Cognitive Robotics最新文献

Generative artificial intelligence (AI) is a form of AI that can autonomously generate new content, such as text, images, audio, and video. Generative AI provides innovative approaches for content production in the metaverse, filling gaps in the development of the metaverse. Products such as ChatGPT have the potential to enhance the search experience, reshape information generation and presentation methods, and become new entry points for online traffic. This is expected to significantly impact traditional search engine products, accelerating industry innovation and upgrading. This paper presents an overview of the technologies and prospective applications of generative AI in the breakthrough of metaverse technology and offers insights for increasing the effectiveness of generative AI in creating creative content.

With the maturity of global positioning technology and the massive popularity of mobile terminals, location-based services can provide people with convenient and efficient assistance. To use such services, mobile users need to provide location information and request query content. However, this process inevitably leads to the leakage of users’ privacy information, which poses a great threat to their property and personal safety. To address the privacy leakage in location services, this paper proposes a location privacy protection method based on ball tree (LP-BT). We first use the ball tree as a spatial index structure, and then do fuzzification on the location information of end users to obtain the maximum primary anonymous entropy, and combine the neural network learning algorithm to predict the corresponding entropy value. Finally, the final entropy is obtained based on the average entropy of the two stages. Experimental results on public dataset manifest that our model is superior to other models such as random selection model and path-based fake location generation model in terms of privacy protection level, user density and anonymization time overhead.

Optimization of energy consumption in industrial robots can reduce operating costs, improve performance and increase the lifespan of the robot during part manufacturing. Choosing energy-efficient components such as motors, drives, and controllers can significantly reduce energy consumption in industrial robots. Over-sized motors and heavy robot arms can waste energy and decrease efficiency of industrial robots. By optimizing the robot programs and reducing idle time in robot operations, the amount of spent time can be reduced to minimize energy consumption of industrial robots. By using energy-efficient motors and drives, the amount of energy consumed by the robot can be reduced. Also, regular maintenance can reduce energy consumption of industrial robots by providing maximum efficiency for the robot's components. By implementing energy management systems, energy consumption of industrial robot can be monitored and analyzed to optimize energy consumption of industrial robot during working conditions. To minimize lost energy and reuse the energy usage during working times, regenerative braking can be used in the robots. The process of part manufacturing can be optimized in order to minimize the robot's movements and energy usage during working times of industrial robots. To analyze and optimize energy consumption in working schedules of industrial robots, different methodologies from recent published papers are reviewed in the study. Proper robot selection, energy-efficient robot motor and low wight robot arms, efficient programming of working schedules, regenerative braking system, regular maintenance of robot elements and optimized process of part production regarding the minimization of energy usage are discussed to optimize the energy consumption in industrial robots. As a result, future research works in the research field can be presented in order to optimize energy consumption, reduce operational costs, and increase sustainability of industrial robot operations in terms of productivity enhancement of part manufacturing.

Rolling machinery is ubiquitous in power transmission and transformation equipment, but it suffers from severe faults during long-term running. Automatic fault diagnosis plays an important role in the production safety of power equipment. This paper proposes a novel cross-domain co-attention network (CDCAN) for fault diagnosis of rolling machinery. Multiscale features cross time and frequency domains are respectively extracted from raw vibration signal, which are then fused with a co-attention mechanism. This architecture fuses layer-wise activations to enable CDCAN to fully learn the shared representation with consistency across time and frequency domains. This characteristic helps CDCAN provide more faithful diagnoses than state-of-the-art methods. Experiments on bearing and gearbox datasets are conducted to evaluate the fault-diagnosis performance. Extensive experimental results and comprehensive analysis demonstrate the superiority of the proposed CDCAN in term of diagnosis correctness and adaptability.

This study presents the method for optimal error tracking in position control for a servo pneumatic actuated robot grasper system using a new adjustable convergence rate prescribed performance control (ACR-PPC). It focuses on improving the feedback controller and the fractional-order proportional-integral-derivative (FOPID) controller used for the position control of each robot's finger. Multiple features were considered such as tracking error, rising time, faster transient response with finite-time convergence, oscillation reduction, and pressure stabilization in the pneumatic system. Experiments were conducted using a single finger of a tri-finger pneumatic gripper (TPG) robot, actuated by a single proportional valve with a double-acting cylinder (PPVDC). Two types of input trajectories were tested: step and sine wave inputs, which are common and critical for pneumatic systems. The results show that the proposed method eliminates oscillation and achieves high tracking performance within the prescribed bounds and minimal overshoot as well. The oscillation was suppressed with minimal overshoot and fast response was achieved by tuning the formulated adjustable prescribe performance function, thus improving the rising time response without significant loss of performance.

