Obstacle avoidance algorithm for UAVs in unknown environment based on distributional perception and decision making

Xu, Zhijie; Wei, Ruixuan; Zhou, Kai; He, Ran

doi:10.1109/cgncc.2016.7828936

Cited by 3 publications

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Marchidan et al [10] proposed a method based on local parameterized guidance vector fields to address the collision avoidance problem. Xu et al [11] proposed a method on the basis of distributed perception and decision-making to solve the obstacle avoidance problem for unmanned aerial vehicles (UAVs). Mujumdar et al [12] proposed a method relying on both the nonlinear geometric guidance and differential geometric guidance to avoid the reactive collision at low altitudes of UAVs.…”

Section: Introductionmentioning

confidence: 99%

Guidance Design for Escape Flight Vehicle Using Evolution Strategy Enhanced Deep Reinforcement Learning

Hu,

Wang,

Gong

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Guidance commands of flight vehicles can be regarded as a series of data sets having fixed time intervals, thus guidance design constitutes a typical sequential decision problem and satisfies the basic conditions for using the deep reinforcement learning (DRL) technique. In this paper, we consider the scenario where the escape flight vehicle (EFV) generates guidance commands based on the DRL technique and the pursuit flight vehicle (PFV) generates guidance commands based on the proportional navigation method. Evasion distance is described as the minimum distance between the EFV and the PFV during the escapeand-pursuit process. For the EFV, the objective of the guidance design entails progressively maximizing the residual velocity, which is described as the EFV's velocity when the evasion distance occurs, subject to the constraint imposed by the given evasion distance. Thus an irregular dynamic max-min problem of extremely large-scale is formulated. In this problem, the time instant when the optimal solution (i.e., the maximum residual velocity satisfying the evasion distance constraint) can be attained is uncertain and the optimum solution is dependent on all the intermediate guidance commands generated before. For solving this challenging problem, a two-step strategy is conceived. In the first step, we use the proximal policy optimization (PPO) algorithm to generate the guidance commands of the EFV. The results obtained by PPO in the global search space are coarse, despite the fact that the reward function, the neural network parameters and the learning rate are designed elaborately. Therefore, in the second step, we propose to invoke the evolution strategy (ES) based algorithm, which uses the result of PPO as the initial value, to further improve the quality of the solution by searching in the local space. Extensive simulation results demonstrate that the guidance design method relying on the proposed ES-enhanced PPO algorithm is highly effective.INDEX TERMS Deep reinforcement learning; max-min problem; guidance design; proximal policy optimization (PPO); evolution strategy (ES).

show abstract

Section: Introductionmentioning

confidence: 99%