Abstract:Purpose
This study aims to propose a novel subjective assessment (SA) method for level 2 or level 2+ advanced driver assistance system (ADAS) with a customized case study in China.
Design/methodology/approach
The proposed SA method contains six dimensions, including perception, driveability and stability, riding comfort, human–machine interaction, driver workload and trustworthiness and exceptional operating case, respectively. And each dimension subordinates several subsections, which describe the correspon… Show more
“…Moving vehicle object detection generally contains different size vehicles that require the accurate and robust detection. The accuracy and rapidity of moving object detection are considered as key factors (Ao and Li, 2022). A CNN is desired to achieve good accuracy but for practical applications, small CNN architectures are desirable with low computational complexity, low memory requirement and low energy consumption.…”
The moving vehicles present different scales in the image due to the perspective effect of different viewpoint distances. The premise of advanced driver assistance system (ADAS) system for safety surveillance and safe driving is early identification of vehicle targets in front of the ego vehicle. The recognition of the same vehicle at different scales requires feature learning with scale invariance. Unlike existing feature vector methods, the normalized PCA eigenvalues calculated from feature maps are used to extract scale-invariant features. This study proposed a convolutional neural network (CNN) structure embedded with the module of multi-pooling-PCA for scale variant object recognition. The validation of the proposed network structure is verified by scale variant vehicle image dataset. Compared with scale invariant network algorithms of Scale-invariant feature transform (SIFT) and FSAF as well as miscellaneous networks, the proposed network can achieve the best recognition accuracy tested by the vehicle scale variant dataset. To testify the practicality of this modified network, the testing of public dataset ImageNet is done and the comparable results proved its effectiveness in general purpose of applications.
“…Moving vehicle object detection generally contains different size vehicles that require the accurate and robust detection. The accuracy and rapidity of moving object detection are considered as key factors (Ao and Li, 2022). A CNN is desired to achieve good accuracy but for practical applications, small CNN architectures are desirable with low computational complexity, low memory requirement and low energy consumption.…”
The moving vehicles present different scales in the image due to the perspective effect of different viewpoint distances. The premise of advanced driver assistance system (ADAS) system for safety surveillance and safe driving is early identification of vehicle targets in front of the ego vehicle. The recognition of the same vehicle at different scales requires feature learning with scale invariance. Unlike existing feature vector methods, the normalized PCA eigenvalues calculated from feature maps are used to extract scale-invariant features. This study proposed a convolutional neural network (CNN) structure embedded with the module of multi-pooling-PCA for scale variant object recognition. The validation of the proposed network structure is verified by scale variant vehicle image dataset. Compared with scale invariant network algorithms of Scale-invariant feature transform (SIFT) and FSAF as well as miscellaneous networks, the proposed network can achieve the best recognition accuracy tested by the vehicle scale variant dataset. To testify the practicality of this modified network, the testing of public dataset ImageNet is done and the comparable results proved its effectiveness in general purpose of applications.
“…First, government subsidies may be introduced because investment costs of green technologies are extremely high, and government subsidies may stimulate liner companies to adopt green technologies [45,46]. Second, the adoption of multiple available new technologies can be investigated to give liner companies more options [35,[47][48][49]. Third, uncertain information [50,51] may also be integrated into the problem in the future.…”
Section: Limitations and Prospects For Further Green Technology Adopt...mentioning
Maritime decarbonization and strict international regulations have forced liner companies to find new solutions for reducing fuel consumption and greenhouse gas emissions in recent years. Green technology is regarded as one of the most promising alternatives to achieve environmental benefits despite its high initial investment costs. Therefore, a scientific method is required to assess the possibility of green technology adoption for liner companies. This study formulates a mixed-integer nonlinear programming model to determine whether to retrofit their ship fleets with green technology and how to deploy ships while taking maritime decarbonization into account. To convert the nonlinear model into a linear model that can be solved directly by off-the-shelf solvers, several linearization techniques are applied in this study. Sensitivity analyses involving the influences of the initial investment cost, fuel consumption reduction rate of green technology, unit fuel cost, and fixed operating cost of a ship on operation decisions are conducted. Green technology may become more competitive when modern technology development makes it efficient and economical. As fuel and fixed operating costs increase, more ships retrofitted with green technology will be deployed on all shipping routes.
“…The sketch map for the vehicle model integrated with impact forces is shown in Figure 3. Referred from Figure 3, the collision forces (F x_impact , F y_impact ) 1 from Equations ( 10) and ( 11) are presumed to be applied to ego vehicle under a horizontal plane with the coordinate (x Q , y Q ), which can be considered as the parameters in V2V impact model. The mathematical equations under Newton second law are described as:…”
Section: Vehicle Dynamics Modelmentioning
confidence: 99%
“…The number of motor vehicles emerges a lot in recent years in China, which generates a societal problem that vehicle‐related accidents also present a big jump [1]. In 2020, about 244,674 traffic accidents happened in mainland China and resulted in about 61,703 deaths according to the data from National Bureau of Statistics of China [2].…”
In terms of the possible secondary or multiple events accidents (MEA) after an initial vehicle to vehicle (V2V) impact, this research proposes an active safety & stability controller for independent-driven electric vehicles towards recovering the vehicle to safety states after an impact. The controller aims to regulate the course angle and lateral deviation that enforce the vehicle back to its original driving path. The upper-level controller is derived based on sliding mode technique. And the low-level controller aims to solve a convex quadratic allocation problem, which inherently incorporates fault-tolerance property. The independent in-wheel-motor (IWM) fault is very likely to happen under a real V2V impact and it contributes much to the driving safety & stability. Two typical real-endcollision scenarios under low-and high longitudinal velocities are designed to verify the proposed controller performance. The low-and high velocities collisions aim to simulate the urban and highway accidents, respectively. Compared to other control methods, i.e. pure braking and static allocation, the RMSEs of safety (lateral deviation, course angle) and stability indexes (yaw rate, sideslip angle) under proposed controller reduce significantly both in urban and highway accidents. Moreover, the controller performs robust enough against the impact-induced IWM fault. It further supports the effectiveness at dealing with the safety and stability of the ego vehicle after an initial impact and prevent possible MEA.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.