Sensor degradation and failure often undermine users’ confidence in adopting a new data-driven decision-making model, especially in risk-sensitive scenarios. A risk assessment framework tailored to classification algorithms is introduced to evaluate the decision-making risks arising from sensor degradation and failures in such scenarios. The framework encompasses various steps, including on-site fault-free data collection, sensor failure data collection, fault data generation, simulated data-driven decision-making, risk identification, quantitative risk assessment, and risk prediction. Leveraging this risk assessment framework, users can evaluate the potential risks of decision errors under the current data collection status. Before model adoption, ranking risk sensitivity to sensor data provides a basis for optimizing data collection. During the use of decision algorithms, considering the expected lifespan of sensors enables the prediction of potential risks the system might face, offering comprehensive information for sensor maintenance. This method has been validated through a case study involving an access control.
This paper focuses on enhancing the resilience of microgrids—localized power systems that integrate multiple energy sources—against challenges such as natural disasters, technological obstacles, and human errors. It begins by defining the specific connotation of microgrid resilience and then proposes an innovative solution centered on the use of advanced sensor technology to continuously monitor the microgrid and its operational environment, ensuring accurate and timely data collection under dynamic conditions. Subsequently, a decision risk assessment framework is constructed, integrating data quality evaluation and operational risk considerations, to drive strategy optimization through in-depth data analysis. At the application level, this framework is successfully applied to two critical decision-making scenarios: the first is to optimize the power allocation strategy between solar energy and the auxiliary grid, aiming to maximize cost efficiency and minimize power outage losses; the second is to develop low-risk maintenance plans based on the predicted failure probabilities of microgrid components with uncertain information. Both decision processes skillfully utilize Monte Carlo simulation and multi-objective genetic algorithms to effectively manage the uncertainty risks in the decision-making process, thereby significantly enhancing the overall resilience of the microgrid.
The ability and rapid access to execution data and information in manufacturing workshops have been greatly improved with the wide spread of the Internet of Things and artificial intelligence technologies, enabling real-time unmanned integrated control of facilities and production. However, the widespread issue of data quality in the field raises concerns among users about the robustness of automatic decision-making models before their application. This paper addresses three main challenges relative to field data quality issues during automated real-time decision-making: parameter identification under measurement uncertainty, sensor accuracy selection, and sensor fault-tolerant control. To address these problems, this paper proposes a risk assessment framework in the case of continuous production workshops. The framework aims to determine a method for systematically assessing data quality issues in specific scenarios. It specifies the preparation requirements, as well as assumptions such as the preparation of datasets on typical working conditions, and the risk assessment model. Within the framework, the data quality issues in real-time decision-making are transformed into data deviation problems. By employing the Monte Carlo simulation method to measure the impact of these issues on the decision risk, a direct link between sensor quality and risks is established. This framework defines specific steps to address the three challenges. A case study in the steel industry confirms the effectiveness of the framework. This proposed method offers a new approach to assessing safety and reducing the risk of real-time unmanned automatic decision-making in industrial settings.
Nowadays it is commonly accepted that knowledge, as resource has become increasingly important in the drive towards knowledge-driven innovation in the manufacturing industry, and which is especially critical for the R&D of complex equipment. In China, there are several fields of manufacturing, such as aviation, ship, electronics etc. Although each of them has its own professional technology, they also take advantage of the basic common manufacturing technology. Thus, it is meaningful for sharing the manufacturing technology knowledge. In order to promote the innovation, a manufacturing technology knowledge retrieval system is developed to make more efficient use of manufacturing technology knowledge. Manufacturing technology knowledge is a phrase with vast meanings, which may include knowledge of different manufacturing process means, machine and process capabilities or the latest technology developments. In this system, the manufacturing technology knowledge is defined and organized. A knowledge retrieval method based on weighted fusion of OSE (ontology-based semantic extension) and UCF (user-based collaborative filtering) is proposed and described detailed. On this basis, the prototype system of manufacturing technology knowledge retrieval system is developed. Its architecture and main functions are elaborated. Results of this system facilitate the sharing of manufacturing knowledge, and it is expected to accelerate the process of knowledge transfer, to improve the level of the overall manufacturing technology.
1 January 2025 in Journal of Mechanical Engineering
Recherche sur l'amélioration de la qualité des données et l'évaluation des risques d'aide à la décision des processus de production: L'industrie de production est essentielle pour le développement durable, la force industrielle soutenant la compétitivité nationale. Le gouvernement chinois a accru ses investissements dans la production industrielle afin d'améliorer la compétitivité manufacturière, renforcer la sécurité et réduire l'exposition des travailleurs aux dangers. Cependant, la transformation sans personnel dans la production dangereuse fait face à des défis liés à la qualité des données et à l'évaluation des risques en temps réel, entravant les opérations stables à faible risque. Soutenue par des programmes nationaux, cette étude aborde ces problèmes en se concentrant sur l'amélioration de la qualité des données et l'évaluation des risques complexes en temps réel. Les domaines de recherche clés incluent :Analyse du modèle PAE : L'étude utilise le modèle PAE pour analyser le rôle humain dans la production, identifiant la prise de décision sans personnel comme centrale à la transformation. Elle met en évidence la qualité des données et l'évaluation des risques en temps réel comme des défis critiques, proposant des solutions telles que l'amélioration de la qualité des données, l'évaluation des risques de décision et l'optimisation. Une approche basée sur le jumeau numérique est introduite pour l'évaluation des risques en temps réel. Imputation des données : Un algorithme d'imputation de données multi-canaux utilisant 1D-CNN WGAN est conçu pour gérer les données manquantes, exploitant l'attention multi-têtes et WGAN pour une imputation efficace. Un mécanisme d'entraînement hybride améliore la précision du modèle de décision dans les scénarios à taux de données manquantes élevé, validé par la classification des défauts industriels. Évaluation des risques de décision : Un cadre basé sur la simulation de Monte Carlo évalue les risques de décision dans des scénarios discrets et continus, quantifiant les risques des algorithmes et guidant la conception des systèmes d'acquisition de données. Une méthode de contrôle tolérant aux défauts des capteurs basée sur les risques minimise les risques potentiels dans les scénarios de décision continue. Amélioration de la résilience : Une méthode de prédiction d'état et d'amélioration de la résilience basée sur l'IoT est proposée, optimisant l'allocation d'énergie et réduisant les risques d'arrêt. Une fonction objectif de pénalité temporelle des risques améliore la récupération du système dans les scénarios à multiples tâches de maintenance. Évaluation des risques en temps réel : Une méthode basée sur le jumeau numérique redéfinit le risque sous des perspectives spatio-temporelles, utilisant la technologie multi-agent et les modèles de jumeaux numériques pour l'évaluation des risques en temps réel. Son efficacité est validée dans une unité de production de nettoyage de projectiles. Cette recherche fournit des bases théoriques et méthodologiques pour la transformation sans personnel dans la production dangereuse, abordant les défis de la qualité des données et de l'évaluation des risques pour assurer des opérations stables à faible risque.