Reduce Convergence Time Research Articles

Extra dimension algorithm: a breakthrough for optimization and enhancing DNN efficiency

Proposing an efficient meta-heuristic to improve the inputs of a trainer in deep neural network (DNNs) is significant. According to the Kaluza’s theory, there exists an extra dimension in the universe. This paper proposes a novel algorithm, extra dimension algorithm (EDA), which is simulated based on this theory. The proposed algorithm utilizes the extra dimension to evaluate the current region of solutions and determine the best direction to follow for the next step of the process. Finally, EDA is used to improve inputs of DNN in the process of solving optimization test problems. The same DNN with and without EDA is used to solve extensive optimization problems, including energy-related tasks. The efficiency of EDA in DNN is assessed by solving some test problems in references, the feasibility and efficiency of solutions, within a suitable number of iterations are demonstrated according to the results. The contributions of this paper are as follows: (1) Introduction of the EDA based on Kaluza’s theory. (2) Application of EDA to enhance the performance of DNNs. (3) Demonstration of EDA’s effectiveness in solving complex optimization problems. (4) Comprehensive evaluation of EDA’s impact on energy optimization problems and other test cases. (5) EDA achieved an average improvement of 15% in optimization accuracy and reduced convergence time compared to the best-performing alternatives.

Evaluation of burning detection using modified VGG19 for LULC classification changes

Abstract. This study aims to evaluate the effectiveness of a modified architecture of convolutional neural network (CNN) VGG19 for detecting fires and changes in land use and land cover classification (LULC). Remote sensing data from the Landsat 8 Operational Land Imagery (OLI) satellite was used to collect images from two distinct regions, one of which was used to obtain a dataset containing 1000 labeled images, and the other region was used to perform inference and verify the generalization of the model in an area with a high annual occurrence of fires. Analyses were conducted using a time series of normalized difference vegetation index (NDVI) and complementary cumulative distribution function (CCDF), to determine the potential for analysis in that area and define the periods of burning, pre-burning and post-burning. The VGG19 architecture was modified to maintain the input sizes of the images, resulting in a significant increase of 20.90 percentage points in the F1 score compared to the original architecture, as well as a 68.76% reduction in convergence time. In addition, the Gradient-weighted Class Activation Mapping (Grad-CAM) technique was used to improve the interpretability of the model at the moment of inference. The proposed methodology offers an approach for detecting burns by altering the LULC classification, and the modified VGG19 showed superior results.

Smart hybrid algorithm for global maximum power point tracking under partial shading conditions

ABSTRACT Maximum power point tracking (MPPT) is a complex task due to the nonlinear behavior of the photovoltaic (PV) array. Moreover, partial shading (PS) is a major problem that significantly decreases the overall system’s efficiency, whereas traditional algorithms are mostly local searches and thus cannot guarantee global optimality. Therefore, Swarm Intelligence (SI) algorithms are developed to circumvent this problem. Existing SI algorithms like Cuckoo Search (CS) and particle swarm optimization (PSO) have been used to address partial shading issues; nevertheless, there is a research gap in developing hybrid methods that combine these two algorithms to enhance MPPT performance under various environmental conditions. This paper fills this research gap by providing a novel hybrid algorithm called Smart Hybrid Algorithm (SHA). The developed optimizer avoids the common disadvantages of conventional MPPT techniques and provides a simple and robust MPPT tracker to handle PS in PV systems effectively. The proposed optimizer features a reset function to restart the search process for the new GMPP at any irradiance change. Multiple performance tests were conducted using 4S, 8S, 10S, 12S, and 4S2P PV array configurations to investigate the performance of our optimizer. The results obtained are compared to those of well-known recent algorithms. The proposed system demonstrates a convergence time of less than one second and a steady state power efficiency exceeding 99.79% across various complex partial shading scenarios. It also offers a convergence time reduction of at least 14% compared to other techniques. The findings demonstrate the robustness and suitability of SHA to handle complex partial shading conditions (CPSCs) and uniform irradiance with high dynamic and steady-state efficiencies and minor transient power fluctuations.

