For vehicle navigation, an inertial navigation system (INS) is often used to assist the global positioning system (GPS); but the positioning accuracy is well known to be sensitive to boththe GPS sampling rate and ine...
详细信息
ISBN:
(纸本)9781839531637
For vehicle navigation, an inertial navigation system (INS) is often used to assist the global positioning system (GPS); but the positioning accuracy is well known to be sensitive to boththe GPS sampling rate and inertial sensor error aggregation. As such, many practical navigation systems adopt both highend INS and high GPS sampling rate, which costs more for hardware on the one hand and demands excessive energy consumption and communications on the other hand. In this work, we propose a learning-based approach that is capable of providing the same level of positioning accuracy but using merely cheap, low-end mobile inertial sensors and integrated GPS with very low sampling rate. the proposed approach consists of a Kalman filter and two multilayer neural networks to realize 3-dimensional, high-precision vehicle navigation. the key to the impressive outcome is the combination of the classic statespace Kalman filter with a new data-driven machine learning mechanism. the proposed approach is evaluated with real trajectory data collected on campus of CUHKSZ, and the results show that it can achieve a positioning error around 8m when the GPS sampling interval is as large as 30s (30 times of nowadays regular sampling interval). Apart from this, the proposed approach also demonstrates good adaptiveness to changing trajectories.
Automatic program repair plays a crucial role in the software development and implementation. While deep learning-based approaches have made significant progress, one inherent challenge is the inefficiency in code rep...
Automatic program repair plays a crucial role in the software development and implementation. While deep learning-based approaches have made significant progress, one inherent challenge is the inefficiency in code representation, which hampers accurate patch generation. Furthermore, the training data used by these data-driven approaches may be limited, and they may not be able to capture the subtle differences between vulnerabilities and patches. To address these issues, FixGPT, we propose a three-tier deep learning model in the study. Specifically, a generative pre-trained transformer model is designed in the first tier to capture code characteristics and programming patterns. the second tier integrates a generation model based on the structure of neural machine translation, for the purpose of generating potential patches. Finally, a contrast model is introduced in the last tier to differentiate between the vulnerability and the patch. We also incorporate the Byte Pair Encoding approach to reduce the search space by converting identifiers into subwords. Detailed experimental studies have been carried out to evaluate the performance of FixGPT on two well-known benchmarking datasets: QuixBugs and Defects4J. the results demonstrated significant improvements in the effectiveness and accuracy in comparison with existing solutions. We complement these findings through the analysis of two case studies.
Power distribution networks are rapidly evolving into large-scale, highly complex, and topologically variable active distribution networks (ADNs). In such ADNs, voltage issues are serious and thus the distributed or d...
详细信息
ISBN:
(数字)9798350377408
ISBN:
(纸本)9798350377415
Power distribution networks are rapidly evolving into large-scale, highly complex, and topologically variable active distribution networks (ADNs). In such ADNs, voltage issues are serious and thus the distributed or decentralized voltage control tends to be called for, whose effectiveness is critically dependent on the proper partitioning of the network into multiple sub-networks, a.k.a., voltage area partition (VAP). Existing partitioning methods are focused on three-phase balanced systems with known line parameters, making them less applicable to real-world distribution networks. To address this limitation, this paper proposes a data-driven VAP method, which can perceive voltage sensitivity from the historical operation data without need of precise distribution line parameters while simultaneously assigning buses into clusters by designing a physics-informed graph neural network architecture. In particular, a phase-completing layer is developed to convert a phase-missing bus to a three-phase bus, which can fairly evaluate the voltage cohesiveness of different buses. Simulation in the modified ieee 123 bus system demonstrates that the proposed method yields partitioning results with higher voltage cohesiveness compared to benchmark methods.
this work is devoted to the development of personalized training systems. A major problem in learning environmens is applying the same approach to all students: teaching materials, time for their mastering, and a trai...
