Low Computational Cost Research Articles

Learning to Guide Particle Search for Dynamic Multiobjective Optimization.

Dynamic multiobjective optimization problems (DMOPs) are characterized by multiple objectives that change over time in varying environments. More specifically, environmental changes can be described as various dynamics. However, it is difficult for existing dynamic multiobjective algorithms (DMOAs) to handle DMOPs due to their inability to learn in different environments to guide the search. Besides, solving DMOPs is typically an online task, requiring low computational cost of a DMOA. To address the above challenges, we propose a particle search guidance network (PSGN), capable of directing individuals' search actions, including learning target selection and acceleration coefficient control. PSGN can learn the actions that should be taken in each environment through rewarding or punishing the network by reinforcement learning. Thus, PSGN is capable of tackling DMOPs of various dynamics. Additionally, we efficiently adjust PSGN hidden nodes and update the output weights in an incremental learning way, enabling PSGN to direct particle search at a low computational cost. We compare the proposed PSGN with seven state-of-the-art algorithms, and the excellent performance of PSGN verifies that it can handle DMOPs of various dynamics in a computationally very efficient way.

Development of a multi-fidelity optimisation strategy based on hybrid methods

A multi-fidelity optimisation strategy has been developed in the present work, and its performance is illustrated through a series of test cases. The strategy is based on hybrid methods such that two genetic optimisation algorithms are employed, each coupled to a different fidelity level with transfer of information between them. The aim is that the low fidelity model, being less accurate but with a lower computational cost, performs a comprehensive search along the design space guiding the high fidelity model to the optimum region. This strategy has been shown to reduce the computational time of an optimisation through analytical test cases as well as numerical cases. The analytical cases have been used to tune the parameters that define the multi-fidelity strategy, while the numerical cases are employed to apply the method to engineering problems, focusing on the aerodynamic performance of an airfoil. The speed-up shows a certain dependency to the models relation, both regarding their similarity level as well as the relative computational cost. For cases exhibiting a significant dissimilarity between models, wherein the low fidelity model is notably inaccurate, the attained speed-up diminishes, and numerous instances demonstrate an absence of speed-up. However, for most cases, even with poor model similarity the optimisations are accelerated by an order of 2, while values up to 3–5 were found for higher similarity levels. Hence, the developed strategy shows a relevant decrease of computational cost of an optimisation procedure although its performance is affected by the models relative accuracy.

Estimating user influence in social networks under independent cascade model

The rapid increase in social applications emphasizes the importance of estimating user influence. Heuristic methods like degree and betweenness centralities usually differ from the actual propagation process and yield unsatisfactory results. Traditional methods like Monte Carlo simulation are time-consuming. We modify the duplicate forwarding model to analyze the propagation process, which is proved to be close to the independent cascade model. We calculate the influence of a given source on a target. This approach allows for relatively accurate user influence estimation. Although this method is more efficient than traditional methods, it still requires traversing all users. Therefore, we introduce a virtual user who is connected to all users. By estimating the influence of any user on the virtual user, we can approximate the user influence efficiently. Experiments on real-world networks demonstrate that our method achieves not only better accuracy in user influence ranking but also lower computational cost.

Resilient Neuroadaptive Distributed Fixed-Time Attitude Coordination Control for Multiple Spacecraft.

This work studies the attitude coordination tracking problem for multiple spacecraft with consideration of unintended faults (communication link faults and actuator faults), inertial uncertainties, and external disturbances under a directed communication graph. A resilient neuroadaptive distributed fixed-time control scheme is investigated to solve this challenging problem. First, an improved adaptive distributed observer is established for followers to estimate the states of the leader when considering communication link faults. The proposed observer improves the resilience against communication link faults. Subsequently, to further cope with the problem of actuator faults, inertial uncertainties, and external disturbances, based on the proposed observer and the technique of adding a power integrator, a neuroadaptive distributed fixed-time attitude coordination controller is developed. Unlike the existing controllers, the proposed one requires less information when dealing with faults and lumped uncertainties, and has a lower-computational cost. Moreover, the fixed-time stability of the closed-loop system is ensured under the designed resilient neuroadaptive distributed control scheme. Finally, comparative simulations are carried out to manifest the effectiveness of the investigated coordination control method.

