A Hybrid Strategy for Target Search Using Static and Mobile Sensors

Kashino, Zendai; Nejat, Goldie; Benhabib, B.

doi:10.1109/tcyb.2018.2875625

Cited by 20 publications

(15 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The search-planning approach proposed in this paper was compared to the sensor-centric WiSAR search-planning method outlined in reference [62]. In reference [62], the motion trajectories of a team of multiple agents were planned to carry out the scheduled deployment of an optimal static-sensor network while allowing the agents to be on the look-out for the target as they were moving from one deployment position to another. Namely, in this approach, the (global) unconstrained optimal sensor network is planned first.…”

Section: Comparative Studymentioning

confidence: 99%

“…As will be shown below through extensive experiments, it is hypothesized here that, although in certain search scenarios, the two hybrid approaches would perform comparably, in cases where the sensors could not be deployed at their (global) optimal times and positions, as dictated by reference [62], due to few available agents, the agent-centric approach proposed herein would perform significantly better than the sensor-centric approach detailed in reference [62]. Since it would be difficult to predict ahead of time which approach the user should select in a real WiSAR scenario, our recommendation is that prior to the start of a search for a lost person, both approaches be simulated in parallel and the hybrid plan produced by the method with the greater potential to succeed quickly be chosen for implementation.…”

Section: Comparative Studymentioning

confidence: 99%

“…Targets were assumed to be found if they passed within the sensing radius of a sensor or an agent. The primary focus of these experiments was to investigate the impact of various agent-to-sensor combinations on the search success of our method versus the one proposed in [62]. The search success rate for a given search scenario was selected as the metric of scrutiny because it most accurately represents the true concern at hand (i.e., the probability of finding the target, in a real-life mobile-target search scenario).…”

Section: Large Sets Of Simulationsmentioning

confidence: 99%

“…As can be noted from Table 1, in cases where the sensors could not be deployed at their (global) optimal times and positions, as dictated by [62], due to few available agents, the proposed agent-centric method performed significantly better (with 98.9% and 99.9% confidence by confidence interval analysis) than the sensor-centric approach-last two rows of the table. Namely, a large gap in detection rate performance is attributable to multiple extremely late sensor deployments, as determined by reference [62]. Moreover, a "snowball effect" was observed where one late sensor deployment led to further and increasingly other late sensor deployments.…”

Section: Large Sets Of Simulationsmentioning

confidence: 99%

“…Moreover, a "snowball effect" was observed where one late sensor deployment led to further and increasingly other late sensor deployments. Unlike the method proposed in reference [62], the approach presented in this paper is robust to low agent-sensor ratios due to the trajectory-constraints sensor deployments are subject to. It should be noted that both hybrid methods performed significantly better when static sensors were present than when they were not.…”

Section: Large Sets Of Simulationsmentioning

confidence: 99%

See 4 more Smart Citations

A Sensor-Network-Supported Mobile-Agent-Search Strategy for Wilderness Rescue

et al. 2019

View full text Add to dashboard Cite

Mobile target search is a problem pertinent to a variety of applications, including wilderness search and rescue. This paper proposes a hybrid approach for target search utilizing a team of mobile agents supported by a network of static sensors. The approach is novel in that the mobile agents deploy the sensors at optimized times and locations while they themselves travel along their respective optimized search trajectories. In the proposed approach, mobile-agent trajectories are first planned to maximize the likelihood of target detection. The deployment of the static-sensor network is subsequently planned. Namely, deployment locations and times are optimized while being constrained by the already planned mobile-agent trajectories. The latter optimization problem, as formulated and solved herein, aims to minimize an overall network-deployment error. This overall error comprises three main components, each quantifying a deviation from one of three main objectives the network aims to achieve: (i) maintaining directional unbiasedness in target-motion consideration, (ii) maintaining unbiasedness in temporal search-effort distribution, and, (iii) maximizing the likelihood of target detection. We solve this unique optimization problem using an iterative heuristic-based algorithm with random starts. The proposed hybrid search strategy was validated through the extensive simulations presented in this paper. Furthermore, its performance was evaluated with respect to an alternative hybrid search strategy, where it either outperformed or performed comparably depending on the search resources available.

show abstract

Section: Comparative Studymentioning

confidence: 99%

Section: Comparative Studymentioning

confidence: 99%

Section: Large Sets Of Simulationsmentioning

confidence: 99%

Section: Large Sets Of Simulationsmentioning

confidence: 99%

Section: Large Sets Of Simulationsmentioning

confidence: 99%

See 3 more Smart Citations

A Sensor-Network-Supported Mobile-Agent-Search Strategy for Wilderness Rescue

et al. 2019

View full text Add to dashboard Cite

show abstract

Balancing the trade-off between cost and reliability for wireless sensor networks: a multi-objective optimized deployment method

Chen

et al. 2022

Appl Intell

View full text Add to dashboard Cite

show abstract

Cooperative multi-agent target searching: a deep reinforcement learning approach based on parallel hindsight experience replay

Zhou

Liu

Shi

et al. 2023

Complex Intell. Syst.

View full text Add to dashboard Cite

Multi-agent multi-target search strategies can be utilized in complex scenarios such as post-disaster search and rescue by unmanned aerial vehicles. To solve the problem of fixed target and trajectory, the current multi-agent multi-target search strategies are mainly based on deep reinforcement learning (DRL). However, the training of agents by the DRL tend to be brittle due to their sensitivity to the training environment, which makes the strategies learned by the agents fall into local optima frequently, resulting in poor system robustness. Additionally, sparse rewards in DRL will lead to the problems such as difficulty in system convergence and low utilization efficiency of the sampled data. To address the problem that the robustness of the agents is weakened and the sparse rewards exist in the multi-objective search environment, we propose a MiniMax Multi-agent Deep Deterministic Policy Gradient based on the Parallel Hindsight Experience Replay (PHER-M3DDPG) algorithm, which adopts the framework of centralized training and decentralized execution in continuous action space. To enhance the system robustness, the PHER-M3DDPG algorithm employs a minimax learning architecture, which adaptively adjusts the learning strategy of agents by involving adversarial disturbances. In addition, to solve the sparse rewards problem, the PHER-M3DDPG algorithm adopts a parallel hindsight experience replay mechanism to increase the efficiency of data utilization by involving virtual learning targets and batch processing of the sampled data. Simulation results show that the PHER-M3DDPG algorithm outperforms the existing algorithms in terms of convergence speed and the task completion time in a multi-target search environment.

show abstract

A Hybrid Strategy for Target Search Using Static and Mobile Sensors

Cited by 20 publications

References 56 publications

A Sensor-Network-Supported Mobile-Agent-Search Strategy for Wilderness Rescue

A Sensor-Network-Supported Mobile-Agent-Search Strategy for Wilderness Rescue

Balancing the trade-off between cost and reliability for wireless sensor networks: a multi-objective optimized deployment method

Cooperative multi-agent target searching: a deep reinforcement learning approach based on parallel hindsight experience replay

Contact Info

Product

Resources

About