• Journal of Auto-vehicle Safety Association
• /
• v.15 no.1
• /
• pp.55-62
• /
• 2023

In this study, a chest deflection is predicted by introducing a deep learning technique with the results of the frontal impact of the USNCAP conducted for 110 car models from MY2018 to MY2020. The 120 data are divided into training data and test data, and the training data is divided into training data and validation data to determine the hyperparameters. In this process, the deceleration data of each vehicle is averaged in units of 10 ms from crash pulses measured up to 100 ms. The performance of the deep learning model is measured by the indices of the mean squared error and the mean absolute error on the test data. A DNN (Deep Neural Network) model can give different predictions for the same hyperparameter values at every run. Considering this, the mean and standard deviation of the MSE (Mean Squared Error) and the MAE (Mean Absolute Error) are calculated. In addition, the deep learning model performance according to the inclusion of CVW (Curb Vehicle Weight) is also reviewed.

• Journal of Auto-vehicle Safety Association
• /
• v.14 no.1
• /
• pp.32-38
• /
• 2022

In the automobile industry, lamps are frequently used as a mean to emphasize each company's brand identity. Therefore, many detailed design models have emerged in order to realize a differentiated image in preparation for competitive vehicles. Among them, the design of a multi compartment lighting image concept that realizes light divided in multiple space also being introduced by various manufacturers. In this study, in order to solve the problem of cost and weight rise that the existing multi compartment image lamp has, using TRIZ method such as functional analysis modeling and trimming. Through this process, an idea to minimize cost and weight was derived. As the idea was designed in detail, the formation of light did not go as desired, and the diffusion of light also proceeded differently than intended. In order to overcome this problem, a new concept of corrosion and diffusion structure was applied. Eventually, it overcomes various problems and successfully applied it to a real vehicle. The idea was actually reflected in the "Santa Fe" model. Later, the media focused on the lamps to which the idea was applied, and contributed to the sale of a large number of vehicles by providing consumers with a new light sensibility. During the research process, it was possible to secure a number of patents and knowledge of new design concepts.

• Journal of Auto-vehicle Safety Association
• /
• v.14 no.1
• /
• pp.45-54
• /
• 2022

Recently, the introduction of Cooperative Intelligent Transport Systems (C-ITS) has been attempted to solve the limitation of only the sensor of the vehicle itself. For example, vehicles traveling at intersections can drive more safely through C-ITS. By using V2X communication of WAVE and LTE, the driver can receive the status and time of traffic lights. However, LTE has a larger transmission delay time than WAVE, so timimg data may not match in real time. In this paper, using the SPaT message, it was confirmed that LTE has a larger C-ITS service transmission delay time than WAVE. Finally, it was confirmed that the timing data of SPaT provided by LTE corrected by the algorithm is similar to SPaT provided by WAVE. It was confirmed that safer intersection driving is possible based on real-time.

• Journal of Auto-vehicle Safety Association
• /
• v.14 no.1
• /
• pp.55-61
• /
• 2022

Along with global environmental issues, the size of the electric vehicle market has recently skyrocketed. Various efforts have been made to extend mileage, one of the biggest problems of the electric vehicles, and development of batteries with high energy densities has led to exponential growth in mileage and performance. However, proper thermal management is essential because these high-performance batteries are affected by continuous heat generation and can cause fires due to thermal runaway phenomena. Therefore, thermal management of the battery is studied through the optimal design of the guide vanes, while utilizing the existing battery casing to ensure the safety of the electric vehicles. A battery from T-company, one of a manufacturer of the electric vehicles, was used for the research, and the commercial CFD software, ANSYS CFX V20.2, was used for analysis. The guide vanes were derived through optimal design based on a genetic algorithm with flow analysis. The optimized guide vanes show improved heat removal performance.

• Journal of Auto-vehicle Safety Association
• /
• v.15 no.2
• /
• pp.6-12
• /
• 2023

For the commercialization and standardization of autonomous vehicles, demand for rigorous safety criteria has been increased over the world. In Korea, the number of extraordinary service permission for automated vehicles has risen since Hyundai Motor Company got its initial license in March 2016. Nevertheless, licensing standards and evaluation factors are still insufficient for operating on public roadways. To assure driving safety, it is significant to verify whether or not the vehicle's decision is similar to human driving. This paper validates the safety of the autonomous vehicle by drawing scenario-based comparisons between manual driving and autonomous driving. In consideration of real traffic situations and safety priority, seven scenarios were chosen and classified into basic and advanced scenarios. All scenarios and safety factors are constructed based on existing ADAS requirements and investigated via a computer simulation and actual experiment. The input data was collected by an experimental vehicle test on the SNU FMTC test track located at Siheung. Then the offline simulation was conducted to verify the output was appropriate and comparable to the manual driving data.

