New Energy Vehicles and Intelligent Connected Vehicles II
Editor's Note
Intelligent connected vehicles are equipped with advanced on-board sensors, controllers, actuators and other devices, and integrate modern communication and network technologies to realize information sharing between vehicles, roads, people, and clouds to achieve "safe, efficient, comfortable and energy-saving" driving. Although the industry shows a positive trend of comprehensive development, it is facing several technical adjustments on core technology level, including bicycle perception and decision-making, vehicle-road cooperation, human-machine co-driving, and various information security threats. In this context, it is crucial to carry out the technological innovation of intelligent networked vehicles.
This new issue “New Energy Vehicles and Intelligent Connected Vehicles II”, which has continued and extended the content of special issue “New Energy Vehicles and Intelligent Connected Vehicles I”(Vol.7, No.6 | 2018), presents cutting-edge research progress in key components of new energy vehicles and bicycle intelligence and vehicle-road collaboration of intelligent connected vehicles from the Center for Automotive Electronics at Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, and that of its partners in the past two years.
Guest Editor
Huiyun Li, Professor
Associate Director of the Institute of Advanced Integration Technology, and Director of Automotive Electronics Research Center, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China. Prof. Li’s research focuses on autonomous driving and intelligent connected vehicles.
Article List
2020, 9(5):1-2.
DOI: 10.12146/j.issn.2095-3135.202005000
Abstract:
智能网联汽车是指搭载先进的车载传感器、控制器、执行器等装置，并融合现代通信与网络技术，实现车与车、路、人、云等信息共享，实现“安全、高效、舒适、节能”行驶。虽然该产业呈现出全面发展的良好局面，但回溯到核心技术层面，该领域仍面临若干技术调整，包括单车感知决策、车路协同、人机共驾、以及各类信息安全威胁。在此背景下，进行智能网联汽车科技创新至关重要。
智能网联汽车是指搭载先进的车载传感器、控制器、执行器等装置，并融合现代通信与网络技术，实现车与车、路、人、云等信息共享，实现“安全、高效、舒适、节能”行驶。虽然该产业呈现出全面发展的良好局面，但回溯到核心技术层面，该领域仍面临若干技术调整，包括单车感知决策、车路协同、人机共驾、以及各类信息安全威胁。在此背景下，进行智能网联汽车科技创新至关重要。
2020, 9(5):3-14.
DOI: 10.12146/j.issn.2095-3135.20200509003
Abstract:
Environmental perception system is an important part of unmanned driving technology, which is the premise of the safety and stability of the unmanned vehicle system. At present, the environmental perception technology is mainly focused on the environmental information acquisition under ideal environment, the high-precision recognition of semantic information, and the multi-sensor fusion etc. A comprehensive attack detection and defense system for the unmanned driving systems is still not available. In this paper, the correlation of multi-sensor perception signals in the perception system in both temporal and spatial domains are considered to establish a mathematical model, which is used to connect the information among different sensors, and to detect the attacked sensors. More important, it can be also used to recover the distorted data based on the matrix completion method. The experimental results showed that, the proposed method can detect the attacked sensor effectively and recover the missed information in the attached perception systems.
Environmental perception system is an important part of unmanned driving technology, which is the premise of the safety and stability of the unmanned vehicle system. At present, the environmental perception technology is mainly focused on the environmental information acquisition under ideal environment, the high-precision recognition of semantic information, and the multi-sensor fusion etc. A comprehensive attack detection and defense system for the unmanned driving systems is still not available. In this paper, the correlation of multi-sensor perception signals in the perception system in both temporal and spatial domains are considered to establish a mathematical model, which is used to connect the information among different sensors, and to detect the attacked sensors. More important, it can be also used to recover the distorted data based on the matrix completion method. The experimental results showed that, the proposed method can detect the attacked sensor effectively and recover the missed information in the attached perception systems.
2020, 9(5):15-26.
DOI: 10.12146/j.issn.2095-3135.20200518001
Abstract:
This paper proposes an optimal speed controlling method to reduce the energy consumption of connected and automated vehicles (CAVs). The proposed method considers not only the surrounding vehicles but also the signal phase and timing (SPAT) information. By determining the economic speed for each vehicle in real time based on the instantaneous optimization, energy consumption and signal waiting time of each vehicle can be reduced. To evaluate feasibility of the proposed method, three benchmark methods are introduced and used for comparison on the Vissim/Autonomie co-simulation platform. The results showed that,energy consumption of the proposed method can be reduced by 14.32%, 9.74%, and 73.72%, respectively in comparison with three benchmark methods.
