Performance Of Systems Research Articles

Target temperature-based simulation method for predicting performance and designing aircraft thermal anti-icing systems

The prevention of ice build-up on aircraft surfaces is crucial for safe operations. Thermal ice protection systems, particularly electro-thermal systems, are commonly employed to mitigate ice formation. However, the weight of electric batteries used in these systems significantly impacts aircraft performance, necessitating the minimization of power load during design. Despite this requirement, effective anti-icing operation mandates a sufficient heat flux to maintain surface temperature above the freezing point and evaporate any runback water film. Previous research has utilized numerical simulation analysis to predict the performance of electro-thermal anti-icing systems. The conjugate heat transfer method has been commonly employed to predict temperature distribution on the protected surface. However, non-isothermal surfaces exhibit distinct local convective heat transfer coefficients that influence local thermal balance, unlike isothermal surfaces that adopt the loosely-coupled method for computational efficiency. In this study, we present a novel target temperature-based anti-icing simulation method that predicts the required heat flux for thermal ice protection systems without employing the conjugate heat transfer method. Our objective was to calculate the necessary heat flux to achieve the desired non-isothermal surface temperature for effective anti-icing. By eliminating the excessive iteration process associated with the coupling method for conjugate heat transfer, we have streamlined the simulation procedure. Through validation, target temperature-based anti-icing simulation method, when compared to the tight coupling method, demonstrated the ability to predict the same surface water flow rates while reducing computation time by up to 90 %. Furthermore, when applied to design problems, our method achieved approximately a 10 % reduction in the required heat capacity for anti-icing compared to experimental values. As a result, target temperature-based anti-icing simulation method offers a more efficient alternative to traditional conjugate heat transfer methods, providing valuable insights for both predicting anti-icing system performance and optimizing aircraft design.

Efficient Object Detection with YOLO: A Comprehensive Guide

Object detection presents itself as a pivotal and complex challenge within the domain of computer vision. Over the past ten years, as deep learning techniques have advanced quickly, researchers have committed significant resources to utilising deep models as the basis to improve the performance of object identification systems and related tasks like segmentation, localization. Two-stage and single-stage detectors are the two basic categories into which object detectors can be roughly divided. Typically, two-stage detectors use complicated structures in conjunction with a selective region proposal technique to accomplish their goals. Conversely, single-stage detectors aim to detect objects across all spatial regions in one shot, employing relatively simpler architectures. Any object detector's inference time and detection accuracy are the main factors to consider while evaluating it. Single-stage detectors offer quicker inference times, but two-stage detectors frequently show better detection accuracy. But since the introduction of YOLO (You Only Look Once) and its architectural offspring, detection accuracy has significantly improved—sometimes even outperforming that of two-stage detectors. The adoption of YOLO in various applications is primarily driven by its faster inference times rather than its detection accuracy alone.

Enhancing Linearity of Light Response in Avalanche Photodiodes by Suppressing Electrode Size Effect

The nonlinear characteristics of avalanche photodiodes (APDs) inhibit their performance in high-speed communication systems, thereby limiting their widespread application as optical detectors. Existing theoretical models have not fully elucidated complex phenomena encountered in actual device structures. In this study, actual APD structures exhibiting lower linearity than their ideal counterparts were revealed. Simulation analysis and physical inference based on GaN APDs reveal that electrode size is a noteworthy factor influencing response linearity. This discovery expands the nonlinear theory of APDs, suggesting that APD linearity can be enhanced by suppressing the electrode size effect. A physical model was developed to explain this phenomenon, which is attributed to charge accumulation at the edge of the contact layer. Therefore, we proposed an improved APD design that incorporates an additional gap layer and a buffer layer to stabilize the internal gain under high-current-density conditions, thereby enhancing linearity. Our improved APD design increases the linear threshold for optical input power by 4.46 times. This study not only refines the theoretical model for APD linearity but also opens new pathways for improving the linearity of high-speed optoelectronic detectors.

