The proceedings contain 71 papers. The topics discussed include: extracting software design from text: a machine learning approach;deep reinforcement learning based actor-critic framework for decision-making actions i...
ISBN:
(纸本)9781665440769
The proceedings contain 71 papers. The topics discussed include: extracting software design from text: a machine learning approach;deep reinforcement learning based actor-critic framework for decision-making actions in production scheduling;model-based test case generation approach for mobile applications load testing using OCL enhanced activity diagrams;deep learning achievements and opportunities in domain of electronic warfare applications;a proposed approach to secure automated teller machine-based financial transactions;astronomical image denoising based on convolutional neural network;an improved deep learning model for early fire and smoke detection on edge vision unit;long-term person re-identification model with a strong feature extractor;satellite imagery road segmentation using a dual network approach with enhancement blocks;multi-limb split learning for tumor classification on vertically distributed data;deep learning methodologies for human activity recognition;unlocking the public perception of COVID-19 vaccination process on social media;and identification of a new topology to enhance the impedance extraction in microfluidic systems.
So as to solve the high time complexity of Zernike moment, we take advamtage of advantage not only a highspeed multi-GPU environment using a unit circle inscribed image transform method but Zernike pyramid cache. Thro...
详细信息
ISBN:
(纸本)9798331517939;9788993215380
So as to solve the high time complexity of Zernike moment, we take advamtage of advantage not only a highspeed multi-GPU environment using a unit circle inscribed image transform method but Zernike pyramid cache. Through experiments using the Python PyTorch package, which provides not only convenience of data transfer between the GPU and GPU, but various tensdor operation function, the cost of recalculation of basia function is eliminated with the unit circle inscribed image transformation merhod and Zernik pyramid *** was verified that the calculation amount of the Zernike moment was redused by more 36% in a multi-GPU environment.
In this paper, we propose an effective low-rank alternating direction doubling algorithm (R-ADDA) for computing numerical low-rank solutions of large-scale sparse continuous-time algebraic Riccati matrix equations. Ou...
详细信息
In this paper, we propose an effective low-rank alternating direction doubling algorithm (R-ADDA) for computing numerical low-rank solutions of large-scale sparse continuous-time algebraic Riccati matrix equations. Our algorithm represents a further extension of the alternating direction doubling algorithm, utilizing the low-rank property of matrices. It is only required to compute one recursion and may apply the associated low-rank structures, solving large-scale problems efficiently. The low-rank formula can save storage space and computational complexity. Finally, we offer theoretical analysis and numerical experiments to illustrate the effectiveness of the derived algorithm.
This study addresses the optimization of grid-connected photovoltaic (PV) systems, particularly focusing on overcoming challenges posed by shading conditions. Employing machine learning (ML) technology, specifically R...
详细信息
ISBN:
(纸本)9798350367607;9798350367591
This study addresses the optimization of grid-connected photovoltaic (PV) systems, particularly focusing on overcoming challenges posed by shading conditions. Employing machine learning (ML) technology, specifically Reinforcement Learning (RL), this research conducts a comparative analysis with traditional optimization techniques such as Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) for Maximum Power Point Tracking (MPPT). Simulation results, executed using Simulink/Matlab, highlight RL's superior performance in terms of convergence speed and effectiveness compared to PSO and GA, without the need for prior system knowledge. This study contributes valuable insights into the application of ML-based algorithms in enhancing PV system efficiency, paving the way for advancements in renewable energy technologies.
Wireless Sensor Networks (WSN) that work hand in hand with IoT are crucial for contemporary smart city solutions. However, as these are dynamic and decentralized, they pose additional issues related to network data an...
详细信息
Model predictive contour control (MPCC) is a widely used control technique in vehicle tracking control. However, its performance highly depends on how accurately the model represents vehicle dynamics. The vehicle mode...