Cone-Traced Supersampling With Subpixel Edge Reconstruction.

While signed distance fields (SDFs) in theory offer infinite level of detail, they are typically rendered using the sphere tracing algorithm at finite resolutions, which causes the common rasterized image synthesis problem of aliasing. Most existing optimized antialiasing solutions rely on polygon mesh representations; SDF-based geometry can only be directly antialiased with the computationally expensive supersampling or with post-processing filters that may produce undesirable blurriness and ghosting. In this work, we present cone-traced supersampling (CTSS), an efficient and robust spatial antialiasing solution that naturally complements the sphere tracing algorithm, does not require casting additional rays per pixel or offline pre-filtering, and can be easily implemented in existing real-time SDF renderers. CTSS performs supersampling along the traced ray near surfaces with partial visibility - object contours - identified by evaluating cone intersections within a pixel's view frustum. We further introduce subpixel edge reconstruction (SER), a technique that extends CTSS to locate and resolve complex pixels with geometric edges in relatively flat regions, which are otherwise undetected by cone intersections. Our combined solution relies on a specialized sampling strategy to minimize the number of shading computations and correlates sample visibility to aggregate the samples. With comparable antialiasing quality at significantly lower computational cost, CTSS is a reliable practical alternative to conventional supersampling.

Attention Retinex Network(A4R-Net) for face detection under low-light environment

The degradation of recognition rates in low-light environments is a critical issue in terms of security when using object and face recognition technologies in various locations. Existing low-light enhancement models have shown limitations in terms of computational cost and performance. However, this paper overcomes these limitations. The experimental results demonstrate that our model achieves the same performance as existing models with 13 times lower computational cost and a face detection performance of 82.2%.The structure of this paper is as follows: Introduction, which provides the background and explains the limitations of existing models. Proposed Method, which details the structure and working principles of A4R-Net. Experimental Results, which present the evaluation of low-light enhancement performance and the comparison of face detection using YOLOv4 [1]. Conclusion, which discusses the contributions of this research and future research directions.The source code and dataset is https://github.com/Obiru2698/obiru2698.github.io/

Review of Recent Additive Manufacturing and Welding Research with Application of Physics-Informed Neural Networks

This review introduces recent research on applying physics-informed neural networks (PINNs) to additive manufacturing and welding. PINNs, which are artificial intelligence models, integrate governing equations containing physical information with artificial neural networks, enabling the modeling of complex physical phenomena at a lower computational cost than traditional numerical models. Although PINNs have been employed in a limited number of studies on welding processes, they have been extensively used in various studies within the field of additive manufacturing. This study reviews the theoretical background of PINNs to explore their effective application to welding processes, examining 12 research cases in additive manufacturing and two research cases in welding processes. The analysis included the structure of the PINN, governing equations, and prediction results of each study. Results indicate that PINNs provide faster computation speeds and higher prediction accuracies than numerical models. Moreover, they could perform analyses without additional training even when process parameters and materials changed. Additionally, PINNs have been effectively applied to predict the mechanical properties of the molten zone. Consequently, PINNs are anticipated to be actively applied in future research on welding process modeling and mechanical property prediction.

Image deraining via invertible disentangled representations

Photos taken on rainy days usually suffer from a loss of background image information due to obstruction by rain streaks and the mist-like effect that rain creates. Recent Invertible Neural Networks (INNs) have shown promise in image deraining, as they are capable of extracting observed features without information loss for the purpose of restoration. However, it remains a challenge for INNs to directly learn the transformation from a rainy image to its rain-free version, given that rain streaks exhibit a diversity in transparency, shape, and brightness, leading to a complex distribution of lost details. To solve the problem, we design a novel Invertible Disentangling network (InvDis) to decompose the complex data distribution into simpler ones and makes the transformation much easier to learning. To be specific, InvDis separates feature channels of rain streaks and those of the polluted image details in the forward mapping. By only discarding rain channels and preserving polluted image details, in the backward mapping, InvDis focuses on generating missing details obstructed by rain streaks with a new sample from a prior distribution, and restores clean details from those polluted ones. Additionally, InvDis incorporates a channel interaction mechanism to facilitate the disentangling. It allows the rain streaks encoded in other channels to easily flow to those discarded channels, and conversely, allows the polluted image details to be directed towards the preserved channels. Experimental results on both synthetic and real-world datasets show that the InvDis not only improves the restoration quality but also has lower computational costs.