Diversification of spider silk properties in an adaptive radiation of Hawaiian orb-weaving spiders

IntroductionThe design of biological structures and the materials composing those structures are intimately connected to performance in biological systems. Spider webs present an excellent example of how design and materials interact during their function in capturing prey. Major shifts in how spider webs capture prey have occurred due to evolutionary changes in both web architecture and silk properties. However, these shifts are mostly described for long timescales deep within the spider’s tree of life. Hawaiian Tetragnatha presents an opportunity to ask if such shifts can occur at much shorter timescales because web design diverges significantly among closely related species on the same island while also converging with more distant relatives on other islands. Here, we provide an initial test of whether or not silk properties diversified during the recent adaptive radiation of Hawaiian Tetragnatha.MethodsWe obtained radial and capture spiral silk from orb webs for spiders on two islands and tested their tensile and adhesive properties. We also used solution-state NMR to compare the composition of low molecular weight compounds in the glue because of their influence on capture spiral stickiness.ResultsResults showed differences in the stiffness of radial silk among four populations of Hawaiian Tetragnatha, while extensibility remained unchanged. Although not statistically different, radial strength and toughness varied twofold among species. Stickiness varied threefold among the four populations of orb weavers. No conspicuous qualitative differences in the low molecular weight compound composition of aggregate glue were found, suggesting that differences in capture spiral stickiness were due to the amount or arrangement of glue droplets on threads.DiscussionWhile our sampling is modest, our data provide the first evidence that silk properties can evolve measurably over the relatively short timescales of the adaptive radiation of Hawaiian Tetragnatha spiders.

3D View Factor Power Output Modelling of Bifacial Fixed, Single, and Dual-Axis Agrivoltaic Systems

This study investigates the performance of agrivoltaic systems employing bifacial photovoltaic modules. A comparison between yield in Sweden and Italy was carried out. Three agrivoltaic system designs were evaluated: vertical fixed, single-axis tracker, and dual-axis tracker. The results showed that the specific production varied between 1090 to 1440 kWh/kWp/yr in Sweden and 1584 to 2112 kWh/kWp/yr in Italy, where the lowest production was obtained with the vertical fixed agrivoltaic system while the highest production was obtained with the dual-axis tracking agrivoltaic system. The vertical fixed design had a higher electricity production during low solar elevation angles, while the single-axis and dual-axis tracking designs had significantly higher power production during mid-day. The electricity production gain using a dual-axis tracker design was mostly during mid-day, but the increase compared to the single-axis tracker was only 1-2%. The study concludes that low-height, fixed agrivoltaic systems without tracking are well-suited for high-latitude countries like Sweden, while elevated systems with tracker solutions are more suitable for locations like Italy. The findings suggest that the performance of agrivoltaic systems with bifacial photovoltaic modules is highly dependent on geographical location and the specific characteristics of the crops grown beneath them.

RECOGNITION VERSUS RESOLUTION CHARTS FOR DEFOCUS CURVE DETERMINATION IN TRIFOCAL INTRAOCULAR LENSES.

To determine the impact of optotype choice on the determination of defocus curve in cataract patients implanted with presbyopia-correcting intraocular lens (IOL). Oftalvist Alicante, Alicante, Spain. Observational case-control study. Patients implanted with Asqelio Trifocal IOL (AST Products, Inc., Billerica, MA, USA) participated in this study. Uncorrected and distance corrected visual acuity and subjective refraction were determined. Monocular defocus curves were obtained 6 months after surgery with the CTS system using either ETDRS or Landolt C charts under photopic conditions (85 cd/m2), and range of vergence from +2.00D to -5.00D in 0.50D steps. A total of 49 patients were enrolled in the study, 24 in the EDTRS group (7 male, 17 female) and 25 in the Landolt C group (5 male, 20 female). Non-significant differences were found between patients conforming both groups, except for preoperative intraocular pressure and white-to-white distance. All patients were within ±1.00D from intended refraction after surgery. 75% of patients in the EDTRS group and 84% in the Landolt C group were within ±0.50D. Average difference between the groups across vergences provided by the defocus curve was 0.12±0.05 logMAR units, significant for all vergences (p<0.05). Differences in VA were significant for distance, intermediate and near between both groups, except for uncorrected distance VA. Defocus curves created using Landolt C charts yield significantly lower through-focus outcomes than those created with standard EDTRS charts. This should be taken into account when comparing the performance of presbyopia-correcting systems accross studies where recognition charts might have been used due to patient characteristics.

