UAV swarm applications are critical for the future, and their mission-planning and decision-making capabilities have a direct impact on their performance. However, creating a dynamic and scalable assignment algorithm that can be applied to various groups and tasks is a significant challenge. To address this issue, we propose the Extensible Multi-Agent Deep Deterministic Policy Gradient (Ex-MADDPG) algorithm, which builds on the MADDPG framework. The Ex-MADDPG algorithm improves the robustness and scalability of the assignment algorithm by incorporating local communication, mean simulation observation, a synchronous parameter-training mechanism, and a scalable multiple-decision mechanism. Our approach has been validated for effectiveness and scalability through both simulation experiments in the Multi-Agent Particle Environment (MPE) and a real-world experiment. Overall, our results demonstrate that the Ex-MADDPG algorithm is effective in handling various groups and tasks and can scale well as the swarm size increases. Therefore, our algorithm holds great promise for mission planning and decision-making in UAV swarm applications.
8 August 2022 in IEEE Transactions on Instrumentation and Measurement
This article describes the technology and realization of an inertial magneto-inductive positioning system (MPS) with the improved GWO-PF algorithm. The system is implemented with a dual-axis magnetic beacon (MB), a three-axis magnetic sensor, and an inertial measurement unit (IMU). Unfortunately, the performance of the magnetic-based PSs is severely impaired by the attitude errors of the magnetic sensor that is directly obtained from IMU. In this article, a positioning method of inertial magneto-inductive is presented to solve the above problem, which is not affected by the attitude errors of the sensors. Furthermore, a particle filter (PF) based on the improved grey wolf optimizer (GWO-PF) algorithm is developed to improve the positioning performance proposed of the moving target. The realized prototype exhibits a maximum positioning error lower than 0.15 m for the static target in an indoor environment with a medium area of 8.4 m $\times $ 6.5 m. The performance of tracking moving target is verified by simulation and the cumulative probability distribution (CPD) indicates that 99% of positioning errors are lower than 0.83 m.
A high accuracy in positioning and navigation is difficult in indoor and underground conditions due to obstacles. To deal with the issues of inaccurate positioning in indoor and underground environments, a novel Magnetic Based Positioning System (MBPS) based on fingerprint and confidence evaluation is presented in this paper. The proposed method promises good penetrability and negligible interference from multipath error, along with solving the sparse fingerprint problem constrained by the a priori condition. Furthermore, to address the problem of variable signal quality, an optimal information selection method based on confidence evaluation is proposed. This method suggests evaluating and selecting the optimal information source to improve the robustness and accuracy of the MBPS. The approach is supported by experiments, where magnetic beacons are detected behind a concrete wall of 0.3 m thickness. According to the experimental results, the positioning accuracy can reach up to 0.040 m. This work provides an impressive positioning solution for indoor, underground, and other scenarios where high-frequency wireless signals cannot be used due to the obstacles.
According to the processing needs of pulse power supply for the micro hole machining, a multi-mode pulse power supply including RC mode, controlled resistance mode, controlled capacitor mode, and other several power supply modes is developed for the micro- EDM system, whose CPU circuit is based on a microprocessor and two complex programmable logic devices (CPLD). In micro- EDM, the micro-energy pulse power not only realizes nanosecond grade electrical pulse signal, but also can achieve a wide range of optional in energy with the cont rol of power transistors and electrical relays which can choose the electrical processing mode, open circuit voltage, resistance, capacitance, pulse-on time, pulse-off time and so on. Finally, 2×128 array holes with deviation of 2μm in diameter can be stably realized with the multi-mode pulse power supply
26 October 2023 in IEEE Transactions on Industrial Electronics
This article introduces visual-depth-inertial-wheel odometry (VDIWO), a robust approach for real-time localization of mobile robots in indoor and outdoor scenarios. Notably, VDIWO achieves accurate localization without relying on prior information. This approach integrates the RGB-D camera, inertial measurement unit, and odometer measurements in a tightly coupled optimization framework. First, we introduce the depth measurement model based on Gaussian mixed model to predict the depth uncertainty of feature points. Then, we propose a hybrid depth estimation method that utilizes both depth measurement fusion and multiview triangulation to estimate the depth of landmarks and simultaneously identify high-quality landmarks. Furthermore, we integrate visual reprojection with depth measurement constraints and odometer preintegration constraints into the tightly coupled optimization framework to further enhance pose estimation accuracy. We evaluate the performance of the VDIWO method using OpenLORIS datasets and real-world experiments. The results demonstrate the high accuracy and robustness of VDIWO for state estimation of mobile robots.
29 May 2024 in IEEE Transactions on Instrumentation and Measurement
This study presents a novel system for accurately positioning underground pipelines by applying rotating permanent magnets, addressing the limitations of traditional wireless systems in effectively penetrating both the pipe and surrounding soil. The system is realized with the improved fingerprint matching algorithm, which is composed of a rotating structure, a permanent magnet, a magnetic sensor, an A/D module, and a positioning solution module. Under the boundary conditions, a continuous propagation model of the magnetic field is analyzed for the medium of different magnetic permeability, which is the basis of the positioning method in the pipelines. To solve the problem of difficulty in obtaining an accurate dense database, a methodology involving measurement and virtual reference points is developed. A matching method based on cluster analysis is applied in the system to improve the matching efficiency, which includes cluster selection, coarse matching, and fine positioning. The realized prototype is used to verify the proposed approach in long metal pipelines, which have a maximum positioning error of less than 0.15m, a mean value of 0.0974m, and a standard deviation of 0.0106m in a factory environment. This affirms the effectiveness of the developed system in addressing issues related to accurate pipeline localization.