Adaptive Dynamic Method Research Articles

Dynamic Adaptive Decision-Making Method for Autonomous Navigation of Ships in Coastal Waters

Dynamic Adaptive Decision-Making Method for Autonomous Navigation of Ships in Coastal Waters

DAR-DRL: A dynamic adaptive routing method based on deep reinforcement learning

Mobile-centric wireless networks offer users a diverse range of services and experiences. However, existing intelligent routing methods often struggle to make suitable routing decisions during dynamic network changes, significantly limiting transmission performance. This paper proposes a dynamic adaptive routing method based on Deep Reinforcement Learning (DAR-DRL) to effectively address these challenges. First, to accurately model network state information in complex and dynamically changing routing tasks, we introduce a link-aware graph learning model (LA-GNN) that efficiently senses network information of varying structures through a hierarchical aggregated message-passing neural network. Second, to ensure routing reliability in dynamic environments, we design a hop-by-hop routing strategy featuring a large acceptance domain and a reliability guarantee reward function. This mechanism adaptively avoids routing holes and loops across various network scenarios while enhancing the robustness of routing under dynamic conditions. Experimental results demonstrate that the proposed DAR-DRL method achieves the network routing task with shorter end-to-end delays, lower packet loss rates, and higher throughput compared to existing mainstream methods across common dynamic network scenarios, including cases with dynamic traffic variations, random link failures (small topology changes), and significant topology alterations.

Study of the Dynamic Adaptive Calculation Method for River Water Environmental Capacity

Controlling the total amount of river pollutant discharge is an important means of water resource protection and management, and it is also a necessary condition for ensuring the normal functioning of water areas. The total amount of pollutant discharge is closely related to the water environmental capacity (WEC). Shifting from the traditional method of calculating WEC to dynamic analyses and calculations, concerning practical applications, in this paper, a dynamic adaptive calculation method is proposed for the river WEC that considers the changes in adaptive demand and hydrological conditions. In this method, the dynamic WEC is represented by intervals based on dynamic changes in different spatial and temporal scales, various calculation methods, hydrological conditions, and parameters. According to the calculation results for the WEC, a variable interval was formed. Taking the Shaanxi section of the main stream of the Wei River as the research object, with the support of an integrated platform, the dynamic adaptive calculation of the WEC in the Shaanxi section of the Wei River was realized, and a corresponding simulation system was constructed. The verification results show that (1) the dynamic calculation of WEC can be realized by freely combining different model methods and calculation conditions; (2) the WEC is described using a variable interval, which has strong applicability and operability; and (3) the simulation system can quickly adapt to the changing needs of practical applications and provide managers with visual and credible decision support. The research results provide a theoretical basis for river water environment pollution prevention and environmental management decision-making and help in the high-quality development of the river basin.

A genetic algorithm-based method to modulate the difficulty of serious games along consecutive robot-assisted therapy sessions

Background and Objective:One of the biggest challenges during neurorehabilitation therapies is finding an appropriate level of therapy intensity for each patient to ensure the recovery of movement of the affected limbs while maintaining motivation. Different studies have proposed adapting the difficulty of exercises based on psychophysiological state, based on success rate, or by modeling the user’s skills. However, all studies propose solutions for a single session, requiring a calibration process before using it in each session. We propose a dynamic adaptation method that can be used during different rehabilitation sessions, without the need for recalibration between sessions. Methods:The adaptation architecture is based on a genetic algorithm that aims to maintain a certain score level and to motivate the user to move. The method has been evaluated with two serious games for five sessions using a rehabilitation robot. A common initial evaluation was made for all the users involved in the study, and the game parameters that best suited each user from the previous session were introduced as the starting point of the next session. In addition, the desired score rate was lowered between sessions to increase the difficulty level. The psychophysiological state of the users was measured based on the Self-Assessment Manikin test, as well as different cardiorespiratory and galvanic skin response signals were analyzed. Results:The adaptation architecture proposed can find those game parameters that maximize the user movement for both games. In one of the games, the score rate set for each session is followed with high fidelity. The degree of personalization in the games increases between sessions as the dispersion of the game parameters grows. The Self-Assessment Manikin test and the physiological signals results would indicate that the psychophysiological state remains equal between sessions despite an increase in game difficulty. Conclusions:The genetic algorithm-based game adaptation has proven efficacy in maximizing the therapy performance through the sessions without needing recalibration. It also can be concluded that the design of the game influences the adaptation performance. Additionally, adaptive game design facilitated by our method does not significantly impact players’ emotional or physiological states.

