Journal of Intelligent & Robotic Systems最新文献

Multicopters are used for a wide range of applications that often involve approaching buildings or navigating enclosed spaces. Opposed to the open spaces in obstacle-free environments commonly flown by fixed-wing unmanned aerial vehicles, multicopters frequently fly close to surfaces and must take into account the airflow variations caused by airflow rebound. Such disturbances must be identified in order to design algorithms capable of compensating them. The evaluation of ground, ceiling and wall effects using two different test stands is proposed in this work. Different propellers and sensors have been considered for testing. The first test setup used was placed inside terraXcube’s large climatic chamber allowing a precise control of temperature and pressure of around 20°C and 1000 hPa, respectively. The second test setup is located at the University of Denver (DU) Unmanned Systems Research Institute (DU(^2)SRI) laboratory with a stable pressure of around 800 hPa. Two different fixed 6 degrees of freedom force-torque sensors have been used for the experiments, allowing to sample forces and moments in three orthogonal axes. The tests simulate a hovering situation of a quadcopter at different distances to either the ground, the ceiling or a wall. The influence of the propeller size, rotation speed, pressure and temperature have also been considered and used for later dimensionless coefficient comparison. A thorough analysis of the measurement uncertainty is also included based on experimental evaluations and manufacturer information. Experimental data collected in these tests can be used for the definition of a mathematical model in which the effect of the proximity to the different surfaces is evaluated.

This research focuses on designing a real-time, flexible gait planner for lower limb exoskeleton robots to assist patients with lower limb disabilities. Given the dynamic nature of gait parameters, which vary according to ground conditions and user intent, the challenge lies in developing a gait planner capable of adapting to these changes in real-time. To avoid planning complications in the cartesian space and to comply with the speed constraints of joint motors, this paper proposes planning in joint space. Furthermore, the approach also considers the maximum speed capabilities of the joint motors, aiming to generate an executable gait pattern and simultaneously enhance the robot’s walking speed by determining the minimum time required for implementation. The introduced gait planner optimizes joint trajectories for minimal angular acceleration. To provide flexibility, generalized boundary conditions suitable for different scenarios are defined. The effectiveness of the proposed planner is validated through comprehensive performance analysis, simulations, and successful implementation trials on the Exoped^® robot in various scenarios.

The collaboration among autonomous mobile robots (AMRs), including unmanned aerial vehicles (UAVs), unmanned ground vehicles (UGVs), and/or unmanned surface vehicles (USVs), significantly enhances their capabilities by enabling them to tackle more complex tasks exceeding those of individual robots. However, to fully exploit this collaboration, a common understanding of exchanged information—referred to as semantic interoperability—is crucial. Achieving semantic interoperability between these robots requires a deep understanding of relevant information and its underlying structure. To address this challenge, this paper presents a platform- and technology-independent information model developed specifically for AMRs. This model aims to facilitate collaboration by structuring information in a way that ensures semantic interoperability. The paper outlines the model's development process, beginning with a structured consolidation of information from pertinent scientific literature, resulting in a foundational framework for representing knowledge and semantics within the domain of AMRs. The practical application of the information model is demonstrated through a use case involving multiple AMRs. Additionally, the paper provides insights into the employed methodology, emphasizing the significance of systematic literature reviews and collaboration with practitioners to refine and validate the model. It also discusses theoretical and practical implications, addressing potential limitations encountered during the research.

At the RoboCup, a robotics soccer tournament, the Small Size League (SSL) is one of its leagues. The thought of a mechanism in this league that could perform unpredictable kicks and passes inspired study into both the physical mechanism required to do it and the algorithms needed to make the most of it.By introducing new ideas and utilizing a Deep Neural Network, this work contributes by improving a prior algorithm that aims to carry out a real-time inversion of the non-linear ordinary differential equation (ODE) that models the ball’s path in order to determine the parameters to hit a target with a curved kick mechanism (DNN). New techniques are also presented. The two suggested DNN achieved accuracy levels of more than 92% in the outcomes of simulation runs in MATLAB.

A control system based on the control paradigm of virtual structure is here proposed for a multi-robot system involving a quadrotor and a ground vehicle, operating in an automated warehouse. The ground robot can either provide extra power to the quadrotor, thus increasing its autonomy, or receive data from it. Therefore, the quadrotor is tethered to the ground robot through flexible cables, thus justifying the adoption of the virtual structure control paradigm, which allows controlling the two vehicles simultaneously. The control approach adopted aims at guiding the virtual vertical line joining the two robots to allow the quadrotor to produce an inventory of goods in an automated warehouse. Therefore, the two robots should visit a sequence of known positions, in front of cabinets of vertically arranged shelves. In each of them the quadrotor should read QR codes, bar-codes or RFID cards corresponding to the stored boxes, to produce the inventory. Therefore, the control objective, the focus of this paper, is to keep the shape of the virtual vertical line linking the two robots while moving. However, when an obstacle appears in the route, such as a box or other robot in the floor or another aerial robot, the formation changes its shape accordingly, to avoid the obstacle. An experiment in lab scale, mimicking a real situation, is run, whose results allow claiming that the proposed system is an effective solution for the problem of controlling a multi-robot system to produce an inventory in an automated warehouse.

