In FDD massive MIMO (mMIMO) systems, although assisted with Rel.16 eType ii feedback, Channel State Information (CSI) accuracy is limited by overhead, quantization error and user mobility. In this paper, we propose Mu...
详细信息
ISBN:
(纸本)9798350310900
In FDD massive MIMO (mMIMO) systems, although assisted with Rel.16 eType ii feedback, Channel State Information (CSI) accuracy is limited by overhead, quantization error and user mobility. In this paper, we propose Multimodal Hetero-frequency Channel Extrapolation (M-HCE) based CSI reconstruction, where channel model with enhanced Levy-Ito decomposition path structure and Karhuen-Loeve expansion are introduced. Downlink channel is reconstructed from the multimodal transformation of uplink Sounding Reference Signal (SRS) channel and CSI feedback, by exploring both partial downlink-uplink reciprocity between hetero-frequency channels and reported downlink CSI. The performance is verified in commercial evaluation platform, which shows 23%-66% throughput gain compared with eType ii CSI for user mobility 10-60 km/h and typical FDD operating bands.
Photovoltaic (PV) system performance is influenced by environmental factors such as irradiation, temperature, and shading effects, which we cannot control. However, to address this aspect, maximum power point tracking...
详细信息
Photovoltaic (PV) system performance is influenced by environmental factors such as irradiation, temperature, and shading effects, which we cannot control. However, to address this aspect, maximum power point tracking (MPPT) algorithms come into play. They keep an eye on the working point and modify it in real-time to comply with the Maximum Power Point (MPP), which assists in reducing the impact of environmental fluctuations. This, in turn, results in steady and ideal power generation from the PV system, increasing its dependability and efficiency regardless of the unpredictable climate conditions. This research compares three commonly used approaches for tracking the MPP in PV systems: the perturb and observe (P&O) method, fuzzy logic control (FLC), and Artificial Neural networks (ANN). This comparison is based on a simulation using MATLAB/Simulink to carry out an in-depth analysis of these three methods. This simulation enables a complete examination of all the three methods, considering parameters like complexity, stability, and efficiency. This study intends to give a comprehensive review of the performance and applicability of the P&O method, fuzzy logic control, and ANN in PV systems by assessing these parameters. The present research adds useful insights to current information by thoroughly examining three popular MPPT approaches with simulations in MATLAB/Simulink The findings can help researchers and professionals choose the best strategy for their PV systems based on their specific requirements. Furthermore, the comparison study highlights prospective possibilities for future research and MPPT technique improvements. Copyright (c) 2024 The Authors.
Synthetic defect generation is an important aid for advanced manufacturing and production processes. Industrial scenarios rely on automated image-based quality control methods to avoid time-consuming manual inspection...
详细信息
ISBN:
(纸本)9781728198354
Synthetic defect generation is an important aid for advanced manufacturing and production processes. Industrial scenarios rely on automated image-based quality control methods to avoid time-consuming manual inspections and promptly identify products not complying with specific quality standards. However, these methods show poor performance in the case of ill-posed low-data training regimes, and the lack of defective samples, due to operational costs or privacy policies, strongly limits their large-scale applicability. To overcome these limitations, we propose an innovative architecture based on an unpaired image-to-image (I2I) translation model to guide a transformation from a defect-free to a defective domain for common industrial products and propose simultaneously localizing their synthesized defects through a segmentation mask. As a performance evaluation, we measure image similarity and variability using standard metrics employed for generative models. Finally, we demonstrate that inspection networks, trained on synthesized samples, improve their accuracy in spotting real defective products.
With the rapid development of autonomous driving technology, the autonomous driving ability of vehicles has become an important research direction in intelligent transportation systems. Convolutional Neural network (C...
详细信息
ISBN:
(纸本)9781510685710
With the rapid development of autonomous driving technology, the autonomous driving ability of vehicles has become an important research direction in intelligent transportation systems. Convolutional Neural network (CNN) has become one of the core technologies in autonomous driving vision systems due to its excellent performance in image recognition and classification. Embedded systems also play an important role in the realization of autonomous vehicle driving due to their high efficiency and real-time nature. In this study, a vehicle autonomous driving scheme based on Convolutional Neural network (CNN) and embedded system is proposed. Firstly, the convolutional neural network was used to perform real-time image processing and feature extraction on the road scene. Specifically, through the superposition of multi-layer convolutional layers and pooling layers, features such as edges, textures, and objects in the image are extracted layer by layer, so as to achieve efficient recognition of complex road environments. In this process, the convolutional layer is used to extract local features, while the pooling layer is used to reduce dimensionality and prevent overfitting to ensure the robustness and efficiency of the model. Secondly, an embedded system was designed and optimized, and the trained CNN model was deployed on the system to ensure real-time processing power and efficient energy consumption management. The design of the embedded system focuses on the optimal allocation of hardware resources and the effective control of energy consumption to meet the real-time operation needs of vehicles under different road conditions. Specifically, the embedded system uses high-performance processors and low-power hardware modules to ensure fast inference and real-time decision-making capabilities of CNN models. In addition, the overall performance and reliability of the system are further improved through the co-design of software and hardware. Combining the above two techn
Model predictive control (MPC) has been widely employed in autonomous driving control, offering advantages including rolling optimization. However, the performance of MPC can be significantly compromised due to the in...
