Guidance, Navigation, and Separation Assurance for Local-Area UAV Networks: Putting the Pieces Together
暂无分享,去创建一个
This paper examines the guidance methodology needed to implement networks of autonomously-flown unmanned aerial vehicles (UAVs) controlled by centralized ground stations (GSs). The intended operations would take place within a local area with a diameter of less than 10 km for most applications but potentially as large as 50 100 km. UAVs in these networks are envisioned to be potentially quite small and inexpensive but capable of automated flight orientation and stability with guidance updates provided by the GS at 0.5 to 2 second intervals. The GS also provides GNSS differential corrections and integrity information to support sub-meter-level 95% navigation accuracy with 10 -7 error bounds in the 3 10 meter range. Position and timing solutions for each UAV are relayed back to the GS and support both operational (route planning) and tactical (path updating) guidance. This guidance needs to insure safe separation between UAVs within the network and (depending on the airspace used) separation from "out-of-network" UAVs as well as manned aircraft. The proposed guidance approach centers around "zones of influence" surrounding each UAV that include allowances for navigation error, UAV guidance error, and groundsystem guidance error. The amount of error allocated to each error source depends upon the degree of error correlation between each UAV and its neighbors as well as the required probabilities of safe separation that must be maintained. The ground system maintains and updates zones of influence for each UAV within its operational area and guides each UAV it controls to remain within a "zone of operations" to insure that all UAV movements it commands avoid collisions with other vehicles or the ground. This paper provides examples of how this is done and how adjustments are made to reflect changes in navigation performance and the influx of UAVs operating outside the network. 1.0 Introduction to UAV Network Concept A large number of applications have been proposed for unmanned aerial vehicles (UAVs). Today, some of these applications, such as taking pictures and monitoring particular locations on the ground, have been implemented to a limited degree with remotely-piloted UAVs operated singly in or small groups. However, requiring each UAV to be controlled by a human pilot who must carry out the UAVs mission while monitoring for and maintaining safe separation from other vehicles greatly constrains the number of vehicles that can participate safely and cost-effectively. Obtaining the full benefits of these applications will most likely involve UAVs operating autonomously and coordinating their activities in large groups. This is particularly true for data-collection and monitoring over large areas. Several examples of applications that are suited for networks of autonomous UAVs are presented in [1]. The simpler of these use the potential cost-effectiveness of autonomous UAV networks to obtain results more quickly and cheaply than existing methods using piloted aircraft or remotely-piloted UAVs. Photography and other forms of passive data collection are good examples, as the data is either not needed in real time or is independent of mission planning. For example, piloted aircraft are used today in major metropolitan areas to monitor road traffic conditions during busy periods. UAV networks could do this job better and more cheaply simply by allowing both denser and more widespread coverage. The outputs of UAV monitoring would be used in near real time to display traffic conditions and warn of accidents and bottlenecks, but the changes to traffic conditions would not require real-time changes in UAV positions or monitoring patterns. Thus, high-level changes to UAV guidance, such as repositioning and UAV air traffic control could be handled by ground personnel, although effective autonomous guidance would be much more cost-effective. Figure 1: Local-Area UAV Network Conceptual Diagram [1,2] Other applications described in [1], such as reconnaissance and surveillance, would place much greater demands on human guidance and would therefore require autonomy guidance for most, if not all, guidance functions. In a surveillance application such as monitoring for potential lawbreaking at a large shopping center or subdivision of houses, UAVs would be spread out in a standard patrol pattern most of the time. When suspicious activity is detected, it would be desirable to reposition several UAVs in real time to get a better view of the area that generated alert to confirm if it is suspicious and, if so, to follow the object of concern until security personnel can arrive. On a small scale, this repositioning could be handled by humans, but as the scale expands to kilometers, most of it will need to be done automatically. This paper expands on the local-area UAV network concept outlined in [1,2] to explain how autonomous guidance of the type described here can be implemented. Section 2.0 provides an overview of this concept, including the use of local-area differential GNSS (LADGNSS) navigation [1] and the derivation of safe separation standards [2]. Section 3.0 describes how the navigation and guidance error models for individual UAVs are combined with the separation standards to generate multiple "zones of influence" (ZoIs) for each UAV that represent the region around each UAV that must be kept clear to avoid collisions. Section 4.0 describes how the current activities of each UAV are reflected in "zones of operation" (ZoOs) that represent the region within which each UAV is allowed to maneuver while performing a particular activity. Section 5.0 describes how zones of influence and operation work together under nominal ("status-quo") conditions to assure that the overall mission is carried out safely. Section 6.0 describes how these zones are used to manage guidance under off-nominal conditions. Section 7.0 addresses the complications of sharing airspace with other users, including other UAV networks (with separate controllers) and manned aircraft. Section 8.0 summarizes the paper and described the next steps in refining this guidance methodology. 2.0 Local-Area UAV Network Concept Figure 1 illustrates the concept of local-area UAV network operations developed in [1,2]. The control station shown at the lower left is the source of local-area differential GNSS corrections and integrity information as well as real-time guidance for each UAV in the network. The LADGNSS and guidance information are separate data messages combined in the same outbound transmission. The guidance function also requires feedback from each UAV in real time, including its current position and velocity. Therefore, a two-way datalink is required and is used to relay GNSS information (such as position-domain protection levels) as well as position and velocity from UAVs to the control station. The maximum operational range of a single network of this type is limited by many factors, including the effective range of the datalink and the range beyond which LADGNSS errors grow unacceptably or become too difficult to reliably detect [1]. Upper buffer (bhigh)
[1] Christopher W. Lum,et al. A Risk Based Paradigm and Model for Unmanned Aerial Systems in the National Airspace , 2011 .
[2] Demoz Gebre-Egziabher,et al. Analysis of Unmanned Aerial Vehicles Concept of Operations in ITS Applications , 2011 .
[3] Jiyun Lee,et al. High-Integrity Local-Area Differential GNSS Architectures Optimized to Support Unmanned Aerial Vehicles (UAVs) , 2013 .