Two-stage Method Research Articles

Urban Road Surface Condition Sensing from Crowd-Sourced Trajectories Based on the Detecting and Clustering Framework.

Roads play a crucial role in urban transportation by facilitating the movement of materials within a city. The condition of road surfaces, such as damage and road facilities, directly affects traffic flow and influences decisions related to urban transportation maintenance and planning. To gather this information, we propose the Detecting and Clustering Framework for sensing road surface conditions based on crowd-sourced trajectories, utilizing various sensors (GPS, orientation sensors, and accelerometers) found in smartphones. Initially, smartphones are placed randomly during users' travels on the road to record the road surface conditions. Then, spatial transformations are applied to the accelerometer data based on attitude readings, and heading angles are computed to store movement information. Next, the feature encoding process operates on spatially adjusted accelerations using the wavelet scattering transformation. The resulting encoding results are then input into the designed LSTM neural network to extract bump features of the road surface (BFRSs). Finally, the BFRSs are represented and integrated using the proposed two-stage clustering method, considering distances and directions. Additionally, this procedure is also applied to crowd-sourced trajectories, and the road surface condition is computed and visualized on a map. Moreover, this method can provide valuable insights for urban road maintenance and planning, with significant practical applications.

Environmental effects of waste electrical and electronic equipment valorization in burkina faso

The valorization of waste electrical and electronic equipment (WEEE) has become a challenge for countries around the world, especially for developing countries like Burkina Faso. The main objective of this article is to analyze the environmental effects of WEEE valorization in Burkina Faso. To do this, the regime-switching model, developed by Lee (1978), based on the Heckman correction model, was used. The data used was collected from 414 actors in the WEEE sector in the three major cities of Burkina Faso, Ouagadougou, Koudougou, and Bobo Dioulasso. The analysis of the results shows that the methods of WEEE valorization in Burkina Faso pollute the air, soil and water. This pollution results in damages amounting to 242,752,192.6 CFA francs per year, which the sector's actors should include in their investment costs in the form of internalization costs. These costs amount to an average of 560,628.62 CFA francs per year and per actor. The WEEE sector could be improved with the implementation of certain economic policies. These policies include raising awareness among the population on WEEE management, training of stakeholders in the sector on the effective valorization of WEEE and the implementation of strict environmental standards to reduce environmental pollution due to poor WEEE valorization in Burkina Faso.

A Simplified Theoretical Method for Estimating the Effect of Capsule Grouting Expansion Technology on Controlling Pile Deformation

Deep excavation and shield tunneling processes often induce significant deformation in adjacent existing pile foundations, posing risks to the safety of superstructures. To address this issue, capsule grouting expansion technology has emerged as an efficient and economical technology to actively control pile deformation during such construction activities. This technology has been applied in some engineering projects aimed at safeguarding nearby buildings or facilities. However, theoretical research on capsule grouting expansion technology is limited and lags behind its practical engineering application. This paper proposed a simplified theoretical approach for estimating the effect of capsule grouting expansion technology on the deformation of neighboring pile foundations. Initially, virtual image technology was employed to obtain an analytical solution for the lateral displacement of soil resulting from the expansion of a cylindrical cavity with infinite height. The pile was considered as an Euler–Bernoulli beam on a Winkler foundation, allowing for the calculation of the loading response of the existing pile foundation. Utilizing the two-stage method, the theoretical analytical method of the lateral deformation of the pile shaft caused by capsule expansion was obtained. The reliability of the proposed theoretical method was verified by comparing the measured and calculated results. Further investigation into the controlling effect of the capsule grouting expansion technology on the deformation of the pile foundation was conducted through parametric analyses. These analyses encompass capsule radius, the distance between the capsule and pile foundation, the depth of the capsule, the length of the capsule, and the pile base constraints.

Computer-aided diagnosis of cystic lung diseases using CT scans and deep learning.