Electron-photon exchange-correlation approximation for quantum-electrodynamical density-functional theory

Quantum-electrodynamical density-functional theory (QEDFT) provides a promising avenue for exploring complex light-matter interactions in optical cavities for real materials. Similar to conventional density-functional theory, the Kohn-Sham formulation of QEDFT needs approximations for the generally unknown exchange-correlation functional. In addition to the usual electron-electron exchange-correlation potential, an approximation for the electron-photon exchange-correlation potential is needed. A recent electron-photon exchange functional [C. Schäfer , ], derived from the equation of motion of the nonrelativistic Pauli-Fierz Hamiltonian, shows robust performance in one-dimensional systems across weak- and strong-coupling regimes. Yet, its performance in reproducing electron densities in higher dimensions remains unexplored. Here we consider this QEDFT functional approximation from one- to three-dimensional finite systems and across weak to strong light-matter couplings. The electron-photon exchange approximation provides excellent results in the ultrastrong-coupling regime. However, to ensure accuracy also in the weak-coupling regime across higher dimensions, we introduce a computationally efficient renormalization factor for the electron-photon exchange functional, which accounts for part of the electron-photon correlation contribution. These findings extend the applicability of photon-exchange-based functionals to realistic cavity-matter systems, fostering the field of cavity QED (quantum-electrodynamics) materials engineering. Published by the American Physical Society 2024

Enhancing MPTCP performance in wireless networks with PRS-MPTCP

In recent years, there has been a significant increasing demand for wireless networks, particularly mobile communication. Additionally, smart devices like mobile phones and tablets are able to use multiple interfaces simultaneously. In this regard, the concept of Multipath TCP (MPTCP) has been introduced, enabling the utilization of multiple interfaces for concurrent communication. However, packet loss and subflow heterogeneity, especially in wireless networks, cause an increase in Out-of-Order (OfO) packets at the receiver node, which leads to a decrease in the total MPTCP throughput. To address these performance-related challenges, numerous schedulers have been proposed. However, most existing methods have primarily focused on improving performance without adequately considering the impact of packet loss. This research paper provides a comprehensive overview of MPTCP schedulers. Subsequently, we propose a Practical and Robust Scheduler for MPTCP (PRS-MPTCP) with the specific aim of minimizing OfO packets to improve MPTCP performance in wireless systems. The PRS-MPTCP scheduler takes into account various characteristics of each subflow, including RTT, CWND, and the number of packet losses, to effectively decide which packets should be assigned to which subflow. By using Mininet-WiFi emulation, fair comparisons between PRS-MPTCP and the state-of-the-art schedulers have been conducted, considering throughput, the number of OfO packets, and retransmission rates as performance metrics. The evaluation results reveal the profound impact of selected parameters on the behavior of the schedulers and the overall performance of MPTCP. Ultimately, the results demonstrate that PRS-MPTCP guarantees acceptable throughput and achieves lower retransmission rates and fewer OfO packets compared to other methods. In the PRS-MPTCP, the number of OfO packets has decreased by 38 %, 37 %, 45 %, and 44 % compared to BLEST, ECF, RR, and the Default scheduler, respectively.