This paper proposes an optimal speed controlling method to reduce the energy consumption of connected and automated vehicles (CAVs). The proposed method considers not only the surrounding vehicles but also the signal phase and timing (SPAT) information. By determining the economic speed for each vehicle in real time based on the instantaneous optimization, energy consumption and signal waiting time of each vehicle can be reduced. To evaluate feasibility of the proposed method, three benchmark methods are introduced and used for comparison on the Vissim/Autonomie co-simulation platform. The results showed that,energy consumption of the proposed method can be reduced by 14.32%, 9.74%, and 73.72%, respectively in comparison with three benchmark methods.
2020, 9(5):27-33.
DOI: 10.12146/j.issn.2095-3135.20200514001
Abstract:
Solid oxide fuel cells (SOFCs) have gained lots of attentions owing to their high energy conversion efficiency, however, because of the complex technology, their application is not mature as compared with other types of fuel cells such as proton-exchange membrane fuel cells and direct methanol fuel cells. The micro-structure is one of important factors on the SOFC performance, therefore, in order to expedite the commercialization of SOFCs, it is crucial to develop an effective method to optimize the complicated microstructure of SOFCs. The experiment of the SOFC performance test is time-consuming and cost-ineffective, thus it is necessary to develop an SOFC simulation model with high reliability to save the time and cost of the micro-structure optimization. This research proposes an artificial neural network (ANN)-based SOFC simulation model according to the experimental data of an anode-supported SOFC performance, in which the polarization characteristics of SOFCs are estimated from their structural characteristics. After training the ANN based on a part of the experimental data, the rest part of data are used to evaluate the effectiveness of the proposed SOFC model. Results show that the proposed SOFC simulation model accurately presents the polarization characteristics of SOFCs according to the structural characteristics, and this indicates that the model is suitable for the micro-structure optimization for SOFCs.
Solid oxide fuel cells (SOFCs) have gained lots of attentions owing to their high energy conversion efficiency, however, because of the complex technology, their application is not mature as compared with other types of fuel cells such as proton-exchange membrane fuel cells and direct methanol fuel cells. The micro-structure is one of important factors on the SOFC performance, therefore, in order to expedite the commercialization of SOFCs, it is crucial to develop an effective method to optimize the complicated microstructure of SOFCs. The experiment of the SOFC performance test is time-consuming and cost-ineffective, thus it is necessary to develop an SOFC simulation model with high reliability to save the time and cost of the micro-structure optimization. This research proposes an artificial neural network (ANN)-based SOFC simulation model according to the experimental data of an anode-supported SOFC performance, in which the polarization characteristics of SOFCs are estimated from their structural characteristics. After training the ANN based on a part of the experimental data, the rest part of data are used to evaluate the effectiveness of the proposed SOFC model. Results show that the proposed SOFC simulation model accurately presents the polarization characteristics of SOFCs according to the structural characteristics, and this indicates that the model is suitable for the micro-structure optimization for SOFCs.
2020, 9(5):34-47.
DOI: 10.12146/j.issn.2095-3135.20200515001
Abstract:
The driving decisions of human drivers have the social intelligence to handle complex conditions in addition to the driving correctness. However, the existing autonomous driving strategies mainly focus on the correctness of the perception-control mapping, which deviates from the driving logic that human drivers follow. To solve this problem, this paper proposes a human-like autonomous driving strategy in an end-toend control framework based on deep deterministic policy gradient (DDPG). By applying rule constraints to the continuous behavior of the agents, an unmanned end-to-end control strategy was established. This strategy can output continuous and reasonable driving behavior that is consistent with the human driving logic. To enhance the driving safety of the end-to-end decision-making scheme, it utilizes the posterior feedback of the policy output to reduce the output rate of dangerous behaviors. To deal with the catastrophic events in the training process, a continuous reward function is proposed to improve the stability of the training algorithm. The results validated in different simulation environments showed that, the proposed human-like autonomous driving strategy has better control performance than the traditional DDPG algorithm. And the improved reward shaping method is more in line with the control strategy to model the catastrophic events of sparse rewards. The optimization expectation of the objective function can be increased by 85.57%. The human-like DDPG autonomous driving strategy proposed in this paper improves the training efficiency of the traditional DDPG algorithm by 21%, the task success rate by 19%, and the task execution efficiency by 15.45%, which significantly reduces collision accidents.
