RESUMO
This study aims to use machine learning methods to examine the causative factors of significant crashes, focusing on accident type and driver's age. In this study, a wide-ranging data set from Jeddah city is employed to look into various factors, such as whether the driver was male or female, where the vehicle was situated, the prevailing weather conditions, and the efficiency of four machine learning algorithms, specifically XGBoost, Catboost, LightGBM and RandomForest. The results show that the XGBoost Model (accuracy of 95.4%), the CatBoost model (94% accuracy), and the LightGBM model (94.9% accuracy) were superior to the random forest model with 89.1% accuracy. It is worth noting that the LightGBM had the highest accuracy of all models. This shows various subtle changes in models, illustrating the need for more analyses while assessing vehicle accidents. Machine learning is also a transforming tool in traffic safety analysis while providing vital guidelines for developing accurate traffic safety regulations.
Assuntos
Acidentes de Trânsito , Aprendizado de Máquina , Acidentes de Trânsito/mortalidade , Humanos , Feminino , Masculino , Fatores de Risco , Pessoa de Meia-Idade , Adulto , Fatores Etários , Idoso , Adulto Jovem , Algoritmos , AdolescenteRESUMO
In this study, the variation of shear strength behavior and particle breakage (after shearing), as a function of moisture state and compaction level, is investigated for recycled concrete aggregate blended with recycled clay masonry. Recycled masonry was blended with concrete aggregate in percentages ranging from 0% to 30% by total weight. Tests include; basic engineering characteristics (particle size, modified compaction, hydraulic conductivity, and California Bearing Ratio, CBR) as well as unconsolidated undrained static triaxial testing. In triaxial tests, moisture levels ranged from 60% to 100% of optimum moisture content, but compaction levels ranged from 90% to 98% of maximum dry density. The hydraulic conductivity for blends is approximately 2x10-6 cm/s, which indicates a relatively low hydraulic conductivity. Results show a proportional linear relationship between the shear strength of blends and the level of compaction. Despite this, both apparent cohesion and shear strength exhibited reverse linear trends. As expected, more compaction effort resulted in more particle breakage. Strict control should be performed over the compaction process to achieve the required compaction level which resulting in pavement materials being stiffer.
Assuntos
Reciclagem , Resistência ao Cisalhamento , Tamanho da Partícula , ArgilaRESUMO
Bridges are among the most vulnerable structures to earthquake damage. Most bridges are seismically inadequate due to outdated bridge design codes and poor construction methods in developing countries. Although expensive, experimental studies are useful in evaluating bridge piers. As an alternative, numerical tools are used to evaluate bridge piers, and many numerical techniques can be applied in this context. This study employs Abaqus/Explicit, a finite element program, to model bridge piers nonlinearly and validate the proposed computational method using experimental data. In the finite element program, a single bridge pier having a circular geometry that is being subjected to a monotonic lateral load is simulated. In order to depict damages, Concrete Damage Plasticity (CDP), a damage model based on plasticity, is adopted. Concrete crushing and tensile cracking are the primary failure mechanisms as per CDP. The CDP parameters are determined by employing modified Kent and Park model for concrete compressive behavior and an exponential relation for tension stiffening. The performance of the bridge pier is investigated using an existing evaluation criterion. The influence of the stress-strain relation, the compressive strength of concrete, and geometric configuration are taken into consideration during the parametric analysis. It has been observed that the stress-strain relation, concrete strength, and configuration all have a significant impact on the column response.
RESUMO
One of the major problems that cause continual trouble in deep learning networks is that training a large network requires massive labelled datasets. The preparation of a massive labelled dataset is a cumbersome task and requires lot of human interventions. This paper proposes a novel generator network 'Sim2Real' transfer is a recent and fast-developing field in machine learning used to bridge the gap between simulated and real data. Training with simulated datasets often converges due to its size but fails to generalize real-world applications. Simulated datasets can be used to train and test deep learning models, enables the development and evaluation of new algorithms and architectures. By simulating road dataset, researchers can generate large amounts of realistic road-traffic dataset that can be used to study and understand several problems such as vehicular object tracking and classification, traffic situation analysis etc. The main advantage of such a transfer algorithm is to use the abundance of a simulated dataset to generate huge realistic-looking datasets to solve data-intense tasks. This work presents a novel, robust sim2real algorithm that converts the labels of a semantic segmentation map to a realistic-looking street view using the Cityscapes dataset and aims to achieve robust urban mobility for smart cities. Further, the generalizability of the Cycle Generative Adversarial Network (CycleGAN) architecture was tested by using an origami robot dataset for sim2real transfer. We show that the results were found to be qualitatively satisfactory for different traffic analysis applications. In addition, road perception was done using a lightweight SVM pipeline and evaluated on the KITTI dataset. We have incorporated Cycle Consistency Loss and Identity Loss as the metrics to evaluate the performance of the proposed Cycle GAN model. We inferred that the proposed Cycle GAN model provides an Identity loss of less than 0.2 in both the Cityscapes dataset and KITTI datasets. Also, we understand that the super-pixel resolution has a good impact on the quantitative results of the proposed Cycle GAN models.