Auxiliary diagnosis of different types of cystic lung diseases (CLDs) is important in the clinic and is instrumental in facilitating early and specific treatments. Current clinical methods heavily depend on accumulated experience, restricting their applicability in regions with less developed medical resources. Thus, how to realize the computer-aided diagnosis of CLDs is of great clinical value. This work proposed a deep learning-based method for automatically segmenting the lung parenchyma in computed tomography (CT) slice images and accurately diagnosing the CLDs using CT scans. A two-stage deep learning method was proposed for the automatic classification of normal cases and five different CLDs using CT scans. Lung parenchyma segmentation is the foundation of CT image analysis and auxiliary diagnosis. To meet the requirements of different sizes of the lung parenchyma, an adaptive region-growing and improved U-Net model was employed for mask acquisition and automatic segmentation. The former was achieved by a self-designed adaptive seed point selection method based on similarity measurement, and the latter introduced multiscale input and multichannel output into the original U-Net model and effectively achieved the lightweight design by adjusting the structure and parameters. After that, the middle 30 consecutive CT slice images of each sample were segmented to obtain lung parenchyma, which was employed for training and testing the proposed multichannel parallel input recursive MLP-Mixer network (MPIRMNet) model, achieving the computer-aided diagnosis of CLDs. A total of 4718 and 16290 CT slice images collected from 543 patients were employed to validate the proposed segmentation and classification methods, respectively. Experimental results showed that the improved U-Net model can accurately segment the lung parenchyma in CT slice images, with the Dice, precision, volumetric overlap error, and relative volume difference of 0.96±0.01, 0.93±0.04, 0.05±0.02, and 0.05±0.03, respectively. Meanwhile, the proposed MPIRMNet model achieved appreciable classification effect for normal cases and different CLDs, with the accuracy, sensitivity, specificity, and F1 score of 0.8823±0.0324, 0.8897±0.0325, 0.9746±0.0078, and 0.8831±0.0334, respectively. Compared with classical machine learning and convolutional neural networks-based methods for this task, the proposed classification method had a preferable performance, with a significant improvement of accuracy of 10.74%. The work introduced a two-stage deep learning method, which can achieve the segmentation of lung parenchyma and the classification of CLDs. Compared to previous diagnostic tasks targeting single CLD, this work can achieve various CLDs' diagnosis in the early stage, thereby achieving targeted treatment and increasing the potential and value of clinical applications.

Reinforcement Learning and Genetic Algorithm-Based Network Module for Camera-LiDAR Detection

Cameras and LiDAR sensors have been used in sensor fusion for robust object detection in autonomous driving. Object detection networks for autonomous driving are often trained again by adding or changing datasets aimed at robust performance. Repeat training is necessary to develop an efficient network module. Existing efficient network module development changes to hand design and requires much module design experience. For this, a neural architecture search was designed, but it takes much time and requires optimizing the design process. To solve this problem, we propose a two-stage optimization method for the offspring generation process in a neural architecture search based on reinforcement learning. In addition, we propose utilizing two split datasets to solve the fast convergence problem as the objective function of the genetic algorithm: source data (daytime, sunny) and target data (day/night, adversary weather). The proposed method is an efficient module generation method requiring less time than the NSGA-NET. We confirmed the performance improvement and the convergence speed reduction using the Dense dataset. Through experiments, it was proven that the proposed method generated an efficient module.

Two-stage stock portfolio optimization based on AI-powered price prediction and mean-CVaR models

With the advancement of prediction methods in the field of artificial intelligence, accurate price predictions can effectively support financial portfolio selection. This paper proposes an intelligent stock portfolio selection method based on a prediction neural network, incorporating signal processing and hyperparameter optimization techniques. The method is divided into two key stages: stock price prediction and portfolio selection. In the first stage, we apply Savitzky–Golay filtering to denoise price data and reveal its patterns, and optimize the hyperparameters of the long short-term memory network using the sparrow search algorithm to achieve high-precision stock price predictions. In the second stage, we use the mean-Conditional Value-at-Risk (mean-CVaR) model to select the optimal stock allocation, considering factors such as potential returns, prediction accuracy, and growth rate. Numerical comparisons based on multiple public financial datasets demonstrate that the proposed two-stage method significantly outperforms seven benchmark methods. Specifically, on the Shanghai and Shenzhen 300 (CSI 300) Index dataset, the proposed method achieves a determination coefficient of 0.9980 and an accuracy rate of 97.05%. Additionally, its cumulative returns reach 9.38%, 8.63%, and 7.54% at different confidence levels.