From Time-Series to Hybrid Models: Advancements in Short-Term Load Forecasting Embracing Smart Grid Paradigm

This review paper is a foundational resource for power distribution and management decisions, thoroughly examining short-term load forecasting (STLF) models within power systems. The study categorizes these models into three groups: statistical approaches, intelligent-computing-based methods, and hybrid models. Performance indicators are compared, revealing the superiority of heuristic search and population-based optimization learning algorithms integrated with artificial neural networks (ANNs) for STLF. However, challenges persist in ANN models, particularly in weight initialization and susceptibility to local minima. The investigation underscores the necessity for sophisticated predictive models to enhance forecasting accuracy, advocating for the efficacy of hybrid models incorporating multiple predictive approaches. Acknowledging the changing landscape, the focus shifts to STLF in smart grids, exploring the transformative potential of advanced power networks. Smart measurement devices and storage systems are pivotal in boosting STLF accuracy, enabling more efficient energy management and resource allocation in evolving smart grid technologies. In summary, this review provides a comprehensive analysis of contemporary predictive models and suggests that ANNs and hybrid models could be the most suitable methods to attain reliable and accurate STLF. However, further research is required, including considerations of network complexity, improved training techniques, convergence rates, and highly correlated inputs to enhance STLF model performance in modern power systems.

Neural adaptive dynamic surface control for position tracking of electrohydraulic servo systems based on extended state observer

High-performance motion tracking of electro-hydraulic servo systems plays an important role in developing industrial technology. However, the system itself has parameter uncertainty, uncertain nonlinearity, and measurement noise, significantly decreasing the performance of conventional controllers designed based on nominal models. To improve the tracking performance of electro-hydraulic servo systems, an adaptive dynamic surface control method combining extended state observer and radial basis function neural network is proposed. Firstly, an extended state observer combining neural network output and parameter adaptation technology is designed without the need for speed and pressure feedback to observe the system’s unknown state variables and nonlinear matched disturbance. The design of neural networks is used to approximate mismatched disturbance with significant nonlinearity online. Then, the proposed extended state observer is introduced into the neural adaptive backstepping design, effectively compensating for simultaneous matched and mismatched disturbances, thereby suppressing most of the uncertain nonlinearity of the system. Meanwhile, online update of uncertain parameters further improves the tracking performance of the controller. In addition, the introduction of dynamic surface overcomes the “explosion of complexity” issue in backstepping design. The rigorous stability of the closed-loop system is proved by the Lyapunov stability theory. Finally, the effectiveness of the proposed controller is validated through comparative experiments on valve-controlled hydraulic cylinder.

Robust design of a two-stage composite sliding mode controller for enhanced performance in large-scale systems

This paper introduces an innovative composite sliding mode control strategy tailored for large-scale system, specifically tackling the intricate task of formulating a sliding mode control law for singularly perturbed systems. This paper delineates the development of a composite controller by dissecting it into subsystem controllers for both ‘slow’ and ‘fast’ subsystems, exploring its effectiveness in managing matched bounded external disturbances within the context of large-scale system. While sliding mode control proves to be a robust solution for handling uncertainty and disturbances, its inherent complexity and discontinuous characteristics pose significant challenges in crafting designs suitable for singularly perturbed systems.

Does Digital Leadership Impact the Performance of Lower-Middle-Class Hospitals?

Even though Indonesia has a large number of hospitals and has experienced significant progress in realizing Universal Health Coverage (UHC), service performance is still relatively low, so many patients seek treatment abroad. The performance of services and care systems in hospitals is often associated with low use of digital technology. The complexity of adopting digital technology in a hospital environment is not an easy thing to do, with the majority of users carrying out digital technology functions being health officers and professionals. This research aims to explore the influence of Digital Leadership in encouraging the adoption of digital technology and its influence on Hospital Performance. Survey data conducted on 285 leaders of Class C and D hospitals spread throughout Indonesia, through Structural Equation Modeling (SEM) analysis using Lisrel 8.80, it was found that there was a positive and significant influence on the role of Digital Adoption Capability as a mediation on Hospital Performance. It was found that Digital Leadership has a positive and significant influence on hospital performance both directly and through the mediation of Digital Adoption Capability. Environmental Dynamism factors have also been proven to significantly influence hospital management’s desire to improve performance through the adoption of digital technology. The model resulting from this research provides a comprehensive view of hospital management in developing effective digital technology adoption strategies in facing an ever-changing environment.