The development of imaging technology has allowed people to move beyond the black-and-white era and into the age of color. However, preserved black-and-white historical footage remains a precious memory for people. We propose a coloring method for historical videos that combines historical image coloring methods with temporal consistency methods, thus achieving color editing for historical videos. The temporal consistency technique uses deep video priors to model the video structure and effectively ensure smoothness between frames after video color editing, even with a small amount of training data. Meanwhile, we have collected a historical video dataset named MHMD-Video, which facilitates further research on colorization of historical videos for researchers. Finally, we demonstrate the effectiveness of the proposed method through objective and subjective evaluation.

The proposed research deals with selection of particle swarm optimization (PSO) algorithm parameters for missile gliding trajectory optimization relying on Taguchi design of experiments, analysis of variance (ANOVA) and artificial neural networks (ANN). Population size, inertial weight and acceleration coefficients of PSO were chosen for the present study. The experiments have been designed as per Taguchi's design of experiments using L₂₅ orthogonal array for selection of better PSO parameters. Missile gliding trajectory is optimized by discretizing angle of attack as control parameter, consequent conversion of optimal control problem to nonlinear programming problem (NLP) and finally solving the problem using PSO with optimized parameters to obtain optimum angle of attack and realization of maximum gliding range. Simulation results portrayed that the gliding range is maximized and missile glide distance is enhanced compared to earlier experiments. The efficiency of proposed approach was verified via different test scenarios.

Some problems still exist in health management and application such as insufficient data, limited technology, and lack of professional evaluation methods by physicians with medical theory. In this study, an intelligent method is based on an analysis of physiological big data collected by wearable smartwatches. Firstly, physiological data such as pulse, heart rate, and blood oxygen were collected continuously from individuals by wearing smartwatches, and the data was digitally transmitted. Secondly, the transmitted data was sent to a health management platform by Narrow Band Internet of Things. Analyzing the data, physicians evaluated individual situations via an intelligent math model. Finally, the results were fed back to individuals through a smartphone APP to finish a medical diagnosis, disease prediction, or warning. The intelligent health management method and technology created via years of studies have been verified and will provide a new and effective strategy for health management.

The conventional application of deep reinforcement learning (DRL) to autonomous racing requires the agent to crash during training, thus limiting training to simulation environments. Further, many DRL approaches still exhibit high crash rates after training, making them infeasible for real-world use. This paper addresses the problem of safely training DRL agents for autonomous racing. Firstly, we present a Viability Theory-based supervisor that ensures the vehicle does not crash and remains within the friction limit while maintaining recursive feasibility. Secondly, we use the supervisor to ensure the vehicle does not crash during the training of DRL agents for high-speed racing. The evaluation in the open-source F1Tenth simulator demonstrates that our safety system can ensure the safety of a worst-case scenario planner on four test maps up to speeds of 6 m/s. Training agents to race with the supervisor significantly improves sample efficiency, requiring only 10,000 steps. Our learning formulation leads to learning more conservative, safer policies with slower lap times and a higher success rate, resulting in our method being feasible for physical vehicle racing. Enabling DRL agents to learn to race without ever crashing is a step towards using DRL on physical vehicles.

Road detection remains a captivating and crucial aspect of any form of autonomous driving. In this manuscript, we furnish a comprehensive appraisal of recent advancements in road lane detection, a fundamental component integral to autonomous driving. Despite numerous methodologies being proposed to augment accuracy while expediting speed, various hindrances, including lane marking variations, lighting fluctuations, and shadowy conditions, necessitate the establishment of dependable detection systems. Model-based and learning-based methods represent the two predominant techniques for lane detection. Model-based methods afford rapid computation speeds, while learning-based methods extend robustness amidst complexity. This paper delves into the techniques of lane detection and forecasts upcoming trends in the field. Collectively, this review offers a sturdy foundation for prospective research in the realm of road lane detection.