Improvement In Accuracy Research Articles

A Short-Term Forecasting Method for High-Frequency Broadcast MUF Based on LSTM

This paper proposes a short-term forecasting method for high-frequency broadcast Maximum Usable Frequency (MUF) based on Long Short-Term Memory (LSTM) to meet the demand for refined and precise high-frequency broadcast coverage. Based on the existing infrastructure of broadcast and television stations, we established an experimental verification system to collect data for approximately three years. Two links were selected based on data quality: Urumqi to Lhasa and Lanzhou to Lhasa. A short-term forecast of MUF was conducted using the data from these two links. Comparison and analysis were conducted between the forecasting results of our model and those of the REC533 model. Our proposed method outperforms the REC533 forecasting results overall, with a reduction in root mean square error (RMSE) of 0.66 MHz and an improvement in forecast accuracy of 14.77%. The comparison result demonstrates the feasibility and accuracy of our model.

Scene-aware data augmentation for ship detection in SAR images

ABSTRACT Ship detection in synthetic aperture radar (SAR) images is widely applied in marine monitoring. In recent years, convolutional neural networks (CNNs) have made significant advances in SAR ship detection. However, the scene imbalance problem, i.e. much more offshore scenes than inshore scenes in SAR images, severely hampers the improvement of detection accuracy. In this article, we propose scene-aware data augmentation to address the problem. First, a novel concept, inshore rate, is introduced to estimate the land area percentage of various scenes in SAR images. We can divide image samples into inshore and offshore ones based on the inshore rate without manual category annotations. Second, to augment inshore samples, we synthesize dummy inshore images by cropping the ground truth bounding boxes (GT bboxes) from offshore images and pasting them onto inshore images. This offline data augmentation strategy can address the scene imbalance problem and improve detection accuracy. The experiments conducted on SAR Ship Detection Dataset (SSDD) and Large-Scale SAR Ship Detection Dataset (LS-SSDD) demonstrate the effectiveness of the proposed method in improving detection performance.

Numerical solution for stochastic Volterra-Fredholm integral equations with delay arguments

We present a method for computing the stochastic operational matrix of integration to advance the study of stochastic Volterra-Fredholm integral equations (SVFIEs) based on delay arguments. First, the method evaluates the combined effects of the delay and its parameters on the accuracy improvement of the convergence rate. Our results can be applied to SVFIEs, with the operational delay matrices of the block pulse function simplified to algebraic ones. Numerical calculations were performed on a PC using Python 3 programs. Results also demonstrate the accuracy of approximate solutions; arithmetic operations are carried out without the need for derivation or integration.

Effect of intra-year Landsat scene availability in land cover land use classification in the conterminous United States using deep neural networks

The Landsat archive having consistent revisit times, near global extent and extensive multi-decadal temporal coverage offers a unique opportunity for land cover land use product generation. Along with this vast volume of freely available data, new classification methods based on deep learning have improved modeling capabilities. This manuscript investigates the effect of intra-annual Landsat scene availability in the accuracy of land cover land use classification in the conterminous United States. More specifically, we seek to quantify the effect of: i) increased monthly scene availability, and ii) specific months that may result in higher classification accuracy across different classes. Identifying specific months with comparable classification accuracy to the entire time series could offer significant computational gains for large-scale mapping. Our experiment incorporated deep learning classifiers and a wide range of reference data across the continental United States. Results were contrasted between five large U.S. climatic regions to further differentiate this intra-annual effect. Our findings indicate that the total number of months can have a highly variable effect in the classification accuracy ranging from minor (a few percentage points in terms of class F1 accuracy) to extremely beneficial (approaching 50% F1 improvement moving from four to twelve month observations). The benefit of increased month observations varied among climatic regions and classes: when all climate regions were combined, the grass/shrub and cultivated classes improved their F1 accuracy up to 30%, while the water class saw the least improvement of about 5%, partially due to its limited room for improvement. The effect of specific month combinations was also examined, where the total number of months was kept constant and the included months varied. The difference between the best month combination and the median combination value was estimated to be as high as about 30% for the four monthly observations scenario and the grass/shrub class. Further validation of the month selection importance comes from an example implementation scenario where F1 improvements can be as high as 10%. Our work demonstrated that month selection may offer such benefits that in some classes and climatic regions this time selection optimization is an inevitable choice due to large accuracy improvements. Also, the potential data reduction with targeted month selection would be particularly appealing to large-scale classification tasks. Due to the large extent of the climatic regions further studies are needed to quantify a more localized effect along with explanation of potential drivers.