STAFFormer: Spatio-temporal adaptive fusion transformer for efficient 3D human pose estimation

Existing two-stage methods for 3D Human Pose Estimation often use 2D poses as input, which are then lifted to obtain 3D representations. This typically involves frame-by-frame estimation, whether in 2D or 3D, resulting in high computational demands unsuitable for edge devices. Due to the continuity of human movement, the differences between adjacent frames can be minimal, a question arises: is frame-by-frame estimation necessary? Previous works demonstrated the feasibility of using Transformer-based models to estimate poses with sparse frames, focusing on either temporal or spatial dependencies but neglecting holistic spatio-temporal correlations. To address this, we introduce the Spatio-Temporal Adaptive Fusion Transformer (STAFFormer). First, STAFFormer recovers dense temporal frames from sparsely sampled ones obtained from a 2D pose estimator through the Temporal Dense Frame Recovery (TDFR) module. This significantly reduces the computational complexity. Second, STAFFormer employs an adaptive fusion attention mechanism, enhancing accuracy by attentively navigating both spatial and temporal dimensions through the Spatio-Temporal Adaptive Fusion (STAF) module. Furthermore, We introduce a kinematic coherence loss to adapt to subtle joint movements, improving pose estimation fidelity. Finally, we explore the possibility of integrating different pre-training strategies using extensive marker-based datasets. Experimental results on challenging datasets show our network achieves state-of-the-art performance with low computational complexity.

Examining AI competence, chatbot use and perceived autonomy as drivers of students' engagement in informal digital learning

PurposeDigital technology's integration into education has transformed learning frameworks, necessitating the exploration of factors influencing students’ engagement in digital informal settings. This study, grounded in self-determination theory (SDT), proposes a model comprising artificial intelligence (AI) competence, chatbot usage, perceived autonomy (PA), digital informal learning (DIL) and students’ engagement.Design/methodology/approachThe study collected survey data from 409 participants at Saudi Arabian universities, ultimately using 387 valid responses for analysis. This dataset was subjected to a thorough examination to confirm the validity of our proposed model. To decipher the complex interactions within our model, we utilized partial least squares structural equation modeling (PLS-SEM). The study adopted a disjoint two-stage method to formulate a reflective-formative higher-order construct (HOC).FindingsThe study's findings showed that cognitive learning (CL), metacognitive learning (MCL) and social and motivational learning (SML) are the essential components of DIL. Significantly, the study determined that AI competence, chatbot usage, PA and DIL markedly affect students’ engagement. Moreover, the R2 value of 0.592 for student engagement indicates the model's robustness in explaining 59.2% of the variance, highlighting its effectiveness in identifying key drivers of student engagement in DIL contexts.Originality/valueThis research enhances understanding by detailing the intricate relationships among AI competence, chatbot usage, and students’ engagement in informal digital learning. It extends SDT to emphasize intrinsic motivations and AI capabilities, introducing reflective-formative HOCs for comprehending educational intricacies. It provides practical strategies for enhancing AI abilities and chatbot use in education, promoting personalized, engaging and autonomous digital learning spaces, thereby advancing educational theory and practice.

Decentralized energy market integrating carbon allowance trade and uncertainty balance in energy communities

With sustained attention to carbon neutrality, the personal carbon trading (PCT) scheme has been embraced as an auspicious paradigm for scaling down carbon emissions. To facilitate the simultaneous clearance of energy and carbon allowance inside the energy community while hedging against uncertainty, a joint trading framework is proposed in this article. The energy trading is implemented in a peer-to-peer (P2P) manner without the intervention of a central operator, and the uncertainty trading is materialized through procuring reserve of conventional generators and flexibility of users. Under the PCT scheme, carbon allowance is transacted via a sharing mechanism. Possible excessive carbon emissions due to uncertainty balance are tackled by obliging renewable agents to procure sufficient carbon allowances, following the consumption responsibility principle. A two-stage iterative method consisting of tightening McCormick envelope and alternating direction method of multipliers (ADMM) is devised to transform the model into a mixed-integer second-order cone program (MISOCP) and to allow for a fully decentralized market-clearing procedure. Numerical results have validated the effectiveness of the proposed market model.

