Volume 13
Published on November 2024This paper explores the numerical study of quantum many-body systems with an emphasis on exact diagonalization techniques. The complexity of strongly correlated systems, often governed by large Hilbert spaces, presents significant computational challenges, making exact solutions difficult. In this work, we examine methods to simplify these systems by leveraging techniques such as the Schrieffer-Wolff transformation, which decouples high-energy and low-energy states, and the use of symmetry operators to block-diagonalize Hamiltonians and so on. These approaches are demonstrated with examples such as the hydrogen atom and a lambda system. The second part of the paper focuses on specific case studies, including a one-dimensional spin model and Bose-Hubbard model. The latter is crucial for understanding the behavior of interacting bosons in lattice systems and phenomena such as the superfluid-Mott insulator transition. We present a detailed investigation of the phase diagram for the one-dimensional Bose-Hubbard model using both exact diagonalization and mean field theory, providing insights into its quantum phase transitions. This study underscores the potential of exact diagonalization in quantum simulations and highlights its relevance for experimental setups involving trapped ions and superconducting qubits.
View pdf
This paper addresses the challenge of sample efficiency in reinforcement learning (RL) for autonomous driving, a domain characterized by long-term dependencies and complex environments. While RL has shown success in various fields, its application to autonomous driving is hindered by the need for numerous samples to learn effective policies. We propose a novel, lightweight reward-shaping method called room-of-adjust to maximize learning progress. This approach separates rewards into continuous tendency rewards for long-term guidance and discrete milestone rewards for short-term exploration. Our method is designed to be easily integrated with other approaches, such as efficient representation, imitation learning, and transfer learning. We evaluate our approach on a hill-climbing task with uneven surfaces, which simulates the spatial-temporal reasoning required in autonomous driving. Results show that our room-of-adjust reward shaping achieves near-human performance (81.93%), whereas other reward shaping and progress maximization methods struggle. When combined with imitation learning, the performance matches human levels (97.00%). The Study also explores the method's effectiveness in formulating control theory, such as 4-wheel independent drive (4WID) systems. With reduced spatial-temporal reasoning, reward shaping can match human performance (89.7%). However, control theory cannot be trained together with complicated spatial-temporal progress maximization.
View pdf
Traffic accidents pose a significant risk factor in urban areas, affecting those on the road, public mobility and the proper functioning of transport infrastructures. This paper presents an innovative framework based on artificial intelligence (AI) and geographic information system (GIS) that aims to predict traffic accidents and optimise emergency responses. The proposed framework focuses on high-risk areas in large cities where it predicts the occurrence of accidents based on historical data combined with traffic densities, weather conditions and proximity to intersections. The AI model trained based on these variables predicts accident zones, while the GIS interface provides spatially accurate visualisations. Early results show that emergency teams respond 20 percent faster to predicted high-risk zones, thereby improving urban traffic safety and efficiency. The study illustrates the practical potential of the combination of AI and GIS for improving transport infrastructures, particularly in accident prediction and emergency responses’ optimisation. The proposed framework will be tested in other cities, aiming at scaling the framework by integrating more variables related to urban mobility and further extending the application domains in city transport management.
View pdf
Vehicle navigation systems are one of the essential tools for automotive intelligence development, playing a crucial role in the process. This study discusses the components, operation principles, classification, and latest technological advances of Vehicle navigation systems, aiming to reveal the current state of the latest technological applications of the system in the automotive industry. The study indicates that the core value of vehicle navigation systems lies in precise positioning, enhanced driving safety, intelligent route planning, and other aspects. At present, the market of vehicle navigation systems is witnessing steady growth and faces intense competition from mobile phone navigation. To hold the upper hand in the competition, the industry should utilize policy support from the government, facing up to challenges and seeking solutions to current problems. In the future, the vehicle navigation system should deeply integrate with artificial intelligence (AI), providing diverse, tailored navigation services for customers. These services should cover driving skills, driving habits, etc. Meanwhile, through constant technological innovation, user experience optimization, and the application of deep learning, the vehicle navigation system is expected to achieve more efficient human-machine interaction and enhanced driving safety and comfortability, thereby improving its competitiveness in the market and turning it into an indispensable intelligent companion for drivers.