Planning of electric vehicle charging stations: An integrated deep learning and queueing theory approach

This study presents a hybrid solution for the charging station location-capacity problem. The proposed approach simultaneously determines the location and capacity of charging stations (i.e., number of charging piles), and assigns piles to electric vehicles based on their level of charge. The problem is formulated as a bi-objective mixed-integer nonlinear programming model to minimize the total cost of establishing charging stations together with the average customers’ waiting time. The proposed solution combines queueing theory with mathematical modelling to estimate the average waiting time. A deep learning algorithm is then developed to enhance the precision of waiting time estimation. Another contribution is involving a deep neural network model in improving NSGA-II algorithm. Numerical experiments are conducted in Halifax, Canada to assess the performance of the proposed framework. The results demonstrate the strong predictive performance of the deep learning algorithm and highlight the limitations of traditional queueing models in estimating waiting times in charging stations (i.e., 99.8% improvement in computation time, as well as accuracy improvement of time estimations from 13% to 1.6% deviation). Several valuable insights are obtained to improve the operational performance of charging stations such as achieving a significant (i.e., 61.5%) drop in the average waiting time across the network by a modest (i.e., 29.2%) increase in the initial investments. Also, it reveals that the variability of service rate significantly impacts the average waiting time (i.e., a 50% increase in the variability of service rate causes a substantial 950.56% surge in the average waiting time). The findings underscore the need to control service rate fluctuations to reduce wait times and boost driver satisfaction. The improved NSGA-II algorithm shows 12.77% improvement in the Pareto front solutions. Finally, the proposed prioritization strategy based on the charging level of vehicles could reduce the average waiting time and cost compared to the first-come-first-served strategy.

Brain tumor segmentation in MRI with multi-modality spatial information enhancement and boundary shape correction

Brain tumor segmentation is currently of a priori guiding significance in medical research and clinical diagnosis. Brain tumor segmentation techniques can accurately partition different tumor areas on multi-modality images captured by magnetic resonance imaging (MRI). Due to the unpredictable pathological process of brain tumor generation and growth, brain tumor images often show irregular shapes and uneven internal gray levels. Existing neural network-based segmentation methods with an encoding/decoding structure can perform image segmentation to some extent. However, they ignore issues such as differences in multi-modality information, loss of spatial information, and under-utilization of boundary information, thereby limiting the further improvement of segmentation accuracy. This paper proposes a multimodal spatial information enhancement and boundary shape correction method consisting of a modality information extraction (MIE) module, a spatial information enhancement (SIE) module, and a boundary shape correction (BSC) module. The above three modules act on the input, backbone, and loss functions of deep convolutional networks (DCNN), respectively, and compose an end-to-end 3D brain tumor segmentation model. The three proposed modules can solve the low utilization rate of effective modality information, the insufficient spatial information acquisition ability, and the improper segmentation of key boundary positions can be solved. The proposed method was validated on BraTS2017, 2018, and 2019 datasets. Comparative experimental results confirmed the effectiveness and superiority of the proposed method over state-of-the-art segmentation methods.

Local spatial and temporal relation discovery model based on attention mechanism for traffic forecasting

Recognizing the evolution pattern of traffic condition and making accurate prediction play a vital role in intelligent transportation systems (ITS). With the massive increase of available traffic data, deep learning-based models have attracted considerable attention for their impressive performance in traffic forecasting. However, the majority of existing approaches neglect to model of asynchronously dynamic spatio-temporal correlation and fail to consider the impact of historical traffic data on future condition. Additionally, the attribute of deep learning method presents challenges in interpreting the explicit spatiotemporal relationships. In order to enhance the accuracy of traffic prediction as well as extract comprehensive and explainable spatial–temporal relevance in traffic networks, we propose a novel attention-based local spatial and temporal relation discovery (ALSTRD) model. Our model firstly implements feature representation learning to effectively express latent input traffic information. Then, a local attention mechanism structure is established to model asynchronous dependencies of historical input data. Finally, another attention network and the Pearson Correlation Coefficient method are introduced to extract the elaborate influence of the historical traffic condition of neighboring roads on the future condition of the target road. The experiment results on several datasets demonstrate that our model achieves significant improvements in prediction accuracy compared to other baseline methods, which can be attributed to its ability to extract the fine-grained correlation among historical traffic data and capture the dynamic association between past and future data. In addition, the incorporation of attention mechanism and Pearson Correlation Coefficient promotes the model’s ability to elucidate spatiotemporal correlations among traffic data, thereby providing a more robust explanation.

