New Energy Vehicles and Intelligent Connected Vehicles I
Editor's Note
In recent years, the Chinese government has provided strong support for new energy vehicles and intelligent connected vehicles in terms of scientific and technological research, industrial development, application demonstration, and market promotion. Interestingly, China has become one of the most active countries in the field of new energy vehicles. Although the new energy vehicle industry has shown a good momentum in China, it has to overcome core technological barriers, including technologies in power system, key components, perceptual decision-making, internet of vehicles and system integration. Under such context, it is crucial to develop technological innovation in new energy vehicles and intelligent connected vehicles. This thematic issue presents cutting-edge research in key components such as battery, motor and electric control of new energy vehicles and perceptual decision-making technology of intelligent connected vehicles, as well as data collection and applications on internet of vehicles.
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
1 Preface
2018, 7(6):1-1.
DOI: 10.12146/j.issn.2095-3135.201806000
Abstract:
当前，中国汽车工业发展面临由大国到强国转型的挑战。《中国制造 2025》 将新能源汽车与智能网联汽车列入十大重点发展领域。近年来，国家对新能源汽 车与智能网联汽车从科技研发、产业发展、应用示范和市场推动等方面进行了全 维度的支持，中国成为新能源汽车领域发展最活跃的国家之一。虽然该产业在中 国呈现出全面发展的良好局面，但回溯到核心技术层面，该领域需要突破若干核 心关键技术，包括动力系统技术、关键零部件、感知决策、车联网及系统集成。 在此背景下，进行新能源汽车与智能网联汽车科技创新至关重要。
当前，中国汽车工业发展面临由大国到强国转型的挑战。《中国制造 2025》 将新能源汽车与智能网联汽车列入十大重点发展领域。近年来