Localization algorithm of soil moisture monitoring in irrigation area based on weighted correction of distance measurement

At present, Wireless Sensor Network technology are widely used in the fields of agricultural environment monitoring. Whether we use ground network or ground-air coordination, the position coordinates of unknown nodes are an important feature of sensor information collection. To achieve the goal of a wireless monitoring system for field soil irrigation and its node positioning, we proposed an RSSI (Received Signal Intensity Indication)-based agricultural environment irrigation monitoring system. The algorithm has three stages: RSSI ranging, error correction, and localization. In the RSSI ranging phase, we obtain inter-node measurement distance by wireless channel modeling. In the distance-weighted correction phase, considering the difference between the measured distance of the beacon nodes and the actual distance, we analyze and select the relative error coefficient for the beacon nodes to correct the measured distance between the unknown nodes and the beacon nodes within their communication range. In the positioning stage for the nodes to be located, we estimated the required node coordinates using a weighted centroid localization method. In addition, by analyzing the monitoring data, we know the changes in soil moisture in real-time, which provides new ideas for irrigation automation and the efficient utilization of water resources. When designing the algorithm, we thoroughly considered the influence of RSSI ranging inaccuracy and the quantity of beacon nodes on the localization precision. The test demonstrated that the method presented in this paper has higher localization precision and lower computation cost.

TabSAL: Synthesizing Tabular data with Small agent Assisted Language models

Tabular data are widely used in machine-learning tasks because of their prevalence in various fields; however, the potential risks of data breaches in tabular data and privacy protection regulations render such data almost unavailable. Tabular data generation methods alleviate data unavailability by synthesizing privacy-free data, and generating data using language models is a novel innovation. Language models can synthesize high-quality datasets by learning knowledge from nondestructive information and recognizing the semantics of table columns. However, when current language models function as generators, their encoding methods are hindered by complicated decoding processes, and the limited predictive ability of language models restricts their generative capability. To this end, we propose an encoding method based on interactive data structures such as JavaScript Object Notation for converting tabular data. We design TabSAL, which is a pluggable tabular data generation framework with small agent assisted language models, to boost the predictive capability, resulting in high-quality synthetic datasets with a much lower computational resource cost. In addition, a benchmark that integrates eight datasets, three methods, and three assessment directions has been issued, which indicates that TabSAL surpasses the state of the art by up to 60%.

Computer vision in the food industry: Accurate, real-time, and automatic food recognition with pretrained MobileNetV2

In contemporary society, the application of artificial intelligence for automatic food recognition offers substantial potential for nutrition tracking, reducing food waste, and enhancing productivity in food production and consumption scenarios. Cutting-edge technologies such as Computer Vision and Deep Learning are highly beneficial, enabling machines to learn automatically, thereby facilitating automatic visual recognition. Despite some research in this field, the challenge of achieving accurate automatic food recognition quickly remains a significant research gap. Some models have been developed and implemented, but maintaining high performance swiftly, with low computational cost and low access to expensive hardware accelerators, is still an area that needs further exploration and improvement. This study employs the pre-trained MobileNetV2 model, which is cost-efficient and fast, for automatic food recognition on the public Food11 dataset, comprising 16643 images. It also utilizes various tools and techniques such as dataset understanding, transfer learning, data augmentation, regularization, dynamic learning rate, hyperparameter tuning, and consideration of images in different sizes to cut time and enhance the model’s performance and robustness. These techniques aid in choosing appropriate metrics, achieving better performance, avoiding overfitting and accuracy fluctuations, speeding up the model, and increasing the generalization of findings, making the study and its results applicable to real-world applications. Despite MobileNetV2′s simpler structure and fewer trainable parameters compared to some deep-dense models in the deep learning area, it quickly achieved approximately 92.97 % accuracy. This underscores the model’s potential for implementation in real-world applications, which is the main intention of this study.