Mobile robot navigation has been a very popular topic of practice among researchers since a while. With the goal of enhancing the autonomy in mobile robot navigation, numerous algorithms (traditional AI-based, swarm intelligence-based, self-learning-based) have been built and implemented independently, and also in blended manners. Nevertheless, the problem of efficient autonomous robot navigation persists in multiple degrees due to the limitation of these algorithms. The lack of knowledge on the implemented techniques and their shortcomings act as a hindrance to further development on this topic. This is why an extensive study on the previously implemented algorithms, their applicability, their weaknesses as well as their potential needs to be conducted in order to assess how to improve mobile robot navigation performance. In this review paper, a comprehensive review of mobile robot navigation algorithms has been conducted. The findings suggest that, even though the self-learning algorithms require huge amounts of training data and have the possibility of learning erroneous behavior, they possess huge potential to overcome challenges rarely addressed by the other traditional algorithms. The findings also insinuate that in the domain of machine learning-based algorithms, integration of knowledge representation with a neuro-symbolic approach has the capacity to improve the accuracy and performance of self-robot navigation training by a significant margin.

This research proposes a novel BLF-based backstepping controller for path tracking of Autonomous Vehicles (AVs) with unknown dynamics and unmeasurable states. The proposed framework includes: (1) forming geometric-dynamic model of the vehicle by combining the dynamics of the vehicle with the kinematics of the visual measurement system, (2) designing a fixed-time Extended-State Observer (ESO) to estimate the unknown dynamics and unmeasurable states, and (3) introducing a BLF-based controller for faster response and more accurate path tracking compared to previous BLF-based controllers. Besides the novelty of the BLF-based controller, by transforming the closed-loop error dynamics into a unified proportional-derivative (PD)-type structure, an intuitive criterion is proposed to provide a systematic procedure for comparing BLF-based controllers. A combined BLF is further proposed based on this performance criterion to eliminate the sensitivity of BLF-based controllers to the magnitude of the constraint. The stability analysis is performed for the fixed-time ESO and the closed-loop control system. MATLAB/CarSim co-simulation is conducted to evaluate the performance of the proposed control system. The outcomes of the work show that the closed-loop control system is exponentially stable. In addition, it can provide a faster response and result in more accurate path tracking compared to previous BLF-based control systems.

Collaborative robotics possesses the potential to revolutionise industrial automation by offering affordable and accessible solutions with reasonable skill requirements. However, identifying the most valuable and appropriate applications for this technology remains a challenge. This study conducted a comprehensive literature review to analyse the existing collaborative robotics applications, and the results showed that only a limited number of applications can be considered true collaboration, with even fewer classified as intelligent collaboration. The study comprised a survey designed to offer valuable insights to not only enhance the state-of-the-art analysis in the identification of existing challenges in the field of collaborative robotics but also to provide motivation to guide future advancements. By leveraging the survey results, researchers and practitioners will be better equipped to navigate the complex landscape of collaborative robotics and develop innovative solutions to tackle the identified challenges. This study also informs on the latest research and development in the field.

To achieve psychological inclusion and skill development orientation in human skill training, this paper proposes a haptic-guided training strategy generation method with Deep Reinforcement Learning (DRL)-based agent as the core and Zone of Proximal Development (ZPD) tuning as the auxiliary. The information of the expert and trainee is stored first with a designed database that can be accessed in real-time, which establishes the data foundation. Then, under the DRL framework, a strategy generation agent is designed, which consists of an actor-network and two Q-networks. The former network generates the agent’s decision policy, while the other two Q-networks work to approximate the state-action value function, and the parameters of all of them are administrated by the Soft Actor-Critic (SAC) algorithm. In addition, for the first time, the psychological ZPD evaluation method is integrated into the strategy generation of the DRL-based agent, which is utilized to describe the relationship between a trainees intrinsic skills and guidance. With it, the problem of transitional guidance or insufficient guidance can be handled well. Finally, simulation experiments validate the proposed method, demonstrating its efficiency in regulating the trainee under favorable training conditions.

In this paper, a novel recursive and efficient algorithm for real-time implementation of the adaptive and passivity-based controllers in model-based control of robot manipulators is proposed. Many of the previous methods on these topics involve the computation of the regressor matrix explicitly or non-recursive computations, which remains as the main challenge in practical applications. The proposed method achieves a compact and fully recursive reformulation without computing the regressor matrix or its elements. This paper is based on a comprehensive literature review of the previously proposed methods, presented in a unified mathematical framework suitable for understanding the fundamentals and making comparisons. The considered methods are implemented on several processors and their performances are compared in terms of real-time computational efficiency. Computational results show that the proposed Adaptive Newton-Euler Algorithm significantly reduces the computation time of the control law per cycle time in the implementation of adaptive control laws. In addition, using the dynamic simulation of an industrial robot with 6-DoF, trajectory tracking performances of the adaptive controllers are compared with those of non-adaptive control methods where dynamic parameters are assumed to be known.