A new method for unmanned aerial vehicle path planning in complex environments

To solve the problems of UAV path planning, such as low search efficiency, uneven path, and inability to adapt to unknown environments, this paper proposes A double-layer optimization A* and dynamic window method for UAV path planning. Firstly, the neighboring node clip-off rule is defined to optimize the node expansion mode of the A* algorithm, and the obstacle coverage model is designed to dynamically adjust the heurizing function of the A* algorithm to improve the path search efficiency. Then, the Bresenham algorithm is adopted for collision detection and critical path nodes are extracted to significantly reduce the number of path turning points. Secondly, a new tracking index is proposed to optimize the evaluation function of the dynamic window method to make the local path fit the global path further. By detecting the dangerous distance, the dynamic adaptive method of evaluation function weight is designed to improve the fixed weight of the dynamic window method. Finally, the key turning point of optimizing the A* algorithm is taken as the temporary target point to improve the DWA algorithm, and the local part follows the global part, and the fusion of the two algorithms is realized. Simulation results show that the proposed method can significantly improve the efficiency and smoothness of mobile robot path planning, enhance the real-time obstacle avoidance and adaptive ability of unknown environments, and better meet the requirements of complex planning tasks.

A Dynamic Adaptive and Resource-Allocated Selection Method Based on TOPSIS and VIKOR in Federated Learning

Federated learning (FL) is a decentralized and privacy-preserving machine learning technique that protects data privacy by learning models locally and not sharing datasets. However, due to limited computing resources on devices and highly heterogeneous data in practical situations, the training efficiency and resource utilization of federated learning is low. In order to resolve these challenges, we introduce a blockchain-assisted dynamic adaptive and personalized federated learning framework (TV-FedAvg) in the presence of restricted computing power resources and data heterogeneity. After each round of local training, we utilize an improved scoring model based on VIKOR and TOPSIS to comprehensively score the devices. The scores are then utilized to choose devices for participation in global aggregation and to carry out model aggregation through blockchain consensus. Furthermore, resources are reallocated for the next round to enhance resource efficiency, model fairness, and performance. Finally, we demonstrate through experimentation that TV-FedAvg outperforms other models such as pFedMe, FedAvg, Per-FedAvg, and TOPSIS in terms of both efficiency and performance.

Dynamic A-MPDU Adaptation Method for Airtime Fairness in Channel Bonding-Ready WLANs

Dynamic A-MPDU Adaptation Method for Airtime Fairness in Channel Bonding-Ready WLANs

Investigation on close-coupled gas atomization for Fe-based amorphous powder production via simulation and industrial trials: Part I. Melt breakup behaviors during primary atomization

The primary atomization process is simulated by Volume of Fluid (VOF) model and dynamic adaptive mesh method. The influence of melt mass flow rate and atomization pressure on the breakup process is investigated, and the effect of hot gas atomization is also evaluated. The results show that the breakup is insufficient when the melt mass flow rate is larger than 0.075 kg s−1, while the liquid film breakup inducing the nozzle clogging occurs when the melt mass flow rate is too low (0.025 kg s−1). The backflow of droplets occurs at low atomization pressure (1.0 MPa), and various defects (satellite powders, hollow powders and needle-shaped powders) appear when the atomization pressure is larger than 3.0 MPa. Although the breakup efficiency can be significantly improved by increasing the gas temperature, severe deformation of the gas-liquid interface is induced due to the Kelvin–Helmholtz wave and gas-liquid interaction, easily leading to the formation of hollow powders. Besides, the gas-to-melt ratio (GMR) is identified as a simple criterion for predicting primary atomization breakup modes, with liquid film breakup occurring when GMR≥4.4 and "fountain" breakup occurring when GMR≤4.3. In this work, not only the gas-liquid interaction is systematically analyzed by establishing a flow-heat transfer-VOF coupling model, but also the GMR is proposed to predict the breakup mode in the industrial production, which can provide theoretical and methodological guidance for the optimization of atomization operational parameters.