Neural equivalent circuit models: Universal differential equations for battery modelling

Current battery modelling methodologies including equivalent circuital modelling and electrochemical modelling do not maintain accuracy over diverse operating conditions of current rates, depth-of-discharge and temperatures. To address this limitation, this article proposes the Universal Differential Equations (UDE) framework from scientific machine learning (SciML) as a methodology to generate battery models with improved generalisability. The effectiveness of UDE in enhancing generalisability is demonstrated through a specific battery modelling example. The approach starts with the Thermal Equivalent Circuital Model with Diffusion (TECMD), a state-of-the-art battery model, which is then enhanced through the integration of neural networks into its state equations, resulting in the Neural-TECMD; a UDE model. Additionally, a two-stage UDE parameterisation method is introduced, combining collocation-based pretraining with mini-batch training. The parameterisation method enables the neural networks in the Neural-TECMD to efficiently learn battery dynamics from multiple time series data sets, covering a wide operating spectrum. Consequently, the Neural-TECMD model offers accurate predictions over broader operating conditions, thus enhancing model generalisability. The Neural-TECMD model was validated using 20 data sets covering current rates of 0 to 2C and temperatures from 0 to 45 °C. This validation revealed substantial improvements in accuracy, with an average of 34.51% decrease in RMSE for voltage and a 24.94% decrease for temperature predictions compared to the standard TECMD model.

Capacity and operation joint optimization for integrated energy system based on Nash bargaining game

Integrated energy systems (IESs) can uniformly coordinate different locations and diversified forms of energy systems, which is critical for ensuring low-carbon and sustainable energy usage. This paper proposed a capacity and operation joint optimization method for an IES with multiple interconnected energy hubs (EHs) based on the Nash bargaining game. Firstly, considering the uncertainty of the EH, robust optimization was carried out for both long-term capacity planning and short-term operational strategies at the same stage. Furthermore, the optimization of the single EH is embedded in the cooperative framework of the IES, and it is formulated as an asymmetrical Nash bargaining game model, in which the players bargain over the transaction energy. Finally, this game model is solved by the alternating direction method of multipliers algorithm. Case studies with a typical IES including three EHs are conducted to verify the effectiveness of the proposed methods. It was found that the proposed joint optimization method has better computational ability than traditional two-stage method. The asymmetrical bargaining solution can not only effectively reduce the cost of the IES, but also help achieve a fair benefit balance among EHs.

GLHDR: HDR video reconstruction driven by global to local alignment strategy

Reconstructing High Dynamic Range (HDR) video from alternating exposure Low Dynamic Range (LDR) sequence is an exceptionally challenging task. It not only demands the reliable reconstruction of missing information caused by occlusion or motion without introducing artifacts but also balances the exposure differences between frames to ensure a visually pleasing reconstructed HDR video. Unfortunately, existing methods are typically complex and struggle with unavoidable artifacts and noise, especially when dealing with low-exposed scenes. To tackle this formidable challenge, we propose a two-stage HDR video reconstruction method that employs a global to local alignment strategy. Firstly, we utilize iterative optical flow estimation and hybrid weighting to achieve global alignment, ensuring well-reconstructed in majority of areas. Secondly, the recursive refinement network further addresses locally misaligned areas, reconstructing HDR frames from bottom to top and recursively refining them to yield faithful reconstruction results. Extensive experimental results demonstrate that our method generates the HDR video with fine details and superior visually, surpassing the state-of-the-art method across diverse scenes.

Lassoed boosting and linear prediction in the equities market

We consider a two-stage estimation method for linear regression. First, it uses the lasso in Tibshirani to screen variables and, second, re-estimates the coefficients using the least-squares boosting method in Friedman on every set of selected variables. Based on the large-scale simulation experiment in Hastie, Tibshirani, and Tibshirani, lassoed boosting performs as well as the relaxed lasso in Meinshausen and, under certain scenarios, can yield a sparser model. Applied to predicting equity returns, lassoed boosting gives the smallest mean-squared prediction error compared to several other methods.

