IET Cyber-Physical Systems: Theory and Applications最新文献

Weeds are a significant challenge to crop quality and quantity and therefore there is a need to adopt effective weed control and management systems. Nowadays, object detection has found extensive applications in the agricultural field such as the detection of weeds through deep learning, machine learning, image processing and IoT. The idea in this paper is to present the proposal of an autonomous rover that can identify and classify weeds in real time using the YOLO object detection method. The dataset that will be utilised in the current research is a collection of 5997 images of weed instances, allowing even more accurate detection and classification of weeds. We also combined the Convolutional Block Attention Module (CBAM) with YOLO to enable the model to pay attention to the useful spatial and channel-wise features, as an evaluation of the performance of various YOLO models is based on inference time and weed detection accuracy. Based on the experiment, YOLOv8 and its variant YOLOv8-X demonstrated the best mean average precision (mAP) of 93.6% with that inference times of 3.4 and 2.2 ms per image, respectively. YOLOv9-E (an extension of YOLOv9) using CBAM, on the other hand, had better mAP of 99.5% with inference times of 10.6 and 2.5 ms, respectively. These findings indicate that YOLOv8 and YOLOv9 hold a good prospective of automatic field-level weed detection and emphasise the significance of high-quality datasets, efficient model architectures and attention mechanisms to the efficient and correct autonomous weed management.

This paper proposes a new framework for the analysis of cyber-physical system security against denial-of-service (DoS) attacks using generalised stochastic Petri nets. Although cyber-physical systems, through increased integration of computational and physical processes, offer great advantages, they are subject to cyber threats that can disrupt their critical operations. Among them, DoS attacks, which overload communication channels and prohibit the exchange of data between system components, are a major concern. Traditional methods of security assessment are inadequate given the unique complexities of cyber-physical system architectures. This research presents a generalised stochastic Petri net-based model able to capture the dynamics of a cyber-physical system under attack scenarios for the comprehensive analysis of system vulnerabilities and defencive mechanisms. The model incorporates immediate and timed transitions, thus mapping both continuous operations of the cyber-physical system and the discrete-event nature of cyber threats. Simulation experiments validate the effectiveness of the model in demonstrating how DoS attacks can degrade system performance. The results reflect the need for improved methodologies for security testing in order to enhance the resilience of cyber-physical systems, particularly in safety-critical applications.

Behaviours and activities are natural concepts (found, e.g., in UML and SysML) for model-driven design of cyber-physical systems (CPS). These concepts are formalised in the activity framework, a model-based framework incorporating a model of activities with determinate timing and behaviour, and a strong mathematical foundation based on max-plus algebra that allows efficient timing analysis and optimisation. It provides a layered view of the system's actions and events, activities built from them, and sequences of activities that capture the overall behaviour of the system. Implementations of supervisory control for CPS to govern the system behaviour are often made by hand. Preserving the specified behaviour and the model-predicted timing in an implementation is challenging, due to the need to simultaneously handle action timing, synchronisation, concurrency, pipelining and plant feedback. We introduce an execution architecture and engine to automatically synthesise an implementation of a supervisory controller directly from a model specification. The execution engine is guaranteed to execute a specification in a time- and behaviour-preserving fashion, even in the presence of action timing variations and including event feedback in a physical execution. We prove that the architecture and engine preserve the specified ordering of actions and events of the model as well as the timing thereof, up to a known bound. We validate our approach on a prototype production system.

The security of industrial control systems (ICSs) is crucial due to their integral role in critical national infrastructure. This study tackles the escalating challenges posed by sophisticated cyberattacks, especially those that are unknown and evade existing detection mechanisms. Despite extensive research, there is a notable gap in systematically comparing supervised and unsupervised learning models for anomaly detection, leading to inconsistent evaluations of their effectiveness. To bridge this gap, we developed a comprehensive anomaly detection framework to systematically evaluate these models, focusing on their capability to detect unknown attacks. Utilising operational data from the Secure Water Treatment (SWaT) testbed, we assessed six unsupervised and five supervised learning methods. Our findings reveal significant performance disparities: supervised models excel in precision but have higher undetected rates, whereas unsupervised models achieve superior recall at the expense of increased false alarm rates. This study provides critical insights into the strengths and limitations of both approaches, guiding the development of more robust ICS security solutions.

Smart grid systems, as modern cyber-physical systems (CPS), introduce new interdependencies between power and communication components that can create new security challenges. One potential challenge that may arise is cascading failures resulting from cyber-attacks or the failure of a component that needs to be detected in a timely manner. In this paper, we propose a novel early-stage failure prediction (ESFP) mechanism that applies machine learning (ML) algorithms to enhance the security of smart grid systems. We use a realistic model to generate a dataset for training ML algorithms and develop a mechanism to predict the state of a system's components in the early stages before failures propagate in the system. ESFP can predict the final state of each power system component with respect to its initial failures. We apply the extreme gradient boosting (XGBoost) algorithm and examine the features of both the communication and power networks that provide high accuracy in predicting failures. We develop a new data generation procedure to construct a dataset containing electrical and network features and characteristics for training ML algorithms. ESFP also identifies the location of the initial failures as this allows for further protection plans and decisions. We evaluate the effectiveness of the proposed mechanism through an analysis conducted on an IEEE 118-bus system. The proposed mechanism achieves 99.4% prediction accuracy in random attacks using the XGBoost algorithm. We also improve the time of the XGBoost algorithm by 75% by combining an unsupervised ML algorithm with this algorithm.