Real time conditioning monitoring of MOSFET using artificial neural network regression

The metal–oxide–semiconductor field-effect transistor (MOSFET) is a critical semiconductor device widely used in electrical and electronic systems. The performance of systems depends on the component’s reliability. More than 30% of failures of electronic and electrical systems are due to MOSFET failure. Industries are paying more apprehension towards the improvement of MOSFET reliability. Temperature stress is one of the key parameters to examine the reliability performance of semiconductor devices and to predict the lifetime of the components. This article provides a quick analysis of the MOSFET degradation models. The primary failure precursor metric that reflects the level of degradation is the MOSFET on-resistance R DS(on). In this paper, a linear, non-linear, and artificial neural network (ANN) deterioration model of MOSFETs is presented with the experimental data. The coefficient of determination (R-square), and root mean square error (RMSE) are used to compare the real data. Finally, it is demonstrated that the proposed model works well with experimental data. More than 500 data were trained for the proposed model. It can therefore predict the real-time conditioning monitoring of MOSFETs to increase the reliability of electrical systems.

A novel method of refining the performance of rail systems: an evaluation of system dynamics using discrete event simulation

PurposeThis paper aims to trial a novel method of improving the performance of rail systems. Accordingly, an evaluation of rail system dynamics (SD) using discrete event simulation (DES) will be undertaken. Globally, cities and their transportation systems face ongoing challenges with many of these resulting from complicated rail SD. To evaluate these challenges, this study utilized DES as the basis of the analysis of Melbourne Metro Rail's SD. The transportation SD processes including efficiency and reliability were also developed.Design/methodology/approachUsing DES, this research examines and determines the Melbourne Metro Rail's SD. Although the Melbourne Metro Rail is still in progress, the DES developed in this research examined the system requirements of functionality, performance and integration. As the basis of this examination, the Melbourne Metro Rail's optimization was simulated using the developed DES. As the basis of the experiment, a total of 50 trials were simulated. This included 25 samples for each of efficiency and reliability. The simulation not only scrutinized the SD but also underlined some of its shortfalls.FindingsThis study found that information and communication technology (ICT) was the pinnacle of system application. The DES development highlighted that both efficiency and reliability rates are the essential SD and thus fundamental for Melbourne Metro Rail system functionality. Specifically, the three elements of SD, capacity, continuity and integration are considered critical in improving the system functionality of Melbourne Metro Rail.Research limitations/implicationsThis particular mega rail infrastructure system was carefully analyzed, and subsequently, the DES was developed. However, since the DES is at its inception, the results are relatively limited without inclusive system calibration or validation process. Nonetheless, with some modifications, such as using different KPIs to evaluate additional systems variables and setting appropriate parameters to test the system reliability measures at different intensities, the developed DES can be modified to examine and evaluate other rail systems. However, if a broader system analysis is required, the DES model subsequently needs to be modified to specific system parameters.Practical implicationsThrough evaluation of Melbourne's Metro Rail in the manner described above, this research has shown the developed DES is a useful platform to understand and evaluate system efficiency and reliability. Such an evaluation is considered important when implementing new transport systems, particularly when they are being integrated into existing networks.Social implicationsEfficient rail networks are critical for modern cities and such systems, while inherently complex, aid local economies and societal cohesion through predictable and reliable movement of people. Through improved system functionality and greater efficiencies, plus improved passenger safety, security and comfort, the traveling public will benefit from the enhanced reliability of the transportation network that results from research as that provided in this paper.Originality/valueThis research paper is the first of its kind specifically focusing on the application of DES on the Melbourne Metro Rail System. The developed model aligns with the efficiency optimization framework, which is central to rail systems. The model shows the relationship between increased efficiency and optimizing system reliability. In comparison with more advanced mathematical modeling, the DES presented in this research provides robust, but yet rapid and uncomplicated system enhancements. These findings can better prepare rail professionals to adequately plan and devise appropriate system measures.

