Effective Real-Time Unmanned Air Vehicle Path Planning in Presence of Threat Netting
Fu Xiaowei,Gao Xiaoguang
DOI: https://doi.org/10.2514/1.i010077
2014-01-01
Journal of Aerospace Information Systems
Abstract:T HE problem of accurately and efficiently planning safe paths is critical for a number of unmanned air vehicle (UAV) applications. This work focuses on UAV path planning through battlefields littered with various enemy threats, such as radar-guided surface-to-air missiles (SAMs). A particular threat may receive information from nearby threat locations with which it can communicate, and the lethality of the threat may increase based on the received information. This process is called threat netting [1,2], and its essential effect is a greater degree of lethality. This Note is devoted to the problem of real-time path planning for UAVs in the presence of threat netting, and each threat is considered a radar-guided SAM(Fig. 1); thismeans that, before themissile launch, the radarmust continuously track theUAV for some response period, and after the launch, accurate missile guidance requires that the radar system maintain tracking of the UAV during the flight of the missile [3]. Although some studies in the literature offer path planning algorithms in the presence of threat netting, there is no approach that integrates the response time and the guidance time of a SAM.Other studies offer path planning algorithms in the presence of radar-guided SAMs, but there is no approach that integrates the process of threat netting. This Note proposes a method of effective real-time UAV path planning in a radar-guided SAM netting environment and presents its results. Although the current literature on UAV path planning does not contain models of integrated threat netting specifically for radar-guided SAMs, some studies describe integrated threat netting and others address radar-guided SAMs. Brief reviews of the relevant results in these studies are as follows. References [1,2] introduce a real-time route planning technique called Sparse A* Search (SAS), which allows a number of mission-dependent route constraints to be incorporated into the search process and allows for an accurate and efficient consideration of the threat netting effect. Two steps must be performed before the search process begins. The first step involves marking the grid cells contained within the various threat intervisible regions. These cells aremarkedwith that particular threat’s identification number in different ways. If a grid cell falls withinmore than one threat’s intervisible region, then that grid cell receivesmultiplemarkings from each of the respective threat locations. The second step involves generating a threat netting cost lookup (TNCLU) table. The TNCLU table contains cost information concerning the lethality effects of netting for the various threat types and locations. Many different factors can be considered when generating the lethality values in the lookup table. Once the appropriate grid cells have been marked and the TNCLU table has been generated, the SAS algorithm is run to find a safe route. During the SAS process, both the total route cost information and the previously traversed threat information are considered. The values in the TNCLU table are computed a priori based on previous experiences with the particular threat types in various scenarios, but in general, it is hard to determine an accurate threat cost to construct the TNCLU table in advance. In these papers, another problem is that the radar cross sections (RCSs) of theUAVs are not discussed. Previousworks [3,4] have introduced a type of real-time path planning algorithm for unmanned combat air vehicles (UCAVs) in the presence of the radar-guidedSAMs. The problem is formulated using the framework of an interaction between three subsystems: the aircraft, the radar, and the missile. The main features of this integrated model are as follows. The aircraft RCS depends explicitly on both the aircraft’s aspect and its bank angles; hence, the RCS and aircraft dynamics are coupled. The probabilistic nature of radar tracking is accounted for in the model; namely, the probability that the aircraft has been continuously tracked depends on the aircraft RCS and range. Finally, the factors required tomaintain tracking before missile launch and during missile flyout are also modeled. Based on this model, the problem of UCAV path planning is formulated as a minimax optimal control problem.However, theseworks do not show how to plan theUCAVpath if a particular SAM receives information from a nearby SAM. Other works in the literature [5–7] discuss how to plan the trajectories for flight vehicles in the presence of threat zones such as radar detection regions. Reference [5] presents the linear approximation of a nonlinear radar detection risk function with integer constraints and indicator (0–1) variables and shows that this approach can yield UAV trajectories depending on the acceptable risk of detection. Reference [6] establishes a methodology tominimize the peak of the RCSs of autonomous precision guidedmunitions (APGMs) as they ingress to a selected target through a radar threat environment. This research demonstrates how route planning may be combined with the simultaneous specification of aerodynamically feasible yaw and bank angles to significantly reduce APGM observability. In [7], the detection model and the UAV’s dynamics are represented as a linear program subject to mixed integer constraints. This mixed integer linear program is then solved to search for all feasible solutions, and it produces the best path based on the user-specified parameters.