Evaluating the Usability of a Gaze-Adaptive Approach for Identifying and Comparing Raster Values between Multilayers

Raster maps provide intuitive visualizations of remote sensing data representing various phenomena on the Earth’s surface. Reading raster maps with intricate information requires a high cognitive workload, especially when it is necessary to identify and compare values between multiple layers. In traditional methods, users need to repeatedly move their mouse and switch their visual focus between the map content and legend to interpret various grid value meanings. Such methods are ineffective and may lead to the loss of visual context for users. In this research, we aim to explore the potential benefits and drawbacks of gaze-adaptive interactions when interpreting raster maps. We focus on the usability of the use of low-cost eye trackers on gaze-based interactions. We designed two gaze-adaptive methods, gaze fixed and gaze dynamic adaptations, for identifying and comparing raster values between multilayers. In both methods, the grid content of different layers is adaptively adjusted depending on the user’s visual focus. We then conducted a user experiment by comparing such adaptation methods with a mouse dynamic adaptation method and a traditional method. Thirty-one participants (n = 31) were asked to complete a series of single-layer identification and multilayer comparison tasks. The results indicated that although gaze interaction with adaptive legends confused participants in single-layer identification, it improved multilayer comparison efficiency and effectiveness. The gaze-adaptive approach was well received by the participants overall, but was also perceived to be distracting and insensitive. By analyzing the participants’ eye movement data, we found that different methods exhibited significant differences in visual behaviors. The results are helpful for gaze-driven adaptation research in (geo)visualization in the future.

Dynamic adaptive autonomous navigation decision-making method in traffic separation scheme waters: A case study for Chengshanjiao waters

A dynamic adaptive decision-making method in traffic separation scheme (TSS) waters (Chengshanjiao) was explored to achieve fully autonomous navigation. A static traffic environment comprising the digital models of the constituent parts of a TSS was established. A dynamic model was constructed using the target ships. Static and dynamic traffic environments serve as a foundation for decision-making. Based on the mathematical evaluations of the 1972 International Regulations for Preventing Collisions at Sea (COLREGs) and good seamanship, the requirements for motor-vessel collision avoidance in different scenarios and stages were presented. An automatic navigation method was established that could control the ship, automatically track the route, and assess the ship motion state. A collision risk identification model and dynamic collision avoidance mechanism for this type of water were built to identify the most dangerous target ship and a feasible manoeuvring plan to ensure safety in a multi-ship environment. To adapt to the random movement of target ships and correct residual errors, an autonomous navigation decision framework was proposed based on time-series rolling calculations. The ship safely moved through the water by altering its course 5° to starboard at 326 s, 6° to starboard at 887 s, route tracking at 1627 s, altering its course 6° to starboard at 4072 s, and route tracking at 4813 s in an experimental scenario with eight target ships. Six of the eight target ships were randomly manipulated. The results demonstrate that this decision-making technique can be utilised in practice and is precise to 1°.

Tangential stretching rate analysis of DRGEP-based automated target species selection dynamic adaptive chemistry method

Automated target species selection dynamic adaptive chemistry (automated DAC) based on the directed relation graph with error propagation is a dynamic mechanism reduction method that can alleviate the huge computational overhead of high-precision large eddy simulations. However, the accuracy of this algorithm is often assessed based on the temperature and ignition delay time, which are a posteriori results and do not reflect the physicochemical nature. Because of this, we proposed a new criterion based on the tangential stretching rate (TSR) value to study the automated DAC method from the perspective of chemical reactions, which can characterize the most energetic reaction mode. First, we tested the new criterion by applying it to the analyses of the simulations adopting the detailed mechanism and automated DAC. The results verified the performance of the new criterion and showed that a new TSR value oscillation error phenomenon was discovered by it. This error is caused by the inappropriate cut-off of the important species and chemical reactions in the DAC-generate skeleton mechanisms. Second, we studied the measures to eliminate this newly emerged error. We found that reducing the search threshold, increasing the number of target species of the automated DAC method, and modifying the automated DAC to the time-correlated automated DAC by adding a time correlation term can alleviate this TSR value oscillation error. In addition, constant pressure autoignition simulations of methane/air were implemented to support this study. The TSR-computational singular perturbation participation indices were also introduced for assistance.