详细信息
ISBN:
(纸本)9798350352573
Model predictive control (MPC) has been widely employed in autonomous driving control, offering advantages including rolling optimization. However, the performance of MPC can be significantly compromised due to the inaccurate system model of the vehicle. In response, a novel MPC model optimization algorithm based on BiLSTM has been proposed to address the issue of inaccurate system models. Firstly, MPC-based pathtracking data is collected under varying speeds, paths, and wheelbases. A BiLSTM network is then trained using this dataset and integrated with a MPC controller. This integration allows for the update of the wheelbase parameters of the kinematic model, thereby improving the accuracy of the vehicle prediction model. The proposed method is validated through simulation experiments. The results demonstrate that the path tracking accuracy of the MPC combined with a BiLSTM method is higher than that of MPC and linear quadratic regulator (LQR). Concurrently, this approach yields a considerable enhancement in path tracking precision, with an improvement of 37.53%.
Motor imagery (MI) decoding methods are pivotal in advancing rehabilitation and motor control research. Effective extraction of spectral-spatial-temporal features is crucial for MI decoding from limited and low signal...
详细信息
Motor imagery (MI) decoding methods are pivotal in advancing rehabilitation and motor control research. Effective extraction of spectral-spatial-temporal features is crucial for MI decoding from limited and low signal-to-noise ratio electroencephalogram (EEG) signal samples based on brain-computer interface (BCI). In this paper, we propose a lightweight Multi-Feature Attention Neural network (M-FANet) for feature extraction and selection of multi-feature data. M-FANet employs several unique attention modules to eliminate redundant information in the frequency domain, enhance local spatial feature extraction and calibrate feature maps. We introduce a training method called Regularized Dropout (R-Drop) to address training-inference inconsistency caused by dropout and improve the model's generalization capability. We conduct extensive experiments on the BCI Competition IV 2a (BCIC-IV-2a) dataset and the 2019 World robot conference contest-BCI Robot Contest MI (WBCIC-MI) dataset. M-FANet achieves superior performance compared to state-of-the-art MI decoding methods, with 79.28% 4-class classification accuracy (kappa: 0.7259) on the BCIC-IV-2a dataset and 77.86% 3-class classification accuracy (kappa: 0.6650) on the WBCIC-MI dataset. The application of multi-feature attention modules and R-Drop in our lightweight model significantly enhances its performance, validated through comprehensive ablation experiments and visualizations.
This paper explores the control of a quadrotor in the presence of attitude tracking and actuator failure, and introduces the Nussbaum function in the design of the backstepping controller in order to eliminate the inf...
详细信息
Event-triggered communication and control provide high controlperformance in networked controlsystems without overloading the communication network. However, most approaches require precise mathematical models of th...
详细信息
Event-triggered communication and control provide high controlperformance in networked controlsystems without overloading the communication network. However, most approaches require precise mathematical models of the system dynamics, which may not always be available. Model-free learning of communication and control policies provides an alternative. Nevertheless, existing methods typically consider single-agent settings. This paper proposes a model-free reinforcement learning algorithm that jointly learns resource-aware communication and control policies for distributed multi-agent systems from data. We evaluate the algorithm in a high-dimensional and nonlinear simulation example and discuss promising avenues for further research.
The transportation industry's focus on improving performance and reducing costs has driven the integration of IoT and machine learning technologies. The correlation between driving style and behavior with fuel con...
详细信息
The transportation industry's focus on improving performance and reducing costs has driven the integration of IoT and machine learning technologies. The correlation between driving style and behavior with fuel consumption and emissions has highlighted the need to classify different driver's driving patterns. In response, vehicles now come equipped with sensors that gather a wide range of operational data. The proposed technique collects critical vehicle performance data, including speed, motor RPM, paddle position, determined motor load, and over 50 other parameters through the OBD interface. The OBD-ii diagnostics protocol, the primary diagnostic process used by technicians, can acquire this information via the car's communication port. OBD-ii protocol is used to acquire real-time data linked to the vehicle's operation. This data are used to collect engine operation-related characteristics and assist with fault detection. The proposed method uses machine learning techniques, such as SVM, AdaBoost, and Random Forest, to classify driver's behavior based on ten categories that include fuel consumption, steering stability, velocity stability, and braking patterns. The solution offers an effective means to study driving behavior and recommend corrective actions for efficient and safe driving. The proposed model offers a classification of ten driver classes based on fuel consumption, steering stability, velocity stability, and braking patterns. This research work uses data extracted from the engine's internal sensors via the OBD-ii protocol, eliminating the need for additional sensors. The collected data are used to build a model that classifies driver's behavior and can be used to provide feedback to improve driving habits. Key driving events, such as high-speed braking, rapid acceleration, deceleration, and turning, are used to characterize individual drivers. Visualization techniques, such as line plots and correlation matrices, are used to compare drivers' performanc
In this research paper we focus on the design and implementation of reliable real time wireless sensor network (WSN) protocols that make use of relay node and proper localization techniques to enhance the performance ...
详细信息
暂无评论