A hybrid gradient climbing algorithm for a swarm robot-based gas leak detector

Methane emissions from leak sources can have a negative climate impact, in addition to contributing to the risk of explosions in urban environments. These risks can be minimized by developing systems that provide for an accurate and timely detection and localization of a gas leakage point. This research used a swarm of robots to detect and locate a leakage point. The localization algorithm derives from further optimization of the gradient climbing algorithm using fireflies acting as opportunistic agents. Firefly agents are characterized by their bioluminescent communication which guides them to dynamically adjust their positions and intensities based on the quality of the gradient information available to them. The proposed research focuses on enhancing gas leak detection through the development of a hybrid gradient climbing algorithm. This algorithm integrates gradient climbing techniques with swarm intelligence, utilizing the strengths of both approaches. This simulation resulted in the hybrid algorithm leading to a reduced convergence time and path lengths when compared to the swarm without opportunistic agents. The suggested approach can be important especially in gas distribution systems or in areas where human intervention is considered to be unsafe.

Estimation of lithium battery state of charge using the LTG-SABO-GRU model

Abstract Accurate estimation of the state of charge (SOC) in lithium batteries is crucial for optimizing energy utilization and ensuring battery safety within Battery Management Systems (BMS). While deep learning techniques have made significant progress, time-series models based on the Gate Recurrent Unit (GRU) have gained widespread application in SOC estimation. However, their performance heavily hinges on the initial hyperparameter settings, impacting both precision and application range. To address this challenge, we propose a novel algorithm—the Logistic-tent-Gold Subtraction Average-based Optimizer (LTG-SABO)—which combines composite chaotic mapping with the golden sine algorithm. The LTG-SABO algorithm aims to optimize key hyperparameters of the GRU model, thereby enhancing precision and robustness in SOC estimation. By leveraging the Logistic-tent composite chaotic mapping for population initialization, our approach not only expands the search space but also effectively prevents algorithm convergence to local optima. Additionally, integrating the Gold-SA strategy further enhances the global search capability of the SABO algorithm, significantly reducing convergence time. The computational results reveal that the proposed LTG-SABO-GRU model outperforms the traditional GRU model in estimating SOC precision under both normal and extreme temperature conditions. Specifically, the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) show a substantial improvement, increasing by over 50% compared to the traditional model. Moreover, the LTG-SABO-GRU model exhibits significantly fewer convergence iterations than existing typical population optimization algorithms. This study introduces a novel, efficient, and practical approach for SOC estimation in BMS applications.

Design and Development of Multi-Agent Reinforcement Learning Intelligence on the Robotarium Platform for Embedded System Applications

This research explores the use of Q-Learning for real-time swarm (Q-RTS) multi-agent reinforcement learning (MARL) algorithm for robotic applications. This study investigates the efficacy of Q-RTS in the reducing convergence time to a satisfactory movement policy through the successful implementation of four and eight trained agents. Q-RTS has been shown to significantly reduce search time in terms of training iterations, from almost a million iterations with one agent to 650,000 iterations with four agents and 500,000 iterations with eight agents. The scalability of the algorithm was addressed by testing it on several agents’ configurations. A central focus was placed on the design of a sophisticated reward function, considering various postures of the agents and their critical role in optimizing the Q-learning algorithm. Additionally, this study delved into the robustness of trained agents, revealing their ability to adapt to dynamic environmental changes. The findings have broad implications for improving the efficiency and adaptability of robotic systems in various applications such as IoT and embedded systems. The algorithm was tested and implemented using the Georgia Tech Robotarium platform, showing its feasibility for the above-mentioned applications.

CHEABC-QCRP: A novel QoS-aware cluster routing protocol for industrial IoT

Clustering routing protocols currently have problems such as Single point of failure of cluster head nodes, poor network dynamics, uneven data transmission, etc., which are critical to the optimization of energy efficiency, network lifespan and network topology control. However, this optimization problem is an NP hard problem that conventional algorithms are difficult to solve. This paper proposes a new multi-objective cluster routing protocol (CHEABC-QCRP) aimed at optimizing network energy consumption, system lifespan, and quality of services (QoS). The protocol is based on a new chaotic hybrid elite artificial bee colony algorithm (CHEABC) proposed in this paper, which has strong search ability and greatly reduces convergence time. At the same time, a new chaotic strategy was designed to effectively prevent falling into local optima and premature convergence. In simulation experiments, compared with multiple routing protocols, a large number of test results show that this protocol significantly reduces network energy consumption, greatly improves system lifespan, and effectively improves QoS in IWSN.