The driving decisions of human drivers have the social intelligence to handle complex conditions in addition to the driving correctness. However, the existing autonomous driving strategies mainly focus on the correctness of the perception-control mapping, which deviates from the driving logic that human drivers follow. To solve this problem, this paper proposes a human-like autonomous driving strategy in an end-toend control framework based on deep deterministic policy gradient (DDPG). By applying rule constraints to the continuous behavior of the agents, an unmanned end-to-end control strategy was established. This strategy can output continuous and reasonable driving behavior that is consistent with the human driving logic. To enhance the driving safety of the end-to-end decision-making scheme, it utilizes the posterior feedback of the policy output to reduce the output rate of dangerous behaviors. To deal with the catastrophic events in the training process, a continuous reward function is proposed to improve the stability of the training algorithm. The results validated in different simulation environments showed that, the proposed human-like autonomous driving strategy has better control performance than the traditional DDPG algorithm. And the improved reward shaping method is more in line with the control strategy to model the catastrophic events of sparse rewards. The optimization expectation of the objective function can be increased by 85.57%. The human-like DDPG autonomous driving strategy proposed in this paper improves the training efficiency of the traditional DDPG algorithm by 21%, the task success rate by 19%, and the task execution efficiency by 15.45%, which significantly reduces collision accidents.
2020, 9(5):103-113.
DOI: 10.12146/j.issn.2095-3135.20200516001
Abstract:
For the permanent magnet synchronous motors, the controlling algorithms are usually complex and the motor parameters identification are difficult. Since the electromagnetic torques are difficult to estimate through mathematical models, which leads to a decline in motor control accuracy and overall performance of the drive system. In this paper, a topological model of the electromagnetic torque network of the motor was investigated based on the dynamic recursive feedback (ELMAN) neural network. At the same time, the neural network is built as a torque observer by the MATLAB / Simulink for accurate estimation of the motor torque. In the experiments, traditional torque calculation method and the back propagation neural network are compared with the proposed approach. In comparison, the proposed torque observer has better performance in both torque estimation accuracy and control precision.
For the permanent magnet synchronous motors, the controlling algorithms are usually complex and the motor parameters identification are difficult. Since the electromagnetic torques are difficult to estimate through mathematical models, which leads to a decline in motor control accuracy and overall performance of the drive system. In this paper, a topological model of the electromagnetic torque network of the motor was investigated based on the dynamic recursive feedback (ELMAN) neural network. At the same time, the neural network is built as a torque observer by the MATLAB / Simulink for accurate estimation of the motor torque. In the experiments, traditional torque calculation method and the back propagation neural network are compared with the proposed approach. In comparison, the proposed torque observer has better performance in both torque estimation accuracy and control precision.
2020, 9(5):48-57.
DOI: 10.12146/j.issn.2095-3135.20200420001
Abstract:
Visualized coverage model is usually applied for the evaluation of site planning and networks robustness in the spatial networks, which can help the users to understand and analyse the spatial data. In this work, a spatial network model of city-wide carparks service capabilities is studied in a visual dynamic coverage way. Firstly, the ParkingRank algorithm was used to model dynamic service capability of each parking lots in real time. Then, an improved weighted Voronoi diagram featuring boundary constraint was introduced to map the ranking value of each carpark to the geospatial space, forming a seamless and nonoverlap coverage over the city. Advantage of the proposed model is that, the coverage scope of each carpark can be adjusted dynamically according to its changing ranking value. By visualizing the parking lots network, it can provide great helps for the city-wide parking management.
Visualized coverage model is usually applied for the evaluation of site planning and networks robustness in the spatial networks, which can help the users to understand and analyse the spatial data. In this work, a spatial network model of city-wide carparks service capabilities is studied in a visual dynamic coverage way. Firstly, the ParkingRank algorithm was used to model dynamic service capability of each parking lots in real time. Then, an improved weighted Voronoi diagram featuring boundary constraint was introduced to map the ranking value of each carpark to the geospatial space, forming a seamless and nonoverlap coverage over the city. Advantage of the proposed model is that, the coverage scope of each carpark can be adjusted dynamically according to its changing ranking value. By visualizing the parking lots network, it can provide great helps for the city-wide parking management.