Learned infinite elements for helioseismology. Learning transparent boundary conditions for the solar atmosphere

Acoustic waves in the Sun are affected by the atmospheric layers, but this region is often ignored in forward models because it increases the computational cost. The purpose of this work is to take the solar atmosphere into account without significantly increasing the computational cost. We solved a scalar-wave equation that describes the propagation of acoustic modes inside the Sun using a finite-element method. The boundary conditions used to truncate the computational domain were learned from the Dirichlet-to-Neumann operator, that is, the relation between the solution and its normal derivative at the computational boundary. These boundary conditions may be applied at any height above which the background medium is assumed to be radially symmetric. We show that learned infinite elements lead to a numerical accuracy similar to the accuracy that is obtained for a traditional radiation boundary condition in a simple atmospheric model. The main advantage of learned infinite elements is that they reproduce the solution for any radially symmetric atmosphere to a very good accuracy at low computational cost. In particular, when the boundary condition is applied directly at the surface instead of at the end of the photosphere, the computational cost is reduced by 20<!PCT!> in 2D and by 60<!PCT!> in 3D. This reduction reaches 70<!PCT!> in 2D and 200<!PCT!> in 3D when the computational domain includes the atmosphere. We emphasize the importance of including atmospheric layers in helioseismology and propose a computationally efficient method to do this.

Development, characterization and validation of a novel physics-informed equivalent circuit model for silicon–graphite battery cells

The widespread deployment of electric vehicles (EVs), as well as the growing requirements both from manufacturers and consumers, necessitate an increase in energy density with regard to current lithium-ion batteries (LIBs). In this regard, one of the most promising alternatives is the addition of silicon to conventional graphite anodes, thus giving rise to energy-dense LIBs. However, this entails a significant increment in modeling, parameterization and computational complexities due to the interactions between both negative electrode materials. In this article, we present a novel physics-informed equivalent circuit model for cells with blended electrodes, which is able to provide accurate results with respect to a state-of-the-art electrochemical model, not only for the output voltage (≤25 mV RMS), but also internal states, namely active material particle surface concentrations (≤1.2 %) and current distribution (≤2.6 %) at a much lower computational cost. Furthermore, this formulation also allows for a notable simplification of model parameterization. Hence, we describe thorough static and dynamic parameterization processes from half-cell experimental data and impedance measurements. Finally, we validate the proposed model in constant-current as well as dynamic operation, highlighting the influence that the choice of hysteresis model has on the latter. We understand that the presented model is an interesting alternative for the modeling of cells with blended electrodes, and is also suitable for its implementation in Battery Management Systems (BMSs) in EVs.

Fast and accurate modelling of twin-screw compressors: A generalised low-order approach

Twin-screw compressors (TSC) are commonly used in heat pump processes due to their robustness and flexibility. They exhibit two core properties, i.e. the swept volume and the built-in volume ratio (BVR), which heavily influence their capacity limits and off-design efficiency. This work presents a new low-order model (i.e. a polynomial model), which can accurately predict a TSC’s behaviour. The model uses the external pressure ratio and volumetric compressor inlet flow rate to calculate isentropic efficiency and compressor speed. The input parameters are normalised with a reference flow rate (calculated from the swept volume) and the BVR, respectively. This results in a generalised model of low numerical cost, which can be used for explorative studies independent of the specific machine size and BVR. A gain in computational speed by a factor of 375 is achieved compared to a semi-empiric reference model. The model displays very good predictive accuracy when used to predict the performance of machines with similar BVRs, but different sizes. A mean deviation from the manufacturer data of 4.29%, 0.88°C and 1.38% for the shaft power, the outlet temperature and the compressor speed can be observed, respectively. When there is a difference in size and BVR, the prediction accuracy is still reasonable but significantly declines for small and very large pressure ratios. Nevertheless, the proposed new approach extends the state-of-the-art by introducing a low-order model, which combines the advantages of low computational cost, high accuracy, physically correct predictions over a wide operational range and scalability to different machine capacities and BVRs. The validation for different fluids indicates a good general prediction accuracy relatively independent of the used fluid.

