Artificial Intelligence and Autonomous Systems
ISSN: 2959-0744 (Print)
ISSN: 2959-0752 (Online)
CODEN: AIASBB
About This Journal
Enhanced safety and efficiency in traction elevators: a real-time monitoring system with anomaly detection
Received
15 Nov 2024
Accepted
21 Jan 2025
Published
13 Feb 2025
This study presents the design and implementation of a real-time monitoring system for traction elevators, leveraging piezoelectric sensors for vibration measurement and speed sensors for velocity data acquisition. The system is powered by a LattePanda dashboard equipped with an integrated Real-Time Clock (RTC), ensuring precise data collection and timestamping. Vibration data is captured through piezoelectric sensors, while velocity data from speed sensors is used to calculate acceleration. The collected data is stored locally and can also be transmitted remotely. Aimed at improving elevator safety and efficiency, the system detects potential issues such as misalignments and mechanical wear. Given the increasing number of elevator accidents, this study focuses on enhancing monitoring capabilities using advanced technologies. Data from an electric elevator was analyzed with three anomaly detection algorithms: Isolation Forest, Support Vector Machine (SVM), and Z-score. The results revealed that Isolation Forest identified 15 anomalies (1.06% of the data), SVM detected 25 anomalies (1.77% of the data), and Z-score identified 86 anomalies (6.08% of the data). This research not only enhances elevator condition monitoring but also lays the groundwork for future digital twin systems in passenger elevator applications.
Deep reinforcement learning Lane-Change Decision-Making for autonomous vehicles based on motion primitives library in hierarchical action space
Received
20 Jun 2024
Accepted
10 Dec 2024
Published
24 Dec 2024
Deep Reinforcement Learning (DRL) is capable of learning a policy with great scene adaptation ability through interactions with the environment, and has application potential in the field of autonomous driving. However, using DRL to directly control the vehicle motion command is easy to lead to fluctuation and non-smoothness. Discrete DRL decision-making method can generate stable behavior but lose some maneuverability. Thus, the trade-off between flexibility and stability to enhance the performance of DRL policy is an important issue. In this paper, a Deep Reinforcement Learning Decision-Making algorithm based on Motion Primitives Library (MPL) in hierarchical action space is proposed to provide flexible and reliable maneuvers for autonomous driving. The upper action space contains discrete lane-changing targets, and the lower action space is mapped as a motion primitives library. In addition, model predictive control method (MPC) based on the vehicle kinematics model is used to optimize the motion primitives instantly. The performance of the proposed method is evaluated through highway simulation. The results show that the method can make the autonomous driving lane changing process safer, more efficient and more comfortable.
Dynamic lane-changing trajectory planning for autonomous vehicles in mixed traffic
Received
14 May 2024
Accepted
30 Oct 2024
Published
05 Dec 2024
This paper presents a novel dynamic lane-changing trajectory planning (DLCTP) model for autonomous vehicle (AV) running in the mixed traffic environment. The proposed model fully considers the dynamics of surrounding human-driven vehicles and can work on both straight and curved roads. The first step of the DLCTP model is to decide when and where to make the lane change based on the car-following model and safety constraints. Upon decision-making, an optimal lane-changing trajectory that accounts for safety, comfort, and efficiency is generated at each time step until the lane-changing procedure is completed. CarSim-Simulink based simulation platform and three typical traffic scenarios are applied to validate the proposed DLCTP model. Experimental results show that the proposed DLCTP model can generate smooth, safe, and comfort trajectories even in complex traffic situations. The proposed DLCTP model can be employed directly on real AVs because it is easy to implement and can adapt to complex traffic environments.
Developing a Machine Intelligence Quotient (MIQ) for evaluating autonomous vehicle intelligence: a conceptual framework
Received
22 Aug 2024
Accepted
21 Oct 2024
Published
26 Oct 2024
This paper presents a methodology to quantify the Machine Intelligence Quotient (MIQ) for autonomous cars. MIQ integrates multi-dimensional categories—Physical, Cognitive, and Functionality Intelligence attributes—to evaluate vehicle intelligence in a comprehensive manner. By focusing on the harmony of these facets with human cognitive and decision-making processes, MIQ provides a transformative approach to understanding and enhancing autonomous vehicle technology. This framework not only offers an empirical method for intelligence assessment but also sets a visionary benchmark, advocating for advancements that parallel human-like intelligence in future autonomous systems.
Cyberattack detection on SWaT plant industrial control systems using machine learning
Received
07 May 2024
Accepted
13 Sep 2024
Published
23 Sep 2024
Detecting cyberattacks is critical for maintaining the security and integrity of industrial control systems (ICSs). This study introduces a machine learning approach for identifying cyberattacks on the Secure Water Treatment (SWaT) plant testbed. The dataset, sourced from the Singapore University of Technology and Design, includes data from 51 sensors and actuators. The research employs a Long Short-Term Memory (LSTM) network alongside traditional machine learning algorithms like Random Forest (R.F.), Support Vector Machine (SVM), and K-Nearest Neighbour (KNN) to classify cyberattacks. The LSTM model outperformed the other methods, achieving a test accuracy of 98.02% (cyberattack: 97.80%, non-attack: 98.30%). Given the imbalanced nature of the dataset, additional metrics such as precision, recall, and F1 score were evaluated, further confirming the LSTM model’s robustness compared to traditional algorithms. This research demonstrates the LSTM network’s effectiveness in enhancing cybersecurity for ICSs and underscores the need for proactive strategies in detecting and mitigating cyber threats.
Cooperative adaptive fault-tolerant control for heterogeneous multiagent systems with guaranteed performance
Received
15 Apr 2024
Accepted
16 Jul 2024
Published
06 Aug 2024
This paper studies the fault-tolerant control problem for the heterogeneous multiagen systems consisting of multiple quadrotors and mobile robots with guaranteed performance in the presence of unknown actuator faults. First, the full-state performance constraints of the position and attitude subsystem of follower vehicles are considered, especially in the case of actuator faults, and then the state constraints of heterogeneous unmanned systems are addressed by combining the performance functions and barrier Lyapunov function method. Then, the constraints-based cooperative adaptive fault-tolerant control strategy is proposed, where the adaptive terms are adopted to compensate for the unknown bounded actuator loss of effectiveness faults and bias faults and the constraint signals are introduced to ensure the performance conditions of system states. Based on the theoretical analysis, the cooperative fault-tolerant time-varying formation convergence performance is discussed. The simulation results on the UAVs-UGVs formation systems composed of quadrotors and mobile robots are presented to validate the effectiveness of the proposed control strategy.
Planning, Control, and Learning in Mobile Robots
Special Issue Editor
Aerial Robotics
Special Issue Editor
Intelligent Transportation Systems
Special Issue Editor