详细信息
ISBN:
(数字)9781728118147
ISBN:
(纸本)9781728118154
this work is devoted to the development of personalized training systems. A major problem in learning environmens is applying the same approach to all students: teaching materials, time for their mastering, and a training program that is designed in the same way for everyone. Although, each student is individual has his own skills, ability to assimilate the material, his preferences and other. Recently, recommendation systems, of which the system of personalized learning is a part, have become widespread in the learning environment. On the one hand, this shift is due to mathematical approaches, such as machine learning and data mining, that are used in such systems while, on the other hand, the requirements of technological standards "validated" by the World Wide Web Consortium (W3C). According to this symbiosis of mathematical methods and advanced technologies, it is possible to implement a system that has several advantages: identifying current skill levels, building individual learning trajectories, tracking progress, and recommending relevant learning material. the conducted research demonstrates how to make learning environment more adaptive to the users according to their knowledge base, behavior, preferences, and abilities. In this research, a model of a learning ecosystem based on the knowledge and skills annotations is presented. this model is a general model of the all life learning. Second, this thesis focuses on the creation of tools for personalized assessment, recommendation, and advising.
the gradual development of China's power grid has formed the current pattern of cross regional interconnected power grid with ultra-high/ultra-high voltage AC/DC hybrid connection, which has effectively promoted t...
详细信息
ISBN:
(数字)9798331523527
ISBN:
(纸本)9798331523534
the gradual development of China's power grid has formed the current pattern of cross regional interconnected power grid with ultra-high/ultra-high voltage AC/DC hybrid connection, which has effectively promoted the optimized allocation and utilization of primary energy nationwide. However, withthe rapid growth of model dimensions in the power system, traditional model driven power grid stability analysis and control decision-making techniques are difficult to achieve real-time perception of power grid safety and stability, and are prone to missing opportunities for power grid regulation. therefore, there is an urgent need for a more effective power system transient stability assessment technology. this article proposes a transient stability assessment method for power systems based on Bayesian deep learning. Using Bayesian deep neural networks to learn and extract the data distribution of transient stability boundaries; Based on variational inference method, Bayesian deep neural network parameters are trained on big data to obtain an artificial intelligence based stability evaluation model for power grid faults. this model can more accurately predict the transient stability of power grid operation mode and objectively measure the reliability of artificial intelligence model prediction results, overcoming the shortcomings of traditional deterministic neural networks in easily misjudging stability in system critical states.
the increasing penetration of renewable energy sources in the electricity grid underscores the critical need for advanced energy management systems in residential settings. Traditional energy management systems are li...
详细信息
ISBN:
(数字)9798350377408
ISBN:
(纸本)9798350377415
the increasing penetration of renewable energy sources in the electricity grid underscores the critical need for advanced energy management systems in residential settings. Traditional energy management systems are limited in handling the complexity and dynamism of energy systems, thus hindering efficient energy utilization in buildings. this study conducts a novel evaluation of Model Predictive control versus traditional rule-based systems within a lab-based energy management system prototype for households. the experimental setup consisted of a bidirectional power supply, lithium-ion batteries, and a programmable logic controller. All systems were rigorously tested over consecutive seven-day periods, taking into account photovoltaic supply, electricity demand, and pricing to optimize energy management and reduce overall electricity costs. the results demonstrate the superior efficiency of the advanced control strategy in minimizing electricity costs compared to traditional controllers, which struggle withthe complexity of sector coupling and dynamic energy environments. this research offers new insights into the practical viability and performance of Model Predictive control in real-world settings, showcasing its potential for advanced energy management and emphasizing the importance of open standards in interface development for future advancements. Recognizing the necessity for comprehensive solutions, model- and data-driven approaches can integrate households as active participants in the dynamic energy system, enabling real-time adaptation to fluctuating grid conditions. By effectively communicating the advantages of advanced control technologies to policymakers, energy providers, and homeowners, we can encourage broader adoption and unlock their potential for creating a more sustainable, efficient, and cost-effective energy future.
暂无评论