重访机制驱动的多无人机协同动目标搜索方法

为实现多无人机高效捕获灰色任务区域内的移动目标，考虑传感器探测概率与虚警概率，提出了重访机制驱动的协同搜索规划（RMD-CSP）方法，以降低目标遗漏与误判概率。考虑无人机飞行性能约束，以最大化任务执行效能为目标建立多无人机协同搜索模型。根据目标先验信息初始化环境搜索信息图（包括目标概率分布图、环境不确定度图与环境搜索状态图），利用无人机实时探测信息，基于贝叶斯准则持续更新搜索信息图。定制基于环境不确定度更新的重访机制，通过增加长时间未被重访区域的环境不确定度，引导无人机搜索该区域，降低移动目标的遗漏概率；定制基于目标函数权重更新的重访机制，引导无人机快速重访发现新的疑似目标的区域，对疑似目标进行再次确认，减少由于传感器虚警概率造成的目标误判概率。采用滚动时域规划架构，将搜索规划问题分解为一系列短时域规划问题，提升了求解效率。在典型任务想定下，通过数值仿真试验验证了所提方法的有效性。仿真结果表明，RMD-CSP能够在秒级时间内生成每个时域的搜索航迹，相比于光栅式搜索方法与标准的概率启发式搜索方法，能够引导无人机捕获更多的移动目标，同时减少误判次数，有效提升了多无人机协同搜索的任务效能。

Revisit mechanism driven multi-UAV cooperative search planning method for moving targets

To efficiently capture moving targets in unknown regions using multi-UAVs, this paper presents a Revisit Mechanism Driven Cooperative Search Planning (RMD-CSP) method to reduce the probability of missing targets and judgmental errors of the sensors. The multi-UAV cooperative search model, subject to the flight performance constraints, is established to maximize the task execution performance. The search maps (i.e., target probability maps, uncertainty maps, and environment search status maps) are initialized according to the prior information of the target, and then updated using Bayes Criterion according to the detected information by the UAVs. The revisit mechanism based on environment-uncertainty-renewal is customized to reduce the missing-target probability. This mechanism guides the UAVs to search the region that has not been revisited for a long time by enlarging the uncertainty of this region. In addition, the revisit mechanism based on objective-function-weight-renewal is customized to direct the UAVs to revisit the region where a new suspected target is found, and check the existence of the target to reduce the judgmental errors caused by the false-alarm probability of the sensors. Based on the receding horizon framework, the search planning problems are divided into a series of short-horizon planning problems to save computational costs. Simulation studies are conducted under classical mission scenarios to verify the effectiveness of the proposed method. Simulation results demonstrate that the RMD-SCP can generate search paths in seconds for each receding horizon. Compared with the scan-search algorithm and the standard probability heuristic algorithm, the RMD-CSP can guide the UAVs to capture more moving targets with fewer judgmental errors, indicating the effectiveness of the proposed method in improving the efficiency of multi-UAV cooperative search missions.