2020, 9(5):93-102.
DOI: 10.12146/j.issn.2095-3135.20200515002
Abstract:
Research in current autonomous driving domain mainly focused on the problems of perception, decision-making and control based on single autonomous vehicle, but the interactions and games among different vehicles are usually ignored. That makes exiting techniques inapplicable to reduce the accident rate and to improve the traffic efficiency of the transportation system. To solve this problem, a decision-making emergence method is proposed for the large-scale autonomous driving system based on the principle of coevolutionary games. We have established a grid road model and a vehicle kinematics model in which each vehicle interacts by indirect interaction. Benefited from the distributed algorithms and the communication method between vehicles, the computational complexity can be kept linear with the simulated vehicle volume. By designing a multi-objectives reward function, and making the co-evolution process in a simulated environment, the emergence of dominant driving strategies can be observed efficiently. Experimental results showed that the accidents rate and the average computation speed can be greatly improved compared with conventional approach. In details, the accident rate can be reduced by 90% and the average speed can be increased by 30%. The proposed method have great potentials to explore the optimal driving strategy for urban traffic up to millions of autonomous vehicles.
Research in current autonomous driving domain mainly focused on the problems of perception, decision-making and control based on single autonomous vehicle, but the interactions and games among different vehicles are usually ignored. That makes exiting techniques inapplicable to reduce the accident rate and to improve the traffic efficiency of the transportation system. To solve this problem, a decision-making emergence method is proposed for the large-scale autonomous driving system based on the principle of coevolutionary games. We have established a grid road model and a vehicle kinematics model in which each vehicle interacts by indirect interaction. Benefited from the distributed algorithms and the communication method between vehicles, the computational complexity can be kept linear with the simulated vehicle volume. By designing a multi-objectives reward function, and making the co-evolution process in a simulated environment, the emergence of dominant driving strategies can be observed efficiently. Experimental results showed that the accidents rate and the average computation speed can be greatly improved compared with conventional approach. In details, the accident rate can be reduced by 90% and the average speed can be increased by 30%. The proposed method have great potentials to explore the optimal driving strategy for urban traffic up to millions of autonomous vehicles.
2020, 9(5):69-80.
DOI: 10.12146/j.issn.2095-3135.20200531001
Abstract:
Road geometry information is an important information source in the autonomous driving perception system, which also plays an important role in the subsequent route planning. To realize the autonomous driving perception while the lane line is invisible and the signal of global positioning system is not available, a road geometry estimation based on the leading vehicle is proposed in this work. By modeling the relationship between the current vehicle, the preceding vehicle and the road, we can obtain the system motion model and the observation model. Then, the unscented Kalman filter framework is applied to filter the observed relative position, relative speed, and relative angel of the preceding vehicle and the angular velocity of the host vehicle, for estimating the curvature of current road. The experimental results on the simulation software car learning to act (Carla) showed that, in congested scenarios where lane line targets cannot be obtained and host vehicle cannot be accurately located, road geometry accuracy by the proposed method can be greatly improved in comparison with conventional map matching methods.
Road geometry information is an important information source in the autonomous driving perception system, which also plays an important role in the subsequent route planning. To realize the autonomous driving perception while the lane line is invisible and the signal of global positioning system is not available, a road geometry estimation based on the leading vehicle is proposed in this work. By modeling the relationship between the current vehicle, the preceding vehicle and the road, we can obtain the system motion model and the observation model. Then, the unscented Kalman filter framework is applied to filter the observed relative position, relative speed, and relative angel of the preceding vehicle and the angular velocity of the host vehicle, for estimating the curvature of current road. The experimental results on the simulation software car learning to act (Carla) showed that, in congested scenarios where lane line targets cannot be obtained and host vehicle cannot be accurately located, road geometry accuracy by the proposed method can be greatly improved in comparison with conventional map matching methods.
2020, 9(5):58-68.
