A Dynamic Task Allocation Algorithm for Heterogeneous UUV Swarms

Abstract

Aiming at the task allocation problem of heterogeneous unmanned underwater vehicle (UUV) swarms, this paper proposes a dynamic extended consensus-based bundle algorithm (DECBBA) based on consistency algorithm. Our algorithm considers the multi-UUV task allocation problem that each UUV can individually complete multiple tasks, constructs a “UUV-task” matching matrix and designs new marginal utility, reward and cost functions for the influence of time, path and UUV voyage. Furthermore, in view of the unfavorable factors that restrict the underwater acoustic communication range between UUVs in the real environment, our algorithm complete dynamic task allocation of UUV swarms with optimization in load balance indicator by the update of the UUV individual and the task completion status in the discrete time stage. The performance indicators (including global utility and task completion rate) of the dynamic task allocation algorithm in the scenario with communication constraints can be well close to the static algorithm in the ideal scenario without communication constraints. The simulation experiment results show that the algorithm proposed in this paper can quickly and efficiently obtain the dynamic and conflict-free task allocation assignment of UUV swarms with great performance.

Optimization of Dynamic Task Allocation forMulti-UAV Systems: Search and Rescue Scenario

“…The authors in [26] proposed a dynamic algorithm for task allocation for unmanned underwater vehicle swarms. They achieved this by extending the CBBA algorithm and considering multiple parameters such as time, path, and UUV tour as marginal utility functions to optimize the baseline CBBA algorithm further.…”

Survey on the Developments of Unmanned Marine Vehicles: Intelligence and Cooperation

“…There have been studies on cooperation between multiple homogeneous UUVs to solve task assignment problems and avoid underwater obstacles to operate in unknown environments. Wu et al [ 173 ] attempted to solve the task assignment problem using three types of UUVs: portable, lightweight, and heavyweight; they perform three different work types: detection, tracking, and structure. By improving consensus-based bundle algorithm (CBBA)-based algorithms, the proposed that extended CBBA (ECBBA) and dynamic ECBBA (DECBBA) can solve the task assignment problem via operation in two iterative steps: task bundle construction and conflict resolution by improving consensus-based bundle algorithm (CBBA)-based algorithms.…”

“…The second is DECBBA, a dynamic task allocation algorithm that changes the entire allocation process from individual time steps to multiple static task allocations to convert the previous static task allocation problem into a dynamic problem. Figure 8 presents a schematic of the UUV swarm network used in [ 173 ]. In addition, Yan et al [ 174 ] tracked multiple AUV formation controls and AUV formation trajectories using the proposed underwater obstacle avoidance algorithm when obstacles were detected using multiple AUVs.…”

Research on the interception capability of distributed UUV swarm

“…Tao and Zhang [4] introduced the key technology necessary and ASW application for the underwater unmanned swarm. Wu and Gao [5] proposed a dynamic extended consensus-based bundle algorithm based on a consistency algorithm. Shu and Zhang [6] proposed a concept of the near-shore unmanned cluster collaborative antisubmarine operation based on USV and UAV.…”