Pursuit-Evasion Problems Involving Two Pursuers and One Evader

Makkapati, Venkata Ramana; Sun, Wei; Tsiotras, Panagiotis

doi:10.2514/6.2018-2107

Cited by 8 publications

(5 citation statements)

References 14 publications

(12 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The agents' initial positions were randomly generated by a 2-variant normal distribution. Furthermore, to estimate the solution of HJI equations (i.e., ( 5) and ( 7)), FPK equations (i.e., (6) and ( 8)), and optimal control input (i.e., (13)), 2 critic NNs, 2 mass NNs, and an actor NN are constructed. Additionally, a random noise is injected to the control input from 0s to 50s to increase the NN approximators' exploration.…”

Section: Simulation Resultsmentioning

confidence: 99%

“…Recently, some of the researches explored a novel type of pursuit-evasion problem for multiple pursuers and evaders due to the enormous gain from the larger population of agents. For instance, [6] studied the pursue evasion problem with two pursuers and one evader; [7] used multiple pursuers, i.e. unmanned aircraft systems (UAS), to capture the ground vehicle.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mean Field Game and Decentralized Intelligent Adaptive Pursuit Evasion Strategy for Massive Multi-Agent System under Uncertain Environment

Zhou

2020

Preprint

View full text Add to dashboard Cite

In this paper, a novel decentralized intelligent adaptive optimal strategy has been developed to solve the pursuit-evasion game for massive Multi-Agent Systems (MAS) under uncertain environment. Existing strategies for pursuit-evasion games are neither efficient nor practical for large population multi-agent system due to the notorious "Curse of dimensionality" and communication limit while the agent population is large. To overcome these challenges, the emerging mean field game theory is adopted and further integrated with reinforcement learning to develop a novel decentralized intelligent adaptive strategy with a new type of adaptive dynamic programing architecture named the Actor-Critic-Mass (ACM). Through online approximating the solution of the coupled mean field equations, the developed strategy can obtain the optimal pursuit-evasion policy even for massive MAS under uncertain environment. In the proposed ACM learning based strategy, each agent maintains five neural networks, which are 1) the critic neural network to approximate the solution of the HJI equation for each individual agent; 2) the mass neural network to estimate the population density function (i.e., mass) of the group; 3) the actor neural network to approximate the decentralized optimal strategy, and 4) two more neural networks are designed to estimate the opponents' group mass as well as the optimal cost function. Eventually, a comprehensive numerical simulation has been provided to demonstrate the effectiveness of the designed strategy.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mean Field Game and Decentralized Intelligent Adaptive Pursuit Evasion Strategy for Massive Multi-Agent System under Uncertain Environment

Zhou

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…In Figure 5, we can point out the attacker-winning cases of Figure 4 (b), (c), (f), for which ρ x a t eventually reaches R Ω S = 10. Noticeably for Figures 5 (b) and (c), ∆ρ x a t |u d DM ≥ ∆ρ x a t |u d P P as in (29). Take (b) for example.…”

Section: Algorithm 1 Attacker Behaviorsmentioning

confidence: 94%

“…Typical and popular strategies utilized in PE games are Constant Bearing (CB), Line of Sight (LoS), and PP guidance laws [28]. Constant Bearing assumes the knowledge of the attacker's instantaneous velocity as well as its position [29], whereas the defender only has access to the noisy observation of the attacker in our problem. Consequently, CB is not suitable to apply to our problem.…”

Section: A Baseline: Pure Pursuit Strategymentioning

confidence: 99%

See 1 more Smart Citation

Protective Mission against a Highly Maneuverable Rogue Drone Using Defense Margin Strategy

Sung¹,

Hiltebrandt-McIntosh²,

Kim³

et al. 2022

Preprint

View full text Add to dashboard Cite

The current paper studies a protective mission to defend a domain called the safe zone from a rogue drone invasion. We consider a one attacker and one defender drone scenario where only a noisy observation of the attacker at every time step is accessible to the defender. Directly applying strategies used in existing problems such as pursuit-evasion games are shown to be insufficient for our mission. We introduce a new concept of defense margin to complement an existing strategy and construct a control strategy that successfully solves our problem. We provide analytical proofs to point out the limitations of the existing strategy and how our defense margin strategy can be used to enhance performance. Simulation results show that our suggested strategy outperforms that of the existing strategy at least by 36.0%p in terms of mission success.

show abstract

Optimal Evading Strategies and Task Allocation in Multi-player Pursuit–Evasion Problems

Makkapati

Tsiotras

2019

Dyn Games Appl

View full text Add to dashboard Cite

Pursuit-Evasion Problems Involving Two Pursuers and One Evader

Cited by 8 publications

References 14 publications

Mean Field Game and Decentralized Intelligent Adaptive Pursuit Evasion Strategy for Massive Multi-Agent System under Uncertain Environment

Mean Field Game and Decentralized Intelligent Adaptive Pursuit Evasion Strategy for Massive Multi-Agent System under Uncertain Environment

Protective Mission against a Highly Maneuverable Rogue Drone Using Defense Margin Strategy

Optimal Evading Strategies and Task Allocation in Multi-player Pursuit–Evasion Problems

Contact Info

Product

Resources

About