under migration applications is not ideal. This work fully analyzes the factors that affect UAV navigation performance, including algorithm performance, training strategy, and state awareness. Based on the analysis results, this article proposes a framework to improve the autonomous navigation performance of UAVs in the migration process from training to application, which consists of three parts: "scenario-perception-algorithm". In addition, this paper proposes improvement strategies for each part from the perspectives of spatial features, temporal features, and perceptual denoising. We combine the proposed framework with navigation algorithms to improve the navigation decision-making performance of UAVs in migration applications under uncertainty perception. Many simulation experiments demonstrate the effectiveness of the proposed framework and its robustness to uncertainty perception.
“…The T&E of sensor fusion engines will require a combination of ground truth data, as well as global and local metrics without ground truth data [ 78 ]. An integral problem to solve with sensor fusion will be ensuring the time synchronization and update rates between all contributing sensors have the precision and frequency required to enhance the blended solution and not inadvertently degrade the solution [ 79 ]. Without fusion or with a suboptimal fusion solution, the UA will be operating with decreased awareness to the environment, which will be a critical consideration for airworthiness officials to consider when conducting the risk assessment of the UA [ 80 ].…”
Section: Review Of Sensor Requirements For A3rmentioning
As technologies advance and applications for uncrewed aircraft increase, the capability to conduct automated air-to-air refueling becomes increasingly important. This paper provides a review of required sensors to enable automated air-to-air refueling for an uncrewed aircraft, as well as a review of published research on the topic. Automated air-to-air refueling of uncrewed aircraft eliminates the need for ground infrastructure for intermediate refueling, as well as the need for on-site personnel. Automated air-to-air refueling potentially supports civilian applications such as weather monitoring, surveillance for wildfires, search and rescue, and emergency response, especially when airfields are not available due to natural disasters. For military applications, to enable the Air Wing of the Future to strike at the ranges required for the mission, both crewed and uncrewed aircraft must be capable of air-to-air refueling. To cover the sensors required to complete automated air-to-air refueling, a brief history of air-to-air refueling is presented, followed by a concept of employment for uncrewed aircraft refueling, and finally, a review of the sensors required to complete the different phases of automated air-to-air refueling. To complete uncrewed aircraft refueling, the uncrewed receiver aircraft must have the sensors required to establish communication, determine relative position, decrease separation to astern position, transition to computer vision, position keep during refueling, and separate from the tanker aircraft upon completion of refueling. This paper provides a review of the twelve sensors that would enable the uncrewed aircraft to complete the seven tasks required for automated air-to-air refueling.
“…In 2020, an interactive multi-model Extended Kalman filter (IMM-EKF) [8] was proposed and used to solve the fusion positioning and noise problems of GPS\INS. In the same year, aiming at the problem of time-varying noise in UAV positioning, an adaptive unscented Kalman filter based on multi-sliding window classification (MWCAUKF) [9] was proposed, which was mainly used to enhance the time stability and accuracy of positioning data. In 2021, an adaptive Federated Kalman filter (AFKF) [10] algorithm was proposed to address the instability of the positioning of a single sensor.…”
Regional positioning can provide position information for objects in a certain area. In order to solve the positioning accuracy problem of regional positioning, a regional positioning system based on the integration of Inertial Navigation System (INS), Beidou System (BDS) and UWB System is proposed, so as to achieve accurate positioning in a certain area. Firstly, a multi-dimensional information fusion framework based on Extended Kalman filter (EKF), redundant information filter (RIF) and weighted fusion was proposed. In this framework, RIF_LP (Liner Prediction) algorithm and ILCTA (Information Lossless Coordinate Transformation Algorithm) are proposed. Finally, the experimental results show that the positioning fluctuation amplitude of the fusion positioning scheme is greatly reduced, and the positioning accuracy is higher than each single positioning scheme. The proposed fusion framework and algorithms have certain reference value for the construction of regional fusion positioning.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.