A novel sequence-to-sequence based deep learning model for satellite cloud image time series prediction

Satellite cloud imagery is pivotal for meteorologists in characterizing weather patterns, detecting climate anomaly regions, and predicting rain effects. The task of satellite cloud image forecasting is crucial, and while deep learning models have shown promise in predicting spatio-temporal data, traditional methods face challenges with extracting long-term spatio-temporal features and high computation costs. To address these issues, we propose the Re-parameterized Sequence-to-Sequence Satellite Cloud Imagery Prediction Network (Rep-SSCIPN). Rep-SSCIPN utilizes Rep-convolution layers to reduce inference-time cost and memory consumption, enhancing efficiency by converting re-parameterized blocks into a single convolution layer during inference. The sequence normalization attention mechanism in Rep-SSCIPN highlights crucial feature sequences and establishes their inter-dependencies. We validate our novel method using a real-world satellite cloud image dataset from the meteorological satellite “Himawari.” Experimental results showcase significant improvements in prediction accuracy and reconstruction quality compared to ConvLSTM, PredRNN, FCLSTM, LMC, SimVP and SCSTque models. The efficiency gains make Rep-SSCIPN a promising advancement for satellite cloud image prediction.ARTICLE INFO.

Capabilities in Physical Education: A Comparative Analysis of School Heads’ and Teachers’ Perspectives

Background: Effective education delivery relies on PE teachers’ abilities. Research indicates that students’ school performance significantly depends on the quality of PE teaching. Objectives: This study focused on assessing instructional capabilities in physical education. Specifically, it examined how school heads and teachers evaluated teachers’ abilities in content knowledge, pedagogical skills, communication, and professionalism. Additionally, the study compared assessments between the two groups. Methods: The study employed a descriptive research design involving 98 physical education teachers and 59 school heads from the Division of Batangas City. Data collection utilized a researcher-made questionnaire and interviews. Statistical analysis involved weighted mean and independent t-tests for data interpretation. Results: The assessment revealed valuable insights regarding PE teachers’ instructional capabilities. Administrators acknowledged strengths in lesson introduction and accurate content delivery. However, areas for improvement included integrating key elements and promoting higher-order thinking skills. PE teachers felt confident introducing lessons but recognized a need to refine performance demonstrations. Administrators praised diverse teaching methods but suggested creating more challenging experiences. Communication skills were positive, with room for improvement in grammatical accuracy. PE teachers expressed dedication to student physical education but identified a need for more decisive leadership in school-based activities. The findings highlighted specific improvement needs while emphasizing contrasting perceptions between administrators and PE teachers. Conclusion: The assessment found that physical education (PE) teachers are capable in delivering junior high school PE. However, they need familiarity with emerging trends to enhance instruction. School heads and PE teachers have differing views on instructional capabilities, attributed to assessment sources: administrators rely on scheduled observations, while PE instructors use self-assessment and daily teaching experiences.

A trajectory data-driven approach for traffic risk prediction: incorporating variable interactions and pre-screening

ABSTRACT Although historical crash data and trajectory data have been widely applied to crash and risk predictions, both types of data have their own limitations. As a solution, this study investigates the impact of the traffic flow parameters and their interaction terms on risk prediction performance, employing a variable pre-screening approach (i.e. Smoothly Clipped Absolute Deviation (SCAD)). A research framework is proposed for more efficient risk prediction, and a detailed case study is further conducted using the proposed approach. In the case study, real vehicle trajectory data from HighD are processed and used, which can be aggregated to extract both traffic flow parameters and corresponding risk data during a specific time interval. As for the risk detection, Time-to-Collision (TTC) index is utilized to identify risky conditions. For different lanes (i.e. inner, middle and outer lanes), the impact of variables, including interaction terms, on risk is explored using the SCAD-logistic models. Furthermore, machine learning methods are employed to compare the risk prediction performance before and after considering interaction terms, as well as before and after variable pre-screening. Finally, the superiority of the machine learning models after SCAD-based variable pre-screening is demonstrated. Results indicate that the interaction terms between traffic flow parameters have significant impacts on the traffic risk. Besides, considering interaction terms and variable pre-screening can improve risk prediction accuracy. Furthermore, the proposed models outperform Random Forest (RF) in terms of predicting traffic risk, achieving a maximum 21.24% accuracy improvement and reducing computational time by up to 31.51%. Findings of this study are expected to contribute to the high-precision prediction of real-time risk in the future.