Optimization study of high-dimensional varying coefficient partially linear model based on elastic network

In this paper, we study the high-dimensional varying coefficient partially linear model and proposes a variable parameter selection method combined with elastic network. The two-stage estimation method is adopted and the alternating direction multiplier method (ADMM) is used to solve the model. Firstly, based on the varying coefficient partially linear model, we use local polynomial estimation method to estimate the non-parametric part of the model and convert it into a parametric model. Secondly, the elastic network regularization term is introduced into the model for variable selection, and the alternating direction multiplier method (ADMM) is used to solve the model. Then, the convergence of the algorithm is proved under certain conditions. Finally, the proposed method has been validated to have good effectiveness and certain practicality through numerical simulation of two types of high-dimensional datasets and example analysis.

A rapid evaluation method of the seismic damage to buildings based on UAV images

With the widespread application of UAVs in post-disaster damage evaluation, feature extraction and analysis based on drone images has become the main trend in seismic assessment. In this research, a fast assessment method that integrates object-oriented 2D classification and 3D assessment is proposed in the paper. Firstly, to segment the regional orthoimage, a multi-scale feature segmentation method is developed to segment the orthoimage of the region and the features are selected based on the attributes of ground objects. Meanwhile, a two-stage object-oriented hierarchical classification method is put forward to establish partition rules at different levels. The classification method takes the overall attributes of the object into account which reduces the amount of data correspondingly resulting in much higher efficiency and accuracy of feature extraction than the pixel-based method. In addition, the integrity of building’s top texture after classification is used as the first criterion to assess building seismic damage. Besides, a 3D elevation analysis method based on regional normalized digital surface model (nDSM) is developed by comparing digital surface model with digital elevation model to serve as the second criterion for determining the vertical changes in building. The proposed method is validated by a field test at the Beichuan earthquake site with an accuracy rate of 86.95 %. The experimental results show that the proposed method can meet the accuracy requirement of seismic assessment and has a high application value in the field of regional disaster evaluation.

Two-stage robust operation of electricity-gas-heat integrated multi-energy microgrids considering heterogeneous uncertainties

With the widespread adoption of combined heat and power and power-to-heat technologies, multi-energy microgrids (MEMGs) have been garnering significant research attention from both industry and academia. However, dealing with uncertainties from renewable energy and load and coordinating multiple energy carriers are the main challenges for MEMG operation. In this regard, a two-stage robust operation method of electricity-gas-heat integrated MEMGs considering heterogeneous uncertainties is proposed in this paper. First, network models for an electricity-gas-heat-based distribution-level MEMG are formulated considering the dynamic characteristics of gas and heat networks. Then, the power-to‑hydrogen-and-heat unit and ladder-type carbon trading mechanism are introduced to reduce the curtailment of wind power and carbon emissions. Further, a two-stage robust optimization (TSRO) method is applied to tackle uncertainties of wind power and load under extreme scenarios in the MEMG operation by iteratively solving the operation problem with the column and constraint generation (C&CG) algorithm. Finally, case studies are conducted to verify our proposed method, demonstrating that it can reduce the multi-energy supply cost while the stepped carbon trading mechanism can also significantly reduce carbon emissions.

Multi-agent cooperative area coverage: A two-stage planning approach based on reinforcement learning

Multi-agent area coverage aims to accomplish the complete traversal of the target area through cooperation between agents. Focusing on the problems of low coverage efficiency and weak practicability in the existing methods, we propose a two-stage area coverage method based on multi-agent deep reinforcement learning. In the first stage, we convert the coverage path planning problem into an optimal grid selection problem, and according to the equivalence of agents in cooperative tasks, we propose a distributed coverage path planning algorithm based on QMIX and a grid coverage map. The second stage is to realize the cooperative navigation control in a constrained environment with obstacles and non-ideal communication conditions. To implement the stage, we design a hybrid attention mechanism to adaptively aggregate important feature information of adjacent agents and obstacles, which efficiently exploits the limited local perception and communication capabilities of agents to perform cooperative control. The experimental results show that the proposed two-stage multi-agent area coverage method can accomplish the area coverage task in the environment with random obstacles, and the area coverage efficiency and robustness are significantly better than other reinforcement learning based or traditional coverage algorithms. In addition, the results also verify that the proposed method has the advantage of adapting to the dynamic changes in the number of agents and the communication range.