Research on autonomous dynamic real-time capture of robot based on hybrid prediction model

The trajectory of the dynamic target is an important basis for the robot to capture the action, and its advanced prediction precision plays a decisive role in achieving efficient capture. In this paper, a dynamic adaptive weighting method is proposed to establish a hybrid prediction model of GM (1,1) and SESM. The weights are allocated in real time, according to the prediction accuracy to improve the prediction accuracy of a single model and enhance the applicability of spatial trajectory prediction. Then, based on the rolling prediction principle, a high-precision advanced prediction method of the hybrid model is created. Compared with the advanced prediction of the model itself, the average prediction errors of the spatial arc and straight line verification trajectories are reduced by 40.26% and 49.77% respectively. The simulation model of robot dynamic target capture is built in Matlab, and the measured trajectory is used as the target trajectory. The robot capture experiment platform is built, and the target motion trajectory is collected by the camera in real time to capture the experiment, which verifies the effectiveness of the proposed method in actual capture.

Dynamic Adaptive Reconfiguration Method for Cloud Manufacturing Service Composition in Practical Multi-constraint Environment

Dynamic Adaptive Reconfiguration Method for Cloud Manufacturing Service Composition in Practical Multi-constraint Environment

Breakup process modeling and production of FeSiAl magnetic powders by close-coupled gas atomization

Close-coupled gas atomization is extensively used in the industrial production of alloy powders due to the advantages of high efficiency, low cost and excellent powder properties. In order to investigate the complex gas-liquid interaction during atomization, primary atomization was simulated through the VOF (Volume of Fluid) model and the dynamic adaptive mesh method, and a two-way coupling DPM (Discrete Particle Model) was used to simulate secondary atomization. The results showed that under the low melt mass flow rate of 0.0050 kg s−1, the flow separation and droplet backflow occurred at the atomization pressure of 1 MPa, while gas-liquid interfaces deformed and rotated drastically when the atomization pressure is above 3 MPa, which led to the formation of hollow powders. When the melt mass flow rate was 0.0150 kg s−1, the d50 decreased from 40.7 to 38.4 μm as the atomization pressure increased from 2 to 3 MPa. However, the chance of collision between particles increased due to the decrease in the spray cone angle, and both the standard deviation of particle size and atomization efficiency deteriorated, which indicated that 2 MPa was the optimum atomization pressure. As the mass flow rate increased from 0.0050 to 0.0150 kg s−1, the gas to melt ratio decreased, but the d50 only increased from 32.3 to 40.7 μm, indicating that the mass flow rate of 0.0150 kg s−1 did not significantly affect the particle size distribution, but also contributed to improve the production efficiency and avoided the risk of the nozzle freeze-off. The d50 of the spherical FeSiAl powders produced by the industrial tests at 2 MPa and 0.0150 kg s−1 was 44.7 ± 2.1 μm, which was basically consistent with the simulation results, indicating that the parameter optimization by the coupling model developed in this study could provide theoretical and practical guidance for the preparation of high-performance alloy powders by gas atomization.

A novel joint distinct subspace learning and dynamic distribution adaptation method for fault transfer diagnosis

Domain adaptation (DA) have achieved phased results in fault transfer diagnosis. However, there is still no unified framework that not only minimize geometrical shift but also minimize distribution shift of source and target data. Besides, existing DA approaches only exploit shared features between domains, which will degenerate when the two domains have a big shift, since a common subspace may not exist. In this paper, a novel joint distinct subspace learning and dynamic distribution adaptation (JDSDDA) method is proposed for rolling bearing fault transfer diagnosis. JDSDDA seeks two coupled projections to map source and target data into distinct subspaces, while adjusting the relative significance of marginal and conditional distributions by an adaptive weight factor. Finally, JDSDDA learns an adaptive classifier by optimizing structural risk minimization (SRM), distinct subspace alignment, dynamic distribution adaptation simultaneously. The experimental results demonstrates that the JDSDDA has significant superiority compared with other DA approaches.