Enhancing accuracy of surface wind sensors in wind tunnel Testing: A Physics-Guided neural network calibration approach

Irwin’s surface wind sensor is widely used in wind tunnel testing for urban and environmental aerodynamics studies. However, the conventional physics-based calibration of this sensor could result in reduced measurement accuracy in regions with low flow velocities and high turbulence intensity. To address this issue, this study proposes a novel physics-guided neural network (PGNN) calibration approach, which couples a physics-based calibration model, derived from extended Taylor series expansions of measured wind speed, with an adaptive, data-driven general regression neural network. Sensors are calibrated within the turbulent boundary layer of an empty flat plate, considering both mean and standard deviation of wind velocity measured by high-accuracy thermal anemometry. The accuracy of calibrated sensors is then assessed using a 1:400 benchmark urban model. Experimental results show significant improvement in measurement accuracy, reducing mean absolute percentage error for wind speed standard deviation from 92.3 % with the current model to 9.8 % using PGNN.

DGVINS: Tightly Coupled Differential GNSS/Visual/Inertial for Robust Positioning Based on Optimization Approach

Abstract Due to the fragility of single-sensor positioning technology in complex scenarios, especially in complex urban areas, multi-sensor positioning technology is becoming increasingly popular. To further improve the robustness of the positioning system by fully utilizing the information from various sensors, this article proposes a differential-GNSS-visual-inertial navigation system (DGVINS) that tightly fuses differential global navigation satellite system (GNSS), vision and inertial information to provide accurate, robust and seamless position information for intelligent navigation applications. DGVINS effectively utilizes all sensor measurements within the factor graph optimization (FGO) framework. When using the carrier phase of GNSS, single-epoch ambiguity optimization is employed to prevent cycle slip detection and adapted to complex environments. We conducted experiments on public datasets with various features and compared the performance of simple differential-GNSS (DGNSS), DGNSS+Inertial, and the state-of-the-art GNSS-visual-inertial navigation systems (GVINS). We also compared the performance of different combinations of GNSS differential factors in various environments. Due to the superiority of differential GNSS and its appropriate integration with visual and inertial measurements, the experimental results demonstrate that DGVINS exhibits significant improvements in accuracy, stability, and continuity in both GNSS-challenged and vision-challenged environments.

Assessing the Interpretability–Performance Trade-Off of Artificial Neural Networks Using Sentinel Fish Health Data

A number of sentinel species are regularly sampled from the environment near the Oil Sands Region (OSR) in Alberta, Canada. In particular, trout-perch are sampled as a proxy for the health of the aquatic ecosystem. As the development of the OSR began before the environmental monitoring program was in place, there is currently no established measure for the baseline health of the local ecosystem. A common solution is to calculate normal ranges for fish endpoints. Observations found to be outside the normal range are then flagged, alerting researchers to the potential presence of stressors in the local environment. The quality of the normal ranges is dependent on the accuracy of the estimates used to calculate them. This paper explores the use of neural networks and regularized regression for improving the prediction accuracy of fish endpoints. We also consider the trade-off between the prediction accuracy and interpretability of each model. We find that neural networks can provide increased prediction accuracy, but this improvement in accuracy may not be worth the loss in interpretability in some ecological studies. The elastic net offers both good prediction accuracy and interpretability, making it a safe choice for many ecological applications. A hybridized method combining both the neural network and elastic net offers high prediction accuracy as well as some interpretability, and therefore it is the recommended method for this application.

A hybrid network of NARX and DS-attention applied for the state estimation of lithium-ion batteries

Monitoring the state of the battery, including the state of charge (SOC) and state of health (SOH), is crucial for ensuring the safety and reliability of electrical equipment. The paper presents a novel hybrid network that combines nonlinear autoregressive model with exogenous inputs (NARX) and DS-attention. The proposed DS-attention method establishes a robust mapping relationship between inputs and outputs, it is a specialized method of the recurrent neural network that enhances the estimation performance by incorporating division function and self-adaptive function into the attention mechanism. The division function is designed to efficiently differentiate between exogenous inputs and state outputs, thereby minimizing the potential for cross-interference between them, and the self-adaptive function optimizes the query features within the attention mechanism. The results demonstrate that the hybrid method of NARX and DS-attention achieves higher prediction accuracy for SOH and SOC estimation of lithium-ion batteries. Specifically, compared with the best-performing similar methods, the proposed hybrid method enhanced the prediction accuracy by 22.9%, 60.7%, and 51.2% across three different SOH datasets, respectively. In terms of long sequence SOC forecasting, the improvements in prediction accuracy under two different working conditions are 82.5% and 60.5%, respectively. The proposed algorithm optimizes the attention mechanism based on the characteristics of the NARX, resulting in higher estimation accuracy for predicting the state of lithium batteries.