Cyber-physical systems (CPS) seamlessly integrate computers, networks and physical devices, enabling machines to communicate, process data and respond to real-world conditions in real time. By bridging the digital and physical worlds, CPS ensures operations that are efficient, safe, innovative and controllable. As smart cities and autonomous machines become more prevalent, understanding CPS is crucial for driving future progress. Recent advancements in edge computing, AI-driven vision and collaborative systems have significantly enhanced CPS capabilities. Synchronisation, optimisation and adaptation are intricate processes that impact CPS performance across different domains. Therefore, identifying emerging trends and uncovering research gaps is essential to highlight areas that require further investigation and improvement. This systematic review and analysis aims to offer a unique point to researchers and facilitates this process by allowing researchers to benchmark and compare various techniques, evaluate their effectiveness and establish best practices. It provides evidence-based insights into optimal strategies for implementation while addressing potential trade-offs in performance, resource usage and reliability. Additionally, such reviews help identify widely accepted standards and frameworks, contributing to the development of standardised approaches.

Task scheduling in real-time multi-core systems is crucial for meeting stringent timing requirements, especially as embedded systems become increasingly prevalent. Battery-powered systems, in particular, require energy-efficient schedulers to regulate processor speed and temperature. Current software-based schedulers face limitations, such as overhead, latency and inefficiencies, making it challenging to balance performance with energy and thermal management. This highlights the need for more dynamic schedulers in heterogeneous multi-core processor environments, particularly for independent task scheduling, which must address these challenges while optimising both energy consumption and temperature control. This paper presents an online distributed hardware scheduler designed specifically for independent tasks, utilising the earliest deadline first (EDF) algorithm to manage hard real-time tasks on multi-core embedded systems. Implementing this scheduler in hardware reduces clock cycles, improves efficiency and lowers latency compared to software-based solutions. It dynamically adapts to real-time tasks without compromising predictability or introducing significant overhead. The scheduler optimises energy consumption and temperature management, addressing both dynamic and static energy demands. Results show a reduction of $��51.16�\%�$ in dynamic energy, $��50.82�\%�$ in static energy and $��20.21�\%�$ in temperature compared to the high-performance real-time hardware scheduler (HRHS), while maintaining overall system performance and efficiency.

In the processing of power grid resource data, incomplete data from service centres lead to data loss, incompleteness, or inaccuracy, which seriously affects the training of neural networks. In terms of intelligent diagnosis of power grid faults, the misjudgement rate of fault diagnosis has significantly increased due to data problems. In some areas, the time for fault analysis has been greatly extended, and maintenance has been delayed, resulting in prolonged power outages. For example, in Guang'an, Sichuan, citizens are charged 192 yuan in electricity bills for one night, causing huge economic losses to residents' lives and industrial production. In this regard, this article chooses a bad data processing scheme based on cosine similarity and a data normalisation scheme based on standardisation to preprocess the full path names in the data model structure of power grid resource business. Then, a linear neural network suitable for the full path name characteristics of power grid equipment is selected, and a neural network framework is set up. The sample data are used to train the neural network to realise the full path name conversion of the remaining large number of power grid equipment. The experiment shows that the optimisation effect of the model established in this article is affected by multiple parameters. When the comprehensive ratio is controlled at 0.7 and the number of neighbours k is 30, the model achieves the best conversion efficiency, and its recommended accuracy can reach 75%. Subsequently, the optimisation model that achieves the best efficiency is compared with traditional collaborative filtering recommendation algorithms or translation models. Compared with the TransH model, the proposed model has added 1.23% and 10% more new data at 45 and 40 min of operation, respectively. The result proves that the neural network model established in this paper can better adapt to power grid data conversion work, ensuring the automation and efficiency of power grid data transmission.

The ability to reduce emissions and improve sustainability in agricultural consumer electronics has been significantly hindered due to the use of energy-intensive technology within the agricultural sector. This study proposes a new enhancement of deep Q-learning (DQN) with principal component analysis (PCA) focused on energy efficiency. PCA helps manage massive operational data by performing dimensionality reduction, whereas DQN, a reinforcement learning paradigm, optimises decision-making during real-world interactions. The main contribution of this study is in the combined use of PCA and DQN to form customisable, precise, contest-responsive energy frameworks powered by real-time analytics on agricultural data—energy management on such a scale has not been approached in the context of sustainable agriculture before. The experiments confirm the optimal model, further achieving a cumulative reward of 72.56, an average emission of 1.83, a Q-value of 24.76 and a total zenith value of 75.40% in ensuring numerous noncriteria-defined efficient energy-dependent operations. This paradigm not only fills the void in the automation of passive intelligent agricultural systems but also serves as a point of reference for other eco-critical domains to strive towards greener technology.