A level set approach using adaptive local pre-fitting energy for image segmentation with intensity non-uniformity

Active contour model (ACM) is considered as one of the most frequently employed models in image segmentation due to its effectiveness and efficiency. However, the segmentation results of images with intensity non-uniformity processed by the majority of existing ACMs are possibly inaccurate or even wrong in the forms of edge leakage, long convergence time and poor robustness. In addition, they usually become unstable with the existence of different initial contours and unevenly distributed intensity. To better solve these problems and improve segmentation results, this paper puts forward an ACM approach using adaptive local pre-fitting energy (ALPF) for image segmentation with intensity non-uniformity. Firstly, the pre-fitting functions generate fitted images inside and outside contour line ahead of iteration, which significantly reduces convergence time of level set function. Next, an adaptive regularization function is designed to normalize the energy range of data-driven term, which improves robustness and stability to different initial contours and intensity non-uniformity. Lastly, an improved length constraint term is utilized to continuously smooth and shorten zero level set, which reduces the chance of edge leakage and filters out irrelevant background noise. In contrast with newly constructed ACMs, ALPF model not only improves segmentation accuracy (Intersection over union(IOU)), but also significantly reduces computation cost (CPU operating time T), while handling three types of images. Experiments also indicate that it is not only more robust to different initial contours as well as different noise, but also more competent to process images with intensity non-uniformity.

Evaluating the reliability of microgrids consisting of renewable energy sources using stochastic scheduling based on the data-driven uncertainty set

The paper addresses challenges arising from the unpredictable and probabilistic nature of wind speed and solar irradiance, leading to variable power production in renewable sources. Power imbalance issues in grids result from the difficulty in accurately estimating these variables. The study introduces an improved data-driven uncertainty set, utilizing a neural network trained with extensive historical data from a multi-microgrid system. This set efficiently identifies strong coupling and extracts characteristic scenarios, ensuring high reliability with a minimal number of scenes. The proposed probabilistic model, implemented in the coordination of pumped storage, wind, and solar power plants, incorporates energy limits and storage characteristics. A four-mode IEEE model for storage units allows quick response to system needs. Simulating storage unit states every hour, considering system load, output power, and energy limitations, optimizes power dispatch based on battery capacity. Additionally, the paper addresses the uncertainty in electric power demand in microgrids, proposing an improved data-driven uncertainty set obtained from a neural network. By eliminating unrealistic worst-case scenarios, the approach accurately captures system characteristics, maintaining high reliability while significantly reducing convergence time. The proposed uncertainty set is applied in a novel two-stage robust optimization dispatch model, transforming the optimization problem into a mixed-integer linear programming problem, and solving it using a column and constraint generation algorithm. Numerical case studies verify the feasibility and superiority of the proposed approach in terms of economy and convergence performance compared to existing robust optimization methods.

Enhanced particle swarm optimization and P&O for MPPT of photovoltaic systems under partial shading conditions

Although Photovoltaic Technologies are largely deployed as a renewable energy source, several factors affect their performance. The major factors that affect PV performance are changes in irradiance and temperature. Maximum PowerPoint Tracking of PV output is essential in giving the maximum photovoltaic outputs at variable levels. Instantaneous variation in irradiance and temperature increases the complexity of tracking maximum power points. Partial shading conditions resulting from shade from trees, tall buildings, and Cloud formation amongst others greatly affect PV systems, especially in large Photovoltaic systems. Under the Partial shading condition, P-V curves become more complex as it is characterized by multiple peaks. The conventional PSO is associated with less accuracy in tracking the Global Maximum Power Point (Global MPP) and slow convergence time in obtaining the Global MPPT and oscillations. In this thesis, a modified Particle Swarm Optimization based Maximum Power Point Tracking technique is designed in MATLAB/Simulink to track the global Maximum Power Point of a Photovoltaic system under partial shading. The proposed modified PSO combines conventional PSO and P&O methods. The particle position in the PSO method is given as the duty cycle value d of the DC-DC converter. Conventional PSO equations are used to update the velocities and duty cycle. Thereafter, the maximum velocity and duty cycle are perturbed to reduce convergence time. The designed PV system was simulated in a MATLAB environment for 10 different irradiation levels and results were compared with results from related works. The average convergence time was 0.99 seconds and efficiency was up to 99.8% with the proposed model which performed better than conventional methods.