Comparison of Optimization Methods for the Attitude Control of Satellites

The definition of multiple operational modes in a satellite is of vital importance for the adaptation of the satellite to the operational demands of the mission and environmental conditions. In this work, three optimization methods were implemented for the initial calibration of an attitude controller based on fuzzy logic with the purpose of performing an initial exploration of optimal regions of the design space: a multi-objective genetic algorithm (GAMULTIOBJ), a particle swarm optimization (PSO), and a multi-objective particle swarm optimization (MOPSO). The performance of the optimizers was compared in terms of energy cost, accuracy, computational cost, and convergence capabilities of each algorithm. The results show that the PSO algorithm demonstrated superior computational efficiency compared to the others. Concerning the exploration of optimum regions, all algorithms exhibited similar exploratory capabilities. PSO’s low computational cost allowed for thorough scanning of specific interest regions, making it ideal for detailed exploration, whereas MOPSO and GAMULTIOBJ provided more balanced performance with constrained Pareto front elements.

Incipient instability real-time warning via adaptive wavelet synchrosqueezed transform: Onboard applications from compressors to gas turbine engines

To address the limitations of existing stall/surge warning methods—including variations in research subjects, sensor configurations, computational complexity, and the universality of thresholds—this study proposes a novel incipient instability warning approach utilizing adaptive wavelet synchrosqueezed transform. Each detection cycle begins with the acquisition of a low-pass filtered and downsampled total pressure signal from the compressor outlet, captured via a sliding window to represent the pressure sample. Then, the adaptive wavelet synchrosqueezed transform (AWSST), employing dynamic discretization with scales and synchrosqueezing technology, analyzes the real-time pressure sample to extract precise instantaneous frequency changes at low computational costs. Features of instability severity are derived from the amplitude, phase, and pressure drop within the time-frequency spectrum, leading to the formulation of a warning logic. The proposed method's effectiveness and universality are validated through hardware-in-loop (HIL) tests using multiple surge test data. Results show that the proposed method outperforms traditional real-time warning methods, which rely on the pressure change rate, by enhancing accuracy and advancing warning times by 15 ms–50 ms. Notably, the threshold is universal for compressors or engines of the same model. By increasing the acceptable performance decline criteria, compressors, turbofan engines, and turboshaft engines can employ a common set of thresholds.

Determining the Optimal Window Duration to Enhance Emotion Recognition Based on Galvanic Skin Response and Photoplethysmography Signals

Automatic emotion recognition using portable sensors is gaining attention due to its potential use in real-life scenarios. Existing studies have not explored Galvanic Skin Response and Photoplethysmography sensors exclusively for emotion recognition using nonlinear features with machine learning (ML) classifiers such as Random Forest, Support Vector Machine, Gradient Boosting Machine, K-Nearest Neighbor, and Decision Tree. In this study, we proposed a genuine window sensitivity analysis on a continuous annotation dataset to determine the window duration and percentage of overlap that optimize the classification performance using ML algorithms and nonlinear features, namely, Lyapunov Exponent, Approximate Entropy, and Poincaré indices. We found an optimum window duration of 3 s with 50% overlap and achieved accuracies of 0.75 and 0.74 for both arousal and valence, respectively. In addition, we proposed a Strong Labeling Scheme that kept only the extreme values of the labels, which raised the accuracy score to 0.94 for arousal. Under certain conditions mentioned, traditional ML models offer a good compromise between performance and low computational cost. Our results suggest that well-known ML algorithms can still contribute to the field of emotion recognition, provided that window duration, overlap percentage, and nonlinear features are carefully selected.