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Abstract In this research study, a novel approach that leverages the power of transfer learning (a famous deep learning technique) is proposed to diagnose multiple types of clutch faults including, worn release fingers, fractured pressure plates, deteriorated pressure plates, loss of friction material and distorted tangential strips using spectrogram plots. To train and validate the diagnostic system, vibration readings were taken from a specially designed test rig with the help of piezoelectric accelerometer while the clutch system was operated under different load conditions of 0 (no load), 5 and 10 kg This procedure of data collection was then repeated to acquire the vibration data for all of the fault conditions by replacing the good with fault components individually. These vibration signals were further processed and transformed into spectrogram plots that serves as the input data for the deep learning models considered. Fine-tuning techniques were applied on pretrained networks to maximise the prediction accuracy of the models to effectively determine and diagnose faults in the clutch system. For this study 12 pre-trained networks were chosen namely, Xception, InceptionResNet, DenseNet, AlexNet, VGG16, GoogLeNet, VGG19, ResNet101, ResNet50, InceptionV3, MobileNetV2 and ShuffleNet. To optimize the performance of deep learning models, a systematic adjustment of hyperparameters such as the train-test split ratio, learning rate, optimizer and batch size for each network model was carried out. Through careful experimentation and analysis, significant improvements in fault classification accuracy were achieved thereby enhancing the reliability and effectiveness of the diagnostic system. From the results it was noted that 100% classification accuracy was displayed by AlexNet (for the no load condition and the 10kg load condition) and GoogLeNet (for 5 kg load condition) with extremely low computation times.

One-dimensional model of manifold microchannels for embedded cooling: Prediction of thermal performance and flow non-uniformity

A one-dimensional model has been developed to accurately predict the thermal performance and flow non-uniformity of the manifold microchannels (MMC) for embedded liquid cooling. The model consists of one-dimensional governing equations derived from the integral relations of momentum and energy over appropriately-defined two separate control volumes. To validate the model, a series of 3-D numerical simulation is conducted over the wide ranges of the Reynolds number (Rem,in) at the manifold inlet from 560 to 3190, the dimensionless hydraulic flow length (x+) from 0.012 to 0.123, and the dimensionless thermal flow length (x∗) from 0.002 to 0.023. It is shown that the model provides accurate predictions of the thermal performance and flow non-uniformity (CV) of MMC heat sinks within the root mean square percentage error (RMSPE) of 6% and 26% for 50 data points, respectively. The significant improvement of the prediction accuracy is made over the earlier models with an error reduction of 82%. Finally, a design guideline for the uniform flow distribution is suggested for the first time based on a newly proposed explicit correlation for predicting the flow non-uniformity: the dynamic pressure at the manifold inlet should be kept smaller than the pressure drop across the microchannels.

Enhanced Brain Tumor Classification: A Hybrid Classifier Approach

Early detection of brain tumors is crucial for enhancing patient survival rates and treatment effectiveness. Analyzing magnetic resonance imaging (MRI) images by manually is a difficult task. Therefore, there is a requirement for digital techniques to improve the accuracy of tumor diagnosis. Study of this work was to enhance the performance of Machine Learning classifiers rather than move to deep learning. Machine Learning classifiers are advantageous compared to deep learning methods because of their ability to train on small datasets, low computing time complexity, affordability, and ease of adoption by individuals with less expertise The proposed methodology encompasses image processing, segmentation, feature extraction, feature reduction, and classification, leveraging techniques such as Otsu’s thresholding, GLCM, BWT, and PCA to extract pertinent information while minimizing dimensionality. By integrating a hybrid classifier that combines SVM, DT, and KNN, along with ensemble techniques, significant improvements in classification accuracy and robustness are achieved. Experimental results on the TCGA-GBM dataset showcase accuracy of 96.9%, with high specificity (95.14%) and sensitivity (97.14%) for classify normal tumors (benign tumors) and cancerous tumors (malignant tumors). Furthermore, the proposed methodology demonstrates efficiency with relatively low execution time, rendering it suitable for real-world applications.