• Journal of Auto-vehicle Safety Association
• /
• v.15 no.4
• /
• pp.39-47
• /
• 2023

Research on the safety of autonomous vehicle is being conducted in various countries, including the European Union, and computer simulation techniques so called 'Virtual Tool Chain' are mainly used. As part of the crash safety study of autonomous vehicle, 25 car to car collision scenarios were provided as a result of a real accident-based accident reproduction analysis study conducted by a domestic research institution, and a vehicle crash analysis was performed using the FE car to car model of the Honda Accord. In order to analyze the results of the car to car simulation and to construct a database, major crash parameters were selected as impact speed, angle, location, and overlap, and a method of defining them in an indexed form was presented. In order to compare the crash severity of each scenario, a value obtained by integrating the resultant acceleration measured by the ACU of the vehicle was applied. The equivalent collision test mode was derived by comparing the crash severity of the regulation test mode, 30 deg rigid barrier mode, in the same way.

• Journal of Auto-vehicle Safety Association
• /
• v.15 no.4
• /
• pp.95-101
• /
• 2023

As the popularization of autonomous vehicles is anticipated, it is expected that the variety of passenger postures will diversify. However, the current vehicle safety system is expected to be inadequate for accommodating these diverse passenger postures, particularly in reclined positions where severe injuries have been reported in frontal collisions. Therefore, it is necessary to investigate the biomechanical responses and tolerances of occupants in reclined postures. In this study, the behavior and injuries of a Hybrid-III dummy model in a reclined position are analyzed through frontal collision sled simulations equipped with the semi-rigid seat provided by the previous study, three-point safety belt with pretensioner and load limiter, and airbag models. The results are evaluated by comparing thouse reponses with post-mortem human surrogate (PMHS) data, and the findings are expected to be applicable to the basic design of a new restraint system suitable for various postures in autonomous vehicles.

• Journal of Auto-vehicle Safety Association
• /
• v.15 no.4
• /
• pp.16-22
• /
• 2023

This paper presents a proactive longitudinal motion planning algorithm for improving the safety of an automated bus. Since the field of view (FOV) of the autonomous vehicle was limited depending on onboard sensors' performance and surrounding environments, it was necessary to implement vehicle-to-vehicle (V2V) communication for overcoming the limitation. After a virtual V2V-equipped target was constructed considering information obtained from V2V communication, the reference motion of the ego vehicle was determined by considering the state of both the V2V-equipped target and the sensor-detected target. Model predictive control (MPC) was implemented to calculate the optimal motion considering the reference motion and the chance constraint, which was deduced from manual driving data. The improvement in driving safety was confirmed through vehicle tests along actual urban roads.

• Journal of Auto-vehicle Safety Association
• /
• v.16 no.2
• /
• pp.20-26
• /
• 2024

The delivery using drones has been attracting attention because it can innovatively reduce the delivery time from the time of order to completion of delivery compared to the current delivery system, and there have been pilot projects conducted for safe drone delivery. However, the current drone delivery system has the disadvantage of limiting the operational efficiency offered by fully autonomous delivery drones in that drones mainly deliver goods to pre-set landing sites or delivery bases, and the final delivery is still made by humans. In this paper, to overcome these limitations, we propose obstacle detection and landing site selection algorithm based on a vision sensor that enables safe drone landing at the delivery location of the product orderer, and experimentally prove the possibility of station-to-door delivery. The proposed algorithm forms a 3D map of point cloud based on simultaneous localization and mapping (SLAM) technology and presents a grid segmentation technique, allowing drones to stably find a landing site even in places without prior information. We aims to verify the performance of the proposed algorithm through streaming data received from the drone.

• Journal of Auto-vehicle Safety Association
• /
• v.16 no.2
• /
• pp.44-50
• /
• 2024

The windshield of a vehicle plays an important role in ensuring driver safety and maintaining visibility. To prevent issues such as frost and mist from occurring inside and outside the vehicle, research related to the defrosting performance of the windshield is being conducted. Evaluating defrosting performance requires accurate thermal flow analyses. Therefore, in this study, a defrosting duct was constructed within a chamber at an actual vehicle scale to evaluate its performance, and a finite element model was developed and verified. To evaluate defrosting performance, the temperature of the windshield was measured under condition with a mass flow rate of 0.1 kg/s, which corresponds to that of a typical midsize vehicle. A total of 45 thermocouples were arranged at equal intervals of 9 widths and 5 lengths on the windshield to measure the temperature and compare it with the temperature predicted through finite element analysis. A volume grid was created in the main flow area to ensure accurate thermal flow analyses, and a prism layer was added at the interface between the windshield and fluid. In total, 6 million grid systems were formed. Comparing the temperature fields of the experimental results and the finite element analysis results confirmed a similar defrosting pattern, with an average temperature difference of 0.64K.