Assuntos
Algoritmos , Benchmarking , Humanos , Cidades , Aprendizado de Máquina , Percepção , Processamento de Imagem Assistida por ComputadorRESUMO
Safety at the gore areas near diverging ramps is very crucial during planning and implementation of highway safety improvement programs. Limited research has been conducted on safety at the gore areas on arterial roads. This study aims at investigating the impact of improving a sharp gore area in Lincoln, Nebraska by performing a micro simulation before-and-after study with respect to its underlying state of safety and congestion. Data on travel times and traffic volumes for peak hours are incorporated after successful calibration to find out how a geometric intervention can decrease mobility issues as well as the likelihood of crash involvement. This study has utilized VISSIM software package to run the simulation however, to perform safety analysis, Surrogate Safety Assessment Model (SSAM) is used which is developed by the Federal Highway Administration (FHWA). During the model calibration, custom driving behaviors are created to represent driving tendencies of familiar drivers. The simulation results indicated that by adding an auxiliary lane near the gore area, the mobility issues such as bottle necks, lane changing dilemmas and queue lengths are substantially decreased. However, geometric interventions such as provision of a separate lane, increasing ramp spacing, nose spacing, deceleration area and queue storage area considerably reduced the likelihood of rear-end and lane changing crashes. Surrogate safety assessment in diverging ramps, particularly for sharp gores, has not previously been studied, and this study can serve as a primary footmark for future research on ramp-gores safety.
RESUMO
Limited information and data are available on the material and structural performance of GC incorporating lightweight fine aggregate. In this research, three types of lightweight fine materials were utilized to partially replace sand volume of GC. These lightweight materials were rubber, vermiculite, or lightweight expanded clay aggregate (LECA) and they were used in contents of 20%, 40%, 60%, and 100%. The variables were applied to better investigate the efficiency of each lightweight material in GC and to recommend GC mixes for structural applications. The concrete workability, compressive strength, indirect tensile strength, freezing and thawing performance, and impact resistance were measured in this study. In addition, three reinforced concrete slabs were made from selected mixes with similar compressive strength of 32 MPa and then tested under a 4-point bending loading regime. The results showed that using LECA as sand replacement in GC increased its compressive strength at all ages and all replacement ratios. Compared with the control GC mix, using 60% LECA increased the compressive strength by up to 44%, 39%, and 27%, respectively at 3, 7, and 28 days. The slabs test showed that partial or full replacement of GC sand adversely affected the shear resistance of concrete and caused premature failure of slabs. The slab strength and deflection capacities decreased by 9% and 30%, respectively when using rubber, and by 23% and 59%, respectively when using LECA, compared with control GC slab. The results indicated the applicability of GC mix with 60% LECA in structures subjected to axial loads. However, rubber would be the best lightweight material to recommend for resisting impact and flexural loads.
RESUMO
This study aims to investigate the two-way shear strength of concrete slabs with FRP reinforcements. Twenty-one strength models were briefly outlined and compared. In addition, information on a total of 248 concrete slabs with FRP reinforcements were collected from 50 different research studies. Moreover, behavior trends and correlations between their strength and various parameters were identified and discussed. Strength models were compared to each other with respect to the experimentally measured strength, which were conducted by comparing overall performance versus selected basic variables. Areas of future research were identified. Concluding remarks were outlined and discussed, which could help further the development of future design codes. The ACI is the least consistent model because it does not include the effects of size, dowel action, and depth-to-control perimeter ratio. While the EE-b is the most consistent model with respect to the size effect, concrete compressive strength, depth to control perimeter ratio, and the shear span-to-depth ratio. This is because of it using experimentally observed behavior as well as being based on mechanical bases.