Multiagent-Based Control for Plug-and-Play Batteries in DC Microgrids with Infrastructure Compensation

The influence of the DC infrastructure on the control of power-storage flow in micro- and smart grids has gained attention recently, particularly in dynamic vehicle-to-grid charging applications. Principal effects include the potential loss of the charge–discharge synchronization and the subsequent impact on the control stabilization, the increased degradation in batteries’ health/life, and resultant power- and energy-efficiency losses. This paper proposes and tests a candidate solution to compensate for the infrastructure effects in a DC microgrid with a varying number of heterogeneous battery storage systems in the context of a multiagent neighbor-to-neighbor control scheme. Specifically, the scheme regulates the balance of the batteries’ load-demand participation, with adaptive compensation for unknown and/or time-varying DC infrastructure influences. Simulation and hardware-in-the-loop studies in realistic conditions demonstrate the improved precision of the charge–discharge synchronization and the enhanced balance of the output voltage under 24 h excessively continuous variations in the load demand. In addition, immediate real-time compensation for the DC infrastructure influence can be attained with no need for initial estimates of key unknown parameters. The results provide both the validation and verification of the proposals under real operational conditions and expectations, including the dynamic switching of the heterogeneous batteries’ connection (plug-and-play) and the variable infrastructure influences of different dynamically switched branches. Key observed metrics include an average reduced convergence time (0.66–13.366%), enhanced output-voltage balance (2.637–3.24%), power-consumption reduction (3.569–4.93%), and power-flow-balance enhancement (2.755–6.468%), which can be achieved for the proposed scheme over a baseline for the experiments in question.

EPPTA: Efficient partially observable reinforcement learning agent for penetration testing applications

AbstractIn recent years, penetration testing (pen‐testing) has emerged as a crucial process for evaluating the security level of network infrastructures by simulating real‐world cyber‐attacks. Automating pen‐testing through reinforcement learning (RL) facilitates more frequent assessments, minimizes human effort, and enhances scalability. However, real‐world pen‐testing tasks often involve incomplete knowledge of the target network system. Effectively managing the intrinsic uncertainties via partially observable Markov decision processes (POMDPs) constitutes a persistent challenge within the realm of pen‐testing. Furthermore, RL agents are compelled to formulate intricate strategies to contend with the challenges posed by partially observable environments, thereby engendering augmented computational and temporal expenditures. To address these issues, this study introduces EPPTA (efficient POMDP‐driven penetration testing agent), an agent built on an asynchronous RL framework, designed for conducting pen‐testing tasks within partially observable environments. We incorporate an implicit belief module in EPPTA, grounded on the belief update formula of the traditional POMDP model, which represents the agent's probabilistic estimation of the current environment state. Furthermore, by integrating the algorithm with the high‐performance RL framework, sample factory, EPPTA significantly reduces convergence time compared to existing pen‐testing methods, resulting in an approximately 20‐fold acceleration. Empirical results across various pen‐testing scenarios validate EPPTA's superior task reward performance and enhanced scalability, providing substantial support for efficient and advanced evaluation of network infrastructure security.

Emergency medical supplies scheduling during public health emergencies: algorithm design based on AI techniques

Based on AI technology, this study proposes a novel large-scale emergency medical supplies scheduling (EMSS) algorithm to address the issues of low turnover efficiency of medical supplies and unbalanced supply and demand point scheduling in public health emergencies. We construct a fairness index using an improved Gini coefficient by considering the demand for emergency medical supplies (EMS), actual distribution, and the degree of emergency at disaster sites. We developed a bi-objective optimisation model with a minimum Gini index and scheduling time. We employ a heterogeneous ant colony algorithm to solve the Pareto boundary based on reinforcement learning. A reinforcement learning mechanism is introduced to update and exchange pheromones among populations, with reward factors set to adjust pheromones and improve algorithm convergence speed. The effectiveness of the algorithm for a large EMSS problem is verified by comparing its comprehensive performance against a super-large capacity evaluation index. Results demonstrate the algorithm's effectiveness in reducing convergence time and facilitating escape from local optima in EMSS problems. The algorithm addresses the issue of demand differences at each disaster point affecting fair distribution. This study optimises early-stage EMSS schemes for public health events to minimise losses and casualties while mitigating emotional distress among disaster victims.