View pdf
Bothkennar clay is a type of soft clay found near the Firth of Forth in Scotland, at the Bothkennar site, which was established in the late 1980s as a national research center for advanced studies on soft clays. This paper analyzes the behavior of Bothkennar clay under both natural and remolded conditions, drawing on data from laboratory and field experiments. Special attention is given to the results of triaxial tests and oedometer tests, which are crucial for understanding the stress-strain response under different stress paths. The analysis largely employs the Cam-clay model, a widely used constitutive model in geotechnical engineering, to investigate the clay’s behavior during loading and unloading. Additionally, this study explores the clay’s internal structure, focusing on plastic anisotropy and destructuration, which reflect changes in the clay's behavior over time. A sensitivity framework and a normalization framework are also applied to better understand the deformation mechanisms and structural evolution of the clay. These studies not only shed light on the clay’s natural characteristics but also provide critical insights for model calibration and foundation stability predictions. As a benchmark material in soil mechanics, Bothkennar clay plays a key role in validating theoretical models and advancing research on the behavior of soft clays in geotechnical engineering.
View pdf
The advent of digital technology has spawned a revolution in software engineering, with artificial intelligence (AI) emerging as a key technology. The integration of advanced techniques, such as natural language processing (NLP) and deep learning has demonstrated AI’s remarkable capabilities throughout the software development lifecycle. Enhancements in areas such as code generation, code inspection, and software testing have significantly elevated both efficiency and quality. In addition, the potential of AI also provides new possibilities for automated software updates and maintenance in the future. However, despite the broad application prospects of AI in software engineering, it still faces some problems to be solved urgently. For example, inadequate adaptability and the challenges of personal data privacy protection limit its wider application. At the same time, the high research cost and immature model technology also bring obstacles to further development. By comprehensively analyzing the existing literature and related cases, this study deeply discusses the application status and limitations of AI in software engineering. The research results show that although AI can greatly improve the efficiency of software development, its shortcomings in data security and adaptability need attention. Future research should address these problems and seek more effective technical solutions to promote the sustainable development of AI in software engineering.
View pdf
This scientific research report focuses on the help of neural networks in fitness promotion. The project background points out that the growth in fitness demand, rich data, technological development, and the popularity of smart devices provide conditions for the application of neural networks in fitness promotion. The project content includes learning neural network knowledge, collecting data, and querying information. Through neural networks, personalized fitness plan recommendations,4 fitness effect prediction and motivation, assistance from smart fitness equipment, and fitness content recommendation and education can be achieved. The project encounters difficulties such as data collection, content organization, and language expression. Teachers provide help in data collection, content organization, and language improvement. The project gains are reflected in realizing the importance of the rigor and systematicness of scientific research, the spirit of innovative exploration, and the importance of patience and perseverance. At the same time, there is a new understanding of neural networks. Their learning ability is strong but there are challenges in interpretability, and combining with other technologies can play a greater role. In short, this project shows the potential of neural networks in the field of fitness promotion and the gains and challenges in the scientific research process.
View pdf
The construction industry in China is a major contributor to the nation’s negative environmental impact as it consumes a significant amount of natural resources and generates substantial greenhouse gas emissions and waste. In response, green building, aimed at designing and operating structures in an environmentally responsible way, has been more and more emphasised. Numerous studies have demonstrated the potential benefits of green buildings but green building practices still face several barriers, including high initial investment, lack of knowledge, incomplete regulations and codes and insufficient professional expertise. This research investigates some of the obstacles for green building practices in China, focusing on stakeholders’ perceptions. It finds that government barriers are the most dominant, followed by human-related and knowledge and operation-related barriers. Lastly, some recommendations regarding government incentives, public awareness initiatives and collaboration between government and construction professionals are proposed to promote the wider practice of green building. By addressing these challenges, the research aims to contribute to the advancement of sustainable construction in the country.
View pdf