Revisiting the resource curse hypothesis from the viewpoint of green growth: The role of Fintech as the de-cursing agent

Conventionally, the resource curse hypothesis highlights the economic growth-inhibiting impacts that are often associated with the expansion of natural resource-based industries. However, not much focus has been given to assessing whether natural resource consumption hinders green growth establishment, as well. Hence, this study makes a novel attempt at evaluating the validity of the resource curse hypothesis from the viewpoint of green growth in the context of selected Sub-Saharan African countries. Notably, this study focuses on whether natural resource abundance upholds green growth in the concerned Sub-Saharan African nations, especially by reducing their annual carbon emission-adjusted economic growth levels. More importantly, this study ascertains whether Fintech development, via the channel of Fintech-based startup financing, can promote green growth in these countries and thereby act as a de-cursing agent for tackling the possible economic and environmental curses imposed by their natural resource sectors. Accordingly, by employing a recently developed two-stage instrumental variable regression method, the resource curse hypothesis is verified from the perspective of green growth establishment. On the other hand, Fintech development is found to not only promote green growth on its own but also act as a de-cursing agent that partially offsets the green growth-impeding effects associated with natural resource consumption. Furthermore, capital accumulation is identified as a green growth-restricting factor while institutional quality improvement is observed to stimulate green growth across the selected Sub-Saharan African countries. Therefore, considering these major findings, a couple of green growth-promoting policy suggestions are put forward.

Resource utilization of sewage sludge as a nitrogen self-doped carbon quantum dot for specific detection of Fe3+ and Hg2+

With the widespread application of sewage biological treatment technology, a large amount of sewage sludge is generated during the sewage treatment process, which not only poses a threat to the environment but also results in wastage of C and N resources. In this study, the nitrogen self-doped fluorescent carbon dots (N-CDs) with an average particle size of 3.43 nm were synthesized green through a two-stage hydrothermal method. It was found that N-CDs can detect Fe3+ through internal filtration effect (IFE), with a linear detection range of 0–250 μM and a detection limit of 2.207 μM. Furthermore, N-CDs can also serve as a “turn off–on” probe for monitoring Hg2+, with a linear detection range of 2–10 μM and a detection limit of 0.989 μM. Interestingly, oxygen-containing groups were shown to be sensitive to Hg2+, which has been demonstrated for the first time. Moreover, the practical application of N-CDs as a double ion sensor in the detection of Fe3+ and Hg2+ has achieved good results for real power plant wastewater.

Two-stage residual networks for damage identification and location of stiffened composite panel based on guided waves

The damage detection of the stiffened composite panel, as a typical aircraft structure, is a research hotspot in Structural Health Monitoring (SHM). where guided waves propagate with multi-modal and dispersion characteristics. The traditional damage detection method manually extracts the potential discriminative features of the signal to achieve damage identification, depending on expert experience. In this paper, we propose a two-stage residual networks (ResNets) framework based on guided waves to locate damage in the stiffened composite panel, which automatically mines the high-dimensional features with sensitive discriminant information. The guided wave signal acquisition system collects four types of data: health data, stringer damage data, damage data on the skin of the stringer-side, and damage data on the skin-side. The first-stage utilizes a ResNet to classify the structure condition, while in the second-stage, three separate ResNets are employed to locate the damage according to the classification results of the first-stage. The experimental results show that the accuracy of the first-stage damage classification and the damage localization of the stringer and the skin of the stringer-side in the second-stage has reached 100%, and that of the skin-side is 99.13%, which significantly outperforms single-stage methods. This strategy of inter-class discrimination and intra-class precise localization of damage can not only identify the damaged regions but also determine the specific location of the damage, which greatly increases the performance of SHM. The present two-stage method is a potential solution for future SHM strategies and further investigation is warranted.