Model-free visual servoing based on active disturbance rejection control and adaptive estimator for robotic manipulation without calibration

Purpose Visual feedback control is a promising solution for robots work in unstructured environments, and this is accomplished by estimation of the time derivative relationship between the image features and the robot moving. While some of the drawbacks associated with most visual servoing (VS) approaches include the vision–motor mapping computation and the robots’ dynamic performance, the problem of designing optimal and more effective VS systems still remains challenging. Thus, the purpose of this paper is to propose and evaluate the VS method for robots in an unstructured environment. Design/methodology/approach This paper presents a new model-free VS control of a robotic manipulator, for which an adaptive estimator aid by network learning is proposed using online estimation of the vision–motor mapping relationship in an environment without the knowledge of statistical noise. Based on the adaptive estimator, a model-free VS schema was constructed by introducing an active disturbance rejection control (ADRC). In our schema, the VS system was designed independently of the robot kinematic model. Findings The various simulations and experiments were conducted to verify the proposed approach by using an eye-in-hand robot manipulator without calibration and vision depth information, which can improve the autonomous maneuverability of the robot and also allow the robot to adapt its motion according to the image feature changes in real time. In the current method, the image feature trajectory was stable in the camera field range, and the robot’s end motion trajectory did not exhibit shock retreat. The results showed that the steady-state errors of image features was within 19.74 pixels, the robot positioning was stable within 1.53 mm and 0.0373 rad and the convergence rate of the control system was less than 7.21 s in real grasping tasks. Originality/value Compared with traditional Kalman filtering for image-based VS and position-based VS methods, this paper adopts the model-free VS method based on the adaptive mapping estimator combination with the ADRC controller, which is effective for improving the dynamic performance of robot systems. The proposed model-free VS schema is suitable for robots’ grasping manipulation in unstructured environments.

Performance Analysis of Moving Target Shadow Detection in Video SAR Systems

The video synthetic aperture radar (ViSAR) system can utilize high-frame-rate scene motion target shadow information to achieve real-time monitoring of ground mobile targets. Modeling the characteristics of moving target shadows and analyzing shadow detection performance are of great theoretical and practical value for the optimization design and performance evaluation of ViSAR systems. Firstly, based on the formation mechanism and characteristics of video SAR moving target shadows, two types of shadow models based on critical size and shadow clutter ratio models are established. Secondly, for the analysis of moving target shadow detection performance in ViSAR systems, parameters such as the maximum detectable speed of moving targets, the minimum clutter backscatter coefficient, and the number of effective shadow pixels of moving targets are derived. Furthermore, the shadow characteristics of five typical airborne/spaceborne ViSAR systems are analyzed and compared. Finally, a set of simulation experiments on moving target shadow detection for the Hamas rocket launcher validates the correctness and effectiveness of the proposed models and methods.

High cyclic stability and low plateau of Ti2V2Zr/NaH/Al hydrogen storage composite system prepared by ultrasound-assisted ball milling

NaH/Al doped with Ti2V2Zr with the aim of lowering the rehydrogenation pressure, lowering the dehydrogenation temperature, and improving the hydrogenation/dehydrogenation kinetics of NaAlH4. It was found that moderate ultrasonic-assisted wet grinding was beneficial to fine particles efficiently, improved uniformity of particle size, and reduced lattice distortion because its crystalline phase can converge more active hydrogen atoms and reduce the energy barrier of H exchange between the alloy and the matrix, thus improving the comprehensive hydrogen storage performance of the composite systems. It could reversibly store 4.81 wt% of hydrogen and completely hydrogenation under a low hydrogen pressure of 55 atm, its starting and ending dehydrogenation temperatures decrease by 81 and 89 °C, respectively, and the activation energies of the two-step dehydrogenation decrease by 26.7 and 52.7 kJ mol−1 respectively, with a dehydrogenation capacity retention of up to 99.58% over five cycles.