Adaptive Satellite Navigation Anti-Interference Algorithm Based on Inverse Cosine Function

Contrasting the dilemma that the traditional time-domain least mean square (LMS) algorithm in the existing satellite navigation receiver anti-interference system cannot satisfy the convergence time is short and maintain a low level of steady-state error at the same time, an inverse cosine variable step size LMS algorithm (ICVS-LMS) is proposed. To begin with, the LMS algorithm, with a fixed step size focuses on its effectiveness in attenuating and suppressing interference signals, is analyzed, and then the proposed ICVS-LMS algorithm is analyzed. In conclusion, both the ICVS-LMS algorithm and the traditional algorithm are simulated and compared in terms of their effectiveness in suppressing interference in satellite navigation signals. The experimental results demonstrate that the improved algorithm significantly reduces convergence time while maintaining a small steady-state error. The improved algorithm demonstrates high robustness and an obvious suppression effect on interference signals. The anti-interference performance is 8.41–12.22% higher than that of the proposed algorithm.

RPIFL: Reliable and Privacy-Preserving Federated Learning for the Internet of Things

Federated Learning for the Internet of Things(FL for IoT) contributes to the enhancement of security in smart cities. However, the privacy disclosure attacks within the FL for IoT framework exposes user’s local data . Although differential privacy mechanism (DP) ensures the privacy of IoT data, one crucial issue is DP affects model accuracy and reduces model reliability. Besides, DP cannot provide comprehensive privacy preservation in the IoT environment. To tackle the above problems, we propose reliable federated learning with privacy-preserving for IoT(RPIFL). Our approach comprises three key components. First, we leverage a lightweight network to hide the private data locally, then use the output to perform the next model training. Second, we propose a multi-modal fine-grained model optimization, which significantly improves the accuracy of the IoT model. Third, we implement adaptive differential privacy mechanisms tailored to dynamic privacy requirements, thereby safeguarding the confidentiality of IoT data. To empirically validate our framework, we conducted a simulation experiment on an image classification dataset. Our methodology concurrently ensures privacy protection and achieves superior accuracy compared to the baseline model with an average accuracy increase of 1.1% over the two prevailing federated privacy protection methods, coupled with an average reduction in convergence time of over 9%.

Autonomous Control System with Passive Positioning for Unmanned-Aerial-Vehicle-Assisted Edge Communication in 6G

UAVs can be deployed in many scenarios to provide various types of services via 6G edge communication. In these scenarios, it is necessary to obtain the position of the UAVs in a timely and accurate manner to avoid UAV collisions. In this paper, we consider improved passive localization algorithms aimed at reducing convergence time and adapting to extreme conditions. For the sake of reducing the complexity of signals and ensuring the reliability of receiving processes, we reconsidered the angle between arrival signals as the feature in positioning. Then, according to the characteristics of the positioning process, we draw on the cyclical process of the iterative greedy algorithm to construct the coding, destruction, and reorganization process to guide the movement of the UAV. Moreover, an improved Metropolis criterion is added to prevent falling into the local optimal solution. Finally, the proposed algorithm is verified in the simulation results. The results show that the algorithm can achieve precise positioning and excellent track planning within a small number of iterations, and it reduces the amount of information carried by the signal and convergence time compared with the traditional method.