Investigations of Functional Groups Effect on CO2 Adsorption on Pillar[5]arenes Using Density Functional Theory Calculations

AbstractThis study investigates how various functional groups affect the adsorption of carbon dioxide (CO2) in pillar[5]arenes (P[5]A) at the aim of enhancing CO2 capture in P[5]A for environmental applications. Density functional theory (DFT) and density‐functional‐based tight‐binding (DFTB) methods were employed to investigate the adsorption behavior of CO2 in pillar[5]arenes (P[5]A) with various functional groups including P[5]A with methyl (P[5]A‐OCH3), aldehyde (P[5]A‐OCOH) functional groups and pillar[5]quinone (P[5]Q). Self‐consistent charge DFTB calculations have been performed using the DFTB+ code, with dispersion corrections included for van der Waals interactions, to ensure accurate results at a lower computational cost. Bader charge analysis have been carried out with VASP code to understand the interaction mechanisms and charge transfer between P[5]A and CO2. The study demonstrates that changing the functional groups of P[5]A affects CO2 adsorption behavior. Specifically, CO2 adsorption is more favorable at the cavity site of P[5]A‐OCH3 compared to other functional groups, due to the combined effects of weak hydrogen bonds and π‐π interactions. The calculated results shows that the functionalization is one of the beneficial factors for CO2 capture in pillararenes, the adsorption energy and configuration of adsorbed CO2 at P[5]A are affected by both the polarization and geometry of the functional groups. Furthermore, the density of states (DOS) calculations confirm that CO2 is only physisorbed in P[5]A, ensuring the reusability of pillararenes after desorption. The findings of this study suggest that modifying the functional groups of pillararens is a promising strategy for developing high‐efficiency CO2 capture materials.

Evaluation of Hand-Crafted Feature Extraction for Fault Diagnosis in Rotating Machinery: A Survey.

This article presents a comprehensive collection of formulas and calculations for hand-crafted feature extraction of condition monitoring signals. The documented features include 123 for the time domain and 46 for the frequency domain. Furthermore, a machine learning-based methodology is presented to evaluate the performance of features in fault classification tasks using seven data sets of different rotating machines. The evaluation methodology involves using seven ranking methods to select the best ten hand-crafted features per method for each database, to be subsequently evaluated by three types of classifiers. This process is applied exhaustively by evaluation groups, combining our databases with an external benchmark. A summary table of the performance results of the classifiers is also presented, including the percentage of classification and the number of features required to achieve that value. Through graphic resources, it has been possible to show the prevalence of certain features over others, how they are associated with the database, and the order of importance assigned by the ranking methods. In the same way, finding which features have the highest appearance percentages for each database in all experiments has been possible. The results suggest that hand-crafted feature extraction is an effective technique with low computational cost and high interpretability for fault identification and diagnosis.

A novel multi-objective generative design approach for sustainable building using multi-task learning (ANN) integration

Building Performance Optimization (BPO) plays a pivotal role in enhancing building performance, guaranteeing comfort while reducing resource consumption. Existing performance-driven generative design is computational demanding and difficult to be generalized to other similar buildings with difficult to be generalized to other building types or climate conditions. To fill this gap, this paper introduces a novel framework, which integrates multitask learning (MTL), code compliance check, and multi-objective optimization through NSGA-III algorithm. This framework is able to identify Paratoo Optimal design solutions, which comply with building codes, at low computation costs. The framework begins with selecting key design variables that are critical to building energy, comfort performance and life cycle cost. It then employs MTL to enhance the model's accuracy while reducing computational costs. Next, we designed a code compliance checking module followed by the NSGA-III optimization process, with the objective of identifying solutions that comply with existing building codes. The results indicate that the proposed MTL network achieved an R2 score of 0.983–0.993 on the test set. In the particular case study where equal weights are preferred, this approach yielded noteworthy reductions of 27.65%, 19.55%, and 31.13% in Building Energy Consumption (BEC), Life Cycle Cost (LCC), and Residue of continuous Daylight Autonomy (RcDA), respectively, for a rural dwelling, and exclude solutions that fail to satisfy regulatory standards. This framework allows designer to input the weights of each objective based on their preference and can be applied to other building types and climate regions. Last, we develop a solution selection tool based on the results output by the framework we proposed, which can be found at https://github.com/LiMingchen159/Village-House-Design-Strategy-in-Hebei-Province-China.