DOI: 10.12146/j.issn.2095-3135.20200511001
Abstract:
Simultaneous localization and mapping (SLAM) technique is sensitive to the environments with laser interference or structural similarity, which usually cause the closed-loop misdetection. To solve this problem, this study proposed a closed-loop coarse matching and geomagnetic feature screening closed-loop detection algorithm. By adding a geomagnetic matching algorithm to the closed-loop detection link to further filter the candidate closed-loop detection pose node set, the false detection phenomenon of traditional lidar closed-loop detection can be reduced. It can also correct the false detection and reflection caused by reflection and transmission interference in the positioning and mapping environment, as well as the map image distortion. This study verified the performance of the algorithm through the lidar point cloud and geomagnetic signal data sets collected in the real environment. Compared with traditional lidar SLAM methods, the proposed method outperformed in both matching speed and accuracy. Compared with Google Cartographer algorithm, the algorithm can improve the closed-loop detection speed by 31%, and the false detection rate of closed-loop detection can be reduced by 23% under the condition of 0.8 recall rate. This research expands the application scenarios of SLAM technology, so that the lidar SLAM has better positioning and mapping effects in the scene contains laser interference.
Simultaneous localization and mapping (SLAM) technique is sensitive to the environments with laser interference or structural similarity, which usually cause the closed-loop misdetection. To solve this problem, this study proposed a closed-loop coarse matching and geomagnetic feature screening closed-loop detection algorithm. By adding a geomagnetic matching algorithm to the closed-loop detection link to further filter the candidate closed-loop detection pose node set, the false detection phenomenon of traditional lidar closed-loop detection can be reduced. It can also correct the false detection and reflection caused by reflection and transmission interference in the positioning and mapping environment, as well as the map image distortion. This study verified the performance of the algorithm through the lidar point cloud and geomagnetic signal data sets collected in the real environment. Compared with traditional lidar SLAM methods, the proposed method outperformed in both matching speed and accuracy. Compared with Google Cartographer algorithm, the algorithm can improve the closed-loop detection speed by 31%, and the false detection rate of closed-loop detection can be reduced by 23% under the condition of 0.8 recall rate. This research expands the application scenarios of SLAM technology, so that the lidar SLAM has better positioning and mapping effects in the scene contains laser interference.
2020, 9(5):81-92.
DOI: 10.12146/j.issn.2095-3135.20200509002
Abstract:
This paper presents an optimal pipeline processing method based on multi-FPGA (field programmable gate array) heterogeneous platform. Firstly, the task is divided according to the dichotomy scheme, so that the task quantity can be deployed in each FPGA unit as evenly as possible. And the balance degree of board-level pipeline can be improved. Secondly, the optimization of pipeline structure is applied for the inter-board transmission delay. While the inter-board delay is large, the inter-board delay can be taken as one stage of the pipeline to improve the throughput of the platform. Finally, the computing unit is optimized in parallel, and the FPGA resources are fully utilized by means of data relation rearrangement, loop unroll and loop pipeline, etc. As the result, throughput and energy efficiency of the data processing system can be improved. The AlexNet was used for the experiment to verify the effectiveness of the proposed method. Experimental results showed that, compared with original pipeline structure, throughput of the optimized pipeline structure can be improved by 215.6%, the energy efficiency can be increased by 105.5%, and the running time of a single task can be reduced by 36.6%.
This paper presents an optimal pipeline processing method based on multi-FPGA (field programmable gate array) heterogeneous platform. Firstly, the task is divided according to the dichotomy scheme, so that the task quantity can be deployed in each FPGA unit as evenly as possible. And the balance degree of board-level pipeline can be improved. Secondly, the optimization of pipeline structure is applied for the inter-board transmission delay. While the inter-board delay is large, the inter-board delay can be taken as one stage of the pipeline to improve the throughput of the platform. Finally, the computing unit is optimized in parallel, and the FPGA resources are fully utilized by means of data relation rearrangement, loop unroll and loop pipeline, etc. As the result, throughput and energy efficiency of the data processing system can be improved. The AlexNet was used for the experiment to verify the effectiveness of the proposed method. Experimental results showed that, compared with original pipeline structure, throughput of the optimized pipeline structure can be improved by 215.6%, the energy efficiency can be increased by 105.5%, and the running time of a single task can be reduced by 36.6%.