详细信息
ISBN:
(纸本)9798350399462
Model predictive contour control (MPCC) is a widely used control technique in vehicle tracking control. However, its performance highly depends on how accurately the model represents vehicle dynamics. The vehicle model uncertainty and the existence of external disturbances will cause uncertainty in the prediction and bring risk to the operation of the vehicle. To address this issue, a nonlinear disturbance observer is designed to estimate the model uncertainty of the vehicle and external disturbances. In addition, the weight coefficients of MPCC also affect its control performance. A fixed set of manually tuned parameters in the objective function is obviously not an optimal choice for different road scenarios. To obtain better control performance, a reinforcement learning tuned adaptive varying weight control scheme is designed. Simulation results show that the proposed method can achieve better control performance. The source code can be downloaded at https://***/ruidev1/AT-NDO-MPCC.
The control of non-holonomic vehicles require a joint control of the vehicle's position and orientation (pose). Pose control of multiple non-holonomic Connected and Automated Vehicles (CAVs) is computationally com...
详细信息
ISBN:
(纸本)9798350399462
The control of non-holonomic vehicles require a joint control of the vehicle's position and orientation (pose). Pose control of multiple non-holonomic Connected and Automated Vehicles (CAVs) is computationally complex. This paper investigates a method for pose control of multiple non-holonomic CAVs that distributes the computations among the CAVs. To this end, we use a Distributed Model Predictive control (DMPC) approach that is based on synchronization. This approach is purely distributed and does not require any central entity. The CAVs follow their reference trajectories while avoiding collisions. We generate the reference trajectories by considering the non-holonomic dynamics of the CAVs. Our demonstration on scaled CAVs shows the applicability of this approach. Furthermore, our evaluation shows the scalability of this approach in the numbers of CAVs.
Dynamically changing access rights of users in large-scale secure data sharing is an important challenge which designers of the secure systems have to address. We focus efficient enforcement of the dynamic access cont...
详细信息
ISBN:
(纸本)9783031505829;9783031505836
Dynamically changing access rights of users in large-scale secure data sharing is an important challenge which designers of the secure systems have to address. We focus efficient enforcement of the dynamic access control using key-aggregate cryptosystem (KAC), an efficient solution to secure data sharing. In this paper, we present a novel KAC construction that, in addition to satisfying all key-aggregate efficiency requirements, allows a data owner to enforce dynamic updates in access rights of a user much more efficiently than the existing ones. In particular, the proposed KAC construction handles the dynamic updates at the level of public parameters, and does not require the data owner to carry out any secure transmissions. This further means that none of the data users, including the one(s) whose access rights are updated, has to update their secrets. Thus, the dynamic update operation of the proposed KAC scheme is free from the one-affects-all problem. We present a formal security proof of the proposed KAC scheme and analyze its performance to further support our claims.
How to achieve autonomous navigation and attitude control for robots in uncertain environments has always been a hot research topic. This study designs an autonomous navigation closed-loop control algorithm based on a...
详细信息
The Internet of Things (IoT) is an innovative technology that encompasses the connectivity of physical, intelligent objects to the Internet, enabling the collection, sharing and analysis of data. Currently regarded as...
详细信息
ISBN:
(纸本)9798350349740;9798350349757
The Internet of Things (IoT) is an innovative technology that encompasses the connectivity of physical, intelligent objects to the Internet, enabling the collection, sharing and analysis of data. Currently regarded as one of the revolutionary technologies of the 21st century, the significant increase of IoT devices and the huge amount of data they produce at the network's edge has led to the Cloud computing (CC) paradigm being over-loaded. Consequently, new computing paradigms are emerging, such as Edge computing and Fog computing (FC). Although these paradigms offer different functionalities and improve Quality of Service (QoS), they also introduce huge risks in terms of data security and privacy. This paper presents a brief review on security in IoT environments based on Fog computing (FC) architecture, with particular attention to security measures such as authentication, confidentiality and data integrity. Furthermore, it addresses attacks targeting these measures, in order to avoid or reduce security issues in Fog computing-IoT environments.
暂无评论