When to transfer: a dynamic domain adaptation method for effective knowledge transfer

When to transfer: a dynamic domain adaptation method for effective knowledge transfer

Development and verification of a high-speed compressible reactive flow solver in OpenFOAM

This paper proposes a newly developed density-based compressible solver called DCRFoam on a basis of OpenFOAM platform for simulating detonation phenomena with detailed chemical kinetics. It employs AUSM+-up scheme to determine the convective fluxes and a four steps low storage Runge-Kutta time integration method for time integration. A total variation diminishing (TVD) and symmetrical flux limiter sgva is implemented for interpolation process. Also, the dynamic mesh adaptation method is applied to refine or coarsen the computational mesh dynamically since the resolution requirements of supersonic combustion are localized. The numerical solver has been validated systematically with many numerical and experimental results reported in the literature, including those from non-reacting cases (Sod’s shock tube problem, two-dimensional shock/cylinder interaction) to reacting flow cases (the ignition sequence in a shock tube, oblique detonation wave (ODW) initiating process and detonation propagating in a bifurcated channel). The results obtained by current schemes are compared with original central upwind Kurganov-Noelle-Petrova (KNP) scheme in official OpenFOAM. The corresponding result demonstrates the robustness and accuracy of the current solver.

Research on occlusion block face recognition based on feature point location

AbstractAiming at improving the poor face recognition accuracy under occlusion, a multipose block occlusion face recognition method based on feature point location is proposed. Face segmentation is carried out according to the location results of face feature points and occlusion areas. The local features of each face block are extracted by deep convolution neural network. The dynamic adaptive weighting method is used to give different weights to the face block information, and the occluded face recognition is completed according to the results of face segmentation, which effectively reduces the impact of pose change and occlusion on face recognition. The experiment is analyzed from two aspects: frontal occlusion and multipose face occlusion. The results show that fewer blocks in the frontal occlusion experiment are conducive to maintaining good local integrity and relatively good recognition performance; When the proportion of frontal occlusion reaches 50%, the recognition rate of our algorithm can still be up to 92.68%, which is significantly better than other algorithms.

Deep multiple auto-encoder with attention mechanism network: A dynamic domain adaptation method for rotary machine fault diagnosis under different working conditions

Intelligent fault diagnosis methods based on deep learning have achieved noteworthy application results in health diagnosis of rotating machinery. However, the fault data distribution discrepancy caused by different working conditions in real industrial scenarios can deteriorate the diagnosis performance of model. And extracted features by multi-ensemble deep auto-encoder neglect the contribution degree of each deep auto-encoder. Inspire by the demands, a dynamic domain adaptation method based on deep multiple auto-encoders with attention mechanism (DMAEAM-DDA) is proposed for rotary machine fault diagnosis under different working conditions. Firstly, combined with attention mechanism, pre-trained multiple deep auto-encoder with six different activation functions are utilized to construct deep multiple auto-encoder with attention mechanism network for extracting feature. Then the dynamic domain factor is calculated to automatically assign the weight of the marginal and conditional distribution for learning domain invariant fault features. Finally, two rotary machine experiments are employed to verify the availability of the proposed DMAEAM-DDA method, and the results show the proposed DMAEAM-DDA method has better superiority and outstanding stability compared to other methods.

College English Teaching Quality Monitoring and Intelligent Analysis Based on Internet of Things Technology

With the continuous development of network communication, modern control, information perception, artificial intelligence, and other technologies, the Internet of Things technology has also been more comprehensively developed and grown, and it is playing an increasingly important role in various fields of production and life. Due to the low efficiency and cumbersome procedures of the existing college English teaching quality assessment, this study intends to propose a new method to deal with this problem. Through this research, we get the following: (1) use static weight equalization method, dynamic weight adaptation method, and iterative weight method to filter and analyze the original data received by mobile phones and discard some obviously wrong values in the preprocessing. Analyzing the number requirements of the game model, the supply operation reference model, and the IoT technology upgrade model for the number of demand points, the results show that the IoT technology upgrade model requires the least number of demand points. (2) Optimize the related parameter information of the IoT system model, including candidate numbers, number of students in class, adjustment coefficient, Di, turnover rate, system cost, and total system consumption. The Internet of Things technology upgrade model performed best. Adjustment coefficient = − 102 ; Di = − 99.6 ; turnover = − 100 ; system cost = − 123.2 ; SL = − 123.5 ; SP = − 99.8 . (3) Passed the test of the number of questions and feedback rate, classroom number, classroom test scores, new vocabulary number, system scoring, and comprehensive scoring to judge the model. It is found that the number of questions, feedback rate, class number, and class test scores, number of new vocabularies, system scores, and comprehensive scores of the upgraded and transformed models of the Internet of Things are above the standard values, indicating that the model is highly efficient and intelligent.