Robust decision making for UAV air-to-ground attack under severe uncertainty

Abstract

As unmanned aerial vehicles (UAVs) are used more and more in military operations, increasing their level of autonomous decision making becomes necessary. In uncertain battlefield environments, when making sovereign decisions, UAVs must choose low-risk options. An integrated framework is proposed for UAV robust decision making in air-to-ground attack missions under severe uncertainty. In the offline part of the framework, the battlefield scenarios are analyzed and an influence diagram is built to represent the decision situation. In the online part, the UAV evaluates the alternative actions for every scenario, and then the optimal robust action is chosen, using the robust decision model. Results of simulation show that the proposed approach is feasible and effective. The framework can support UAVs in making independent robust decisions under circumstances which require immediate responses under severe uncertainty, and it can also be extended to applications in more complex situations.

Keywords

unmanned aerial vehicle / severe uncertainty / robust decision making / influence diagram

References

[1] KimM H, BaikH, LeeS. Resource welfare based task allocation for UAV team with resource constraints [J]. Journal of Intelligent & Robotic Systems, 2014, 8: 1-17

[2] PongpunwattanaA A, RysdykR. Real-time planning for multiple autonomous vehicles in dynamic uncertain environments [J]. Journal of Aerospace Computing, Information, and Communication, 2004, 1(12): 580-604

[3] YangY, PolycarpouM M, MinaiA A. Multi-UAV cooperative search using an opportunistic learning method [J]. Journal of Dynamic Systems Measurement and Control-Transactions of the ASME, 2007, 129(50): 716-728

[4] UreN K, ChowdharyG, ChenY F. Distributed learning for planning under uncertainty problems with heterogeneous teams [J]. Journal of Intelligent & Robotic Systems, 2014, 74(1/2): 529-544

[5] LianZ, DeshmukhA. Performance prediction of an unmanned airborne vehicle multi-agent system [J]. European Journal of Operational Research, 2006, 172(2): 680-695

[6] AlighanbariM, HowJ P. A robust approach to the UAV task assignment problem [J]. International Journal of Robust and Nonlinear Control, 2008, 18(2): 118-134

[7] BertuccelliL F. Robust decision-making with model uncertainty in aerospace systems [D]. Boston: Massachusetts Institute of Technology, 2008

[8] BertuccelliL F, WuA, HowJ P. Robust adaptive Markov decision processes: Planning with model uncertainty [J]. IEEE Control Systems Magazine, 2012, 32(5): 96-109

[9] ReganH M, Ben-HaimY, LangfordB, WilsonW G, LundvergP, AndelmanS J, BurgmanM A. Robust decision-making under severe uncertainty for conservation management [J]. Ecological Applications, 2005, 15(4): 1471-1477

[10] LempertR J, GrovesD G, PopperS W, BankesS C. A general, analytic method for generating robust strategies and narrative scenarios [J]. Management Science, 2006, 52(4): 514-528

[11] DoD. Formulating and modeling robust decision-making problems under severe uncertainty [D]. Melbourne: The University of Melbourne, 2008

[12] HowardR A, MathesonJ E. Influence diagrams [J]. Decision Analysis, 2005, 2(3): 127-143

[13] DetwarasitiA, ShachterR D. Influence diagrams for team decision analysis [J]. Decision Analysis, 2005, 2(4): 207-228

[14] GÓMezM, BielzaC, DEL PozoJ A F, Rios-InsuaS. A graphical decision-theoretic model for neonatal jaundice [J]. Medical Decision Making, 2007, 27(3): 250-265

[15] KjaerulffU B, MadsenA LBayesian networks and influence diagrams: A guide to construction and analysis [M], 2008New YorkSpringer114-139

[16] ShachterR D. Evaluating influence diagrams [J]. Operations Research, 1986, 34(6): 871-882

[17] CharnesJ M, ShenoyP P. Multistage Monte Carlo method for solving influence diagrams using local computation [J]. Management Science, 2004, 50(3): 405-418

[18] CanoA, GomezM, MoralS. A forward-backward Monte Carlo method for solving influence diagrams [J]. International Journal of Approximate Reasoning, 2006, 42(1): 119-135

[19] BielzaC, GÓMezM, ShenoyP P. A review of representation issues and modeling challenges with influence diagrams [J]. Omega, 2011, 39(3): 227-241

[20] BielzaC, GÓMezM, ShenoyP P. Modeling challenges with influence diagrams: Constructing probability and utility models [J]. Decision Support Systems, 2010, 49(4): 354-364

[21] RenooijS, WittemanC. Talking probabilities: Communicating probabilistic information with words and numbers [J]. International Journal of Approximate Reasoning, 1999, 22(3): 169-194

[22] WittemanC, RenooijS. Evaluation of a verbal-numerical probability scale [J]. International Journal of Approximate Reasoning, 2003, 33(2): 117-131

[23] ChengC, LinY. Evaluating the best main battle tank using fuzzy decision theory with linguistic criteria evaluation [J]. European Journal of Operational Research, 2002, 142(1): 174-186

[24] HwangC, LaiY, LiuT Y. A new approach for multiple objective decision making [J]. Computers & Operations Research, 1993, 20(8): 889-899

[25] WangX-m, QinJ-c, ZhangQ-l, ChenW-j, ChenX-long. Mining method optimization of Gu Mountain stay ore based on AHP-TOPSIS evaluation model [J]. Journal of Central South University: Science and Technology, 2013, 44(3): 1131-1137

[26] BoranF E G S, KurtM, AkayD. A multi-criteria intuitionistic fuzzy group decision making for supplier selection with TOPSIS method [J]. Expert Systems with Applications, 2009, 36(8): 11363-11368

[27] DağdevirenM, YavuzS, KikincN. Weapon selection using the AHP and TOPSIS methods under fuzzy environment [J]. Expert Systems with Applications, 2009, 36(4): 8143-8151