Optimizing platoon safety through key node selection in pinning control strategy

The advent of Connected and Automated Vehicles (CAVs) and vehicular platooning has ignited high expectations for improving the safety performance of future transportation systems. Present control strategies for CAV platoons primarily concentrate on regulating each vehicle within the platoon. Although theoretically equipped to optimize a variety of performance metrics, this method is heavily dependent on communication and computational resources, which may lead to exorbitant costs. This study proposes a pinning control strategy that focuses on the selection of pivotal nodes, thereby facilitating the optimization of safety indexes for the entire platoon through the control of a select group of vehicles. The strategy encompasses two critical aspects: the identification of key control nodes and the application of node-based control. Initially, a feature matrix that includes both the intrinsic dynamics of the vehicles and the dynamics relative to other vehicles is developed. Spectral clustering is then utilized to identify these pivotal nodes. Subsequently, specialized control mechanisms for these nodes are devised, based on Proportional-Integral-Derivative (PID) controllers, with the objective of improving platoon safety. The efficacy of this strategy is corroborated through numerical simulations. To assess the safety performance, four categories of Surrogate Safety Measures (SSMs), which consider distance and acceleration, are applied. The findings confirm that the pinning control strategy significantly enhances performance across various SSMs, considerably reducing the risk of collisions during platoon operations. Moreover, it improves operational efficiency and platoon stability, establishing its value as an effective method for safety enhancement.

Research on emergency logistics information traceability model and resource optimization allocation strategies based on consortium blockchain.

In response to increasingly complex social emergencies, this study realizes the optimization of logistics information flow and resource allocation by constructing the Emergency logistics information Traceability model (ELITM-CBT) based on alliance blockchain technology. Using the decentralized, data immutable and transparent characteristics of alliance blockchain technology, this research breaks through the limitations of traditional emergency logistics models and improves the accuracy and efficiency of information management. Combined with the hybrid genetic simulated Annealing algorithm (HGASA), the improved model shows significant advantages in emergency logistics scenarios, especially in terms of total transportation time, total cost, and fairness of resource allocation. The simulation results verify the high efficiency of the model in terms of timeliness of emergency response and accuracy of resource allocation, and provide innovative theoretical support and practical scheme for the field of emergency logistics. Future research will explore more efficient consensus mechanisms, and combine big data and artificial intelligence technology to further improve the performance and adaptability of emergency logistics systems.

Advancing Federated Learning by Addressing Data and System Heterogeneity

In the emerging field of federated learning (FL), the challenge of heterogeneity, both in data and systems, presents significant obstacles to efficient and effective model training. This talk focuses on the latest advancements and solutions addressing these challenges. The first part of the talk delves into data heterogeneity, a core issue in FL, where data distributions across different clients vary widely and affect FL convergence. We will introduce the FedCor framework addressing this by modeling loss correlations between clients using Gaussian Process and reducing expected global loss. External covariate shift in FL is uncovered, demonstrating that normalization layers are crucial, and layer normalization proves effective. Additionally, class imbalance in FL degrades performance, but our proposed Federated Class-balanced Sampling (Fed-CBS) mechanism reduces this imbalance by employing homomorphic encryption for privacy preservation. The second part of the talk shifts focus to system heterogeneity, an equally critical challenge in FL. System heterogeneity involves the varying computational capabilities, network speeds, and other resource-related constraints of participating devices in FL. To address this, we introduce FedSEA, which is a semi-asynchronous FL framework that addresses accuracy drops by balancing aggregation frequency and predicting local update arrival. Additionally, we discuss FedRepre, a framework specifically designed to enhance FL in real-world environments by addressing challenges including unbalanced local dataset distributions, uneven computational capabilities, and fluctuating network speeds. By introducing a client selection mechanism and a specialized server architecture, FedRepre notably improves the efficiency, scalability, and performance of FL systems. Our talk aims to provide a comprehensive overview of the current research and advancements in tackling both data and system heterogeneity in federated learning. We hope to highlight the path forward for FL, underlining its potential in diverse real-world applications while maintaining data privacy and optimizing resource usage.