RESUMO
Resilient modulus (Mr) of subgrade soils is one of the crucial inputs in pavement structural design methods. However, the spatial variability of soil properties and the nature of test protocols, the laboratory determination of Mr has become inexpedient. This paper aims to design an accurate soft computing technique for the prediction of Mr of subgrade soils using the hybrid least square support vector machine (LSSVM) approaches. Six swarm intelligence algorithms, namely particle swarm optimization (PSO), grey wolf optimizer (GWO), symbiotic organisms search (SOS), salp swarm algorithm (SSA), slime mould algorithm (SMA), and Harris hawks optimization (HHO) have been applied and compared to optimize the LSSVM parameters. For this purpose, a literature dataset (891 datasets) of different types of soils has been used to design and evaluate the proposed models. The input variables in all of the proposed models included confining stress, deviator stress, unconfined compressive strength, degree of soil saturation, soil moisture content, optimum moisture content, plasticity index, liquid limit, and percent of soil particles (P #200). The accuracy of the proposed models was assessed by comparing the predicted with the observed of Mr values with respect to different statistical analyses, i.e., root means square error (RMSE) and determination coefficient (R2). For modeling the Mr of subgrade soils, percent passing No. 200 sieve, optimum moisture content, and unconfined compressive strength were found to be the most significant variables. It is observed that the performance of LSSVM-GWO, LSSVM-SOS, and LSSVM-SSA outperforms other models in predicting accurate values of Mr. The (RMSE and R2) of the LSSVM-GWO, LSSVM-SSA, and LSSVM-SOS are (6.79 MPa and 0.940), (6.78 MPa and 0.940), and (6.72 MPa and 0.942), respectively, and hence, LSSVM-SOS can be used for high estimating accuracy of Mr of subgrade soils.
Assuntos
Solo , Máquina de Vetores de Suporte , Algoritmos , Inteligência , Análise dos Mínimos QuadradosRESUMO
Unreinforced masonry (URM) is one of the most popular construction materials around the world, but vulnerable during earthquakes. Due to its brittle nature, the URM structures may lead to a possible collapse of the wall of a building during earthquake events causing casualties. In the current research, an attempt is made to enhance the seismic capacity of URM structures by proposing a new innovative composite material that can improve the shear strength and deformation capacity of the URM wall systems. The results revealed that the fiber-reinforced plastic having high tensile and shear stiffness can significantly increase in-plane as well as out-of-plane bending strength of the URM wall. It was recorded that the bending moment of the prism increased up to 549.5% by increasing the bending moment from 490 N*mm to 3183 N*mm per mm deflection of prism upon using glass fibers. Moreover, the ductility ratio amplified up to 5.73 times while the stiffness ratio increased up to 4.16 times with the aid of glass fibers. Since the material used in this research work is low cost, easily available, and no need for any skilled labor, which is economically good. Therefore, the URM walls retrofitted with fiber-reinforced plastic is an economical solution.
RESUMO
Steel fibers are widely extracted from scrap tyres, causing environmental concerns. This paper presents the use of steel fibers in variable proportions extracted from scrap tyres. The enhancement of the confinement was envisaged through the addition of steel fibers obtained from scrap tyres. The study included an experimental program for the development of constitutive material models for ordinary Portland cement (OPC) concrete and concrete with added steel fibers. A mix design was carried out for OPC, targeting a compressive strength of 3000 psi. Steel fibers were added to OPC in ratios of 1.0% to 3.0%, with an increment of 0.5%. Concrete columns, with cross-sectional dimensions of 6 × 6 inches and a length of 30 inches, were cast with both OPC and fiber-reinforced concrete. The column confinement was evaluated with a different spacing of ties (3- and 4-inch center-to-center). Compression tests on the concrete columns indicate that the addition of steel fibers to a concrete matrix results in an appreciable increase in strength and ductility. Overall, increasing the percentage of steel fibers increased the compression strength and the ductility of concrete. The maximum strain in the concrete containing 2.5% steel fibers increased by 285% as compared to the concrete containing 1% of steel fibers. An optimum percentage of 2.5% steel fibers added to the concrete resulted in a 39% increase in compressive strength, accompanied by a significant improvement in ductility. The optimum content of steel fibers, when used in confined columns, showed that confined compression strength increased with the addition of steel fibers. However, it is recommended that additional columns on the basis of the optimum steel fiber content shall be tested to evaluate their effectiveness in reducing the stirrup spacing.