Federated learning with hyper-parameter optimization

Federated Learning is a new approach for distributed training of a deep learning model on data scattered across a large number of clients while ensuring data privacy. However, this approach faces certain limitations, including a longer convergence time compared to typical deep learning models. Existing federated optimization algorithms often employ the same hyper-parameters for all clients, disregarding potential system heterogeneity and varying local data availability, which contributes to an even longer convergence time and more communication rounds. To address this challenge, we propose FedHPO, a new federated optimization algorithm that adaptively modifies hyper-parameters of each client’s local model during training, such as learning rate and epochs. This adaptability facilitates quicker convergence of the client’s local model, which in turn helps the global model converge faster, consequently reducing overall convergence time and the required communication rounds. In addition, FedHPO does not require any additional complexity since each client adjusts hyper-parameters independently based on the training results obtained in each epoch. In our evaluation, we compare FedHPO with other algorithms, such as FedAVG, FedAVGM, FedProx, and FedYogi, using both IID and non-IID distributed datasets. The results demonstrate the promising outcomes of FedHPO, showcasing reduced convergence time and fewer required communication rounds in comparison to alternative algorithms.

Go beyond the limits of genetic algorithm in daily covariate selection practice

Covariate identification is an important step in the development of a population pharmacokinetic/pharmacodynamic model. Among the different available approaches, the stepwise covariate model (SCM) is the most used. However, SCM is based on a local search strategy, in which the model-building process iteratively tests the addition or elimination of a single covariate at a time given all the others. This introduces a heuristic to limit the searching space and then the computational complexity, but, at the same time, can lead to a suboptimal solution. The application of genetic algorithms (GAs) for covariate selection has been proposed as a possible solution to overcome these limitations. However, their actual use during model building is limited by the extremely high computational costs and convergence issues, both related to the number of models being tested. In this paper, we proposed a new GA for covariate selection to address these challenges. The GA was first developed on a simulated case study where the heuristics introduced to overcome the limitations affecting currently available GA approaches resulted able to limit the selection of redundant covariates, increase replicability of results and reduce convergence times. Then, we tested the proposed GA on a real-world problem related to remifentanil. It obtained good results both in terms of selected covariates and fitness optimization, outperforming the SCM.

Experimental validation of an advanced metaheuristic algorithm for maximum power point tracking of a shaded photovoltaic system: A comparative study between three approaches

Under partially shaded conditions, the power–voltage characteristic curve of the Photovoltaic System (PVS) presents more than one peak, so the Global Maximum Power Point (GMPP) cannot be detected using the conventional Maximum Power Point Tracking (MPPT) algorithms, such as the Perturb-and-Observe (P&O) algorithm. In order to overcome the limitations of the conventional MPPT algorithms, this paper suggests a metaheuristic MPPT called the Crow Search Algorithm (CSA) for the performance optimization of a standalone PVS. The CSA algorithm has the capability of attenuating the negative effects of the partial shading on the performance of the PVS by the accurate detection of the GMPP. The principle of the latter algorithm uses the crow skills and behaviors in the process of locating places to hide its food. Relative to other metaheuristic methods, the CSA utilizes only two tuning parameters that combine between simplicity of implementation and good efficiency. The simulation and experimental results under partial shading applications demonstrates the better performance of the suggested CSA algorithm compared to the particle swarm optimization and P&O algorithms. In fact, the comparison is carried out in terms of high efficiency, good accuracy, low convergence time and simplicity of implementation. Indeed, the proposed CSA-based MPPT approach extracts the maximum power produced by the PVS with an estimated average efficiency of 99.87%, whereas the PSO and P&O methods record average efficiencies of 99.39% and 95.23%, respectively. Furthermore, as compared to the PSO and P&O MPPT methods, the suggested CSA-based MPPT approach reduces convergence time by an average of 48.41% and 49.63%, respectively.

An adaptive stochastic resonance detection method based on a fast artificial fish swarm algorithm

A new stochastic resonance method that is based on a fast artificial fish swarm algorithm has been proposed in an effort to address the adaptive parameter-induced stochastic resonance for weak signal detection’s slow convergence time. The target evaluation function for the system is the output signal-to-noise ratio. The method of scale transformation and amplitude compression is used to pre-process the high frequency and large parameter signals. To achieve fast adaptive detection that applies to weak communication signals, the stochastic resonance system’s characteristics are used to constrain the optimization iteration rules.According to the simulation results, the fast artificial fish swarm method has significantly better optimization efficiency and achieves the same optimization results as the basic artificial fish swarm algorithm while reducing convergence time by 74.89%.