A dynamical trap made of target-tracking chasers

TL;DR

The paper introduces a discrete-element-method model to trap a fleeing target using a swarm of chasers that apply velocity- and position-based tracking forces. A four-group cross-domain arrangement successfully captures the target by forming a confinement around a repulsive-energy minimum, whereas a single-group approach fails due to force imbalance. The velocity-tracking component is essential for tight pursuit, while the target-prediction (TDP) position guidance mainly improves capture efficiency. The findings suggest a non-lethal, UAV-assisted wildlife management strategy and advance understanding of pursuit-evasion dynamics in multi-agent systems.

Abstract

We propose a dynamical trapping system composed of multiple chasers subject to target-tracking forces utilizing the velocity and position information of a single escaping target. To successfully capture the target, dividing chasers into multiple groups while each group approaching its assigned destination in the proper vicinity of the target is essential. Moving direction synchronization between the target and its chasers is crucial to the capturing process, while guiding chasers to the predicted position of the target in future only improves the efficiency of capture but is not indispensable. Potential applications of our trapping system include capturing live animals such as bears invading a human residential area.

Figures (7)

• Figure 1: (Color online) Specifications of the position-related term $\beta \hat{n}^{ct}$ of the tracking force. A chaser pursuing the target is subject to a force of magnitude $\beta$ along the center-to-center direction $\hat{n}^{ct}$ between the chaser's current position $c$ (filled green circle) and a random point $t$ (open red circle) in (a) a square domain of side length $L$ or (b) a cross-like domain, with the shortest distance $L_1$ away from the target and composed of four identical sub-domains of size $L_2-L_1$ by $L_3$, centered on $T$ (filled red circle), the target's current position (CP strategy) or waypoint (after time $dt$, TDP strategy). The $N-1$ chasers are evenly divided into four groups from $i$ to $iv$ with their $t$'s in the corresponding sub-domains labeled accordingly if the cross-like domain is used.
• Figure 2: (Color online) Representative data of a failed capture. (a) Separation distance $\Delta$ between the target and the center of mass of a single chaser group as a function of time $t$. The corresponding configuration of the system when the value of $t$ is at 1, 2, or 3 (red open circle) is shown in Fig. \ref{['fig:1division_details']}. (b) Velocity $v$ of the target (red) or the center of mass of its chasers (green) as a function of time $t$. The data are obtained using a single realization.
• Figure 3: (Color online) Representative configurations of a failed capture, where the system is composed of one target (red) and a single chaser group (green or blue so that different configurations are visually discernible), when the value of time $t$ is at 1, 2, or 3, as labelled in Fig. \ref{['fig:1division']}a. All particles are drawn with identical arbitrary size and their relative positions preserved; added arrows show each particle's current direction of movement, with length being proportional to its speed.
• Figure 4: (Color online) Representative data of a successful capture. (a) Separation distance $\Delta$ between the target and the center of mass of the four chaser groups as a function of time $t$. The corresponding configuration of the system when the value of $t$ is at 1, 2, or 3 (red open circle) is shown in Fig. \ref{['fig:4division_bis_details']}. (b) Velocity $v$ of the target (red) or the center of mass of its chasers (green) as a function of time $t$. The data are obtained using a single realization.
• Figure 5: (Color online) Representative configurations of a successful capture, where the system is composed of one target (red) and four chaser groups (i: light green, ii: light blue, iii: green, and iv: blue) when the value of time $t$ is at 1, 2, or 3, as labelled in Fig. \ref{['fig:4division_bis']}a. All particles are drawn in the same way as in Fig. \ref{['fig:1division_details']}.