An adaptive transition probability matrix with quality seeds for cellular automata models

ABSTRACT The cellular automata (CA) model is the predominant method for predicting land use and land cover (LULC) changes. The accuracy of this model critically depends on well-defined transition rules, which encapsulate the local dynamics of complex systems and facilitate the manifestation of organized global patterns. While current studies largely concentrate on land use transition matrices as core elements of these rules, exclusive reliance on these matrices is insufficient for capturing the full spectrum of land use change potential. Addressing this gap, our research introduces the adaptive transition probability matrix with quality seeds (ATPMS) model, which incorporates both the Markov model and the genetic algorithm (GA) into the traditional CA framework. Furthermore, an artificial neural network (ANN) is utilized to determine land suitability. Implemented in Beijing, Wuhan, and the Pearl River Delta (PRD), our results indicate that the ATPMS-ANN-CA model surpasses the standard Markov-ANN-CA model in various validation metrics, displaying improvements in overall accuracy (OA) by 0.03% to 0.74% and figure of merit (FoM) by 3.67% to 63.14%. Additionally, the ATPMS-ANN-CA model excels in providing detailed landscape analysis.

Interactive attack-defense for generalized person re-identification

Generalized Person Re-Identification (GReID) aims to develop a model capable of robust generalization across unseen target domains, even with training on a limited set of observed domains. Recently, methods based on the Attack-Defense mechanism are emerging as a prevailing technology to this issue, which treats domain transformation as a type of attack and enhances the model’s generalization performance on the target domain by equipping it with a defense module. However, a significant limitation of most existing approaches is their inability to effectively model complex domain transformations, largely due to the separation of attack and defense components. To overcome this limitation, we introduce an innovative Interactive Attack-Defense (IAD) mechanism for GReID. The core of IAD is the interactive learning of two models: an attack model and a defense model. The attack model dynamically generates directional attack information responsive to the current state of the defense model, while the defense model is designed to derive generalizable representations by utilizing a variety of attack samples. The training approach involves a dual process: for the attack model, the aim is to increase the challenge for the defense model in countering the attack; conversely, for the defense model, the focus is on minimizing the effects instigated by the attack model. This interactive framework allows for mutual learning between attack and defense, creating a synergistic learning environment. Our diverse experiments across datasets confirm IAD’s effectiveness, consistently surpassing current state-of-the-art methods, and using MSMT17 as the target domain in different protocols resulted in a notable 13.4% improvement in GReID task average Rank-1 accuracy. Code is available at: https://github.com/lhf12278/IAD.

Predicting Bus Travel Time in Cheonan City Through Deep Learning Utilizing Digital Tachograph Data

Urban transportation systems are increasingly burdened by traffic congestion, a consequence of population growth and heightened reliance on private vehicles. This congestion not only disrupts travel efficiency but also undermines productivity and urban resident’s overall well-being. A critical step in addressing this challenge is the accurate prediction of bus travel times, which is essential for mitigating congestion and improving the experience of public transport users. To tackle this issue, this study introduces the Hybrid Temporal Forecasting Network (HTF-NET) model, a framework that integrates machine learning techniques. The model combines an attention mechanism with Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) layers, enhancing its predictive capabilities. Further refinement is achieved through a Support Vector Regressor (SVR), enabling the generation of precise bus travel time predictions. To evaluate the performance of the HTF-NET model, comparative analyses are conducted with six deep learning models using real-world digital tachograph (DTG) data obtained from intracity buses in Cheonan City, South Korea. These models includes various architectures, including different configurations of LSTM and GRU, such as bidirectional and stacked architectures. The primary focus of the study is on predicting travel times from the Namchang Village bus stop to the Dongnam-gu Public Health Center, a crucial route in the urban transport network. Various experimental scenarios are explored, incorporating overall test data, and weekday and weekend data, with and without weather information, and considering different route lengths. Comparative evaluations against a baseline ARIMA model underscore the performance of the HTF-NET model. Particularly noteworthy is the significant improvement in prediction accuracy achieved through the incorporation of weather data. Evaluation metrics, including root mean squared error (RMSE), mean absolute error (MAE), and mean squared error (MSE), consistently highlight the superiority of the HTF-NET model, outperforming the baseline ARIMA model by a margin of 63.27% in terms of the RMSE. These findings provide valuable insights for transit agencies and policymakers, facilitating informed decisions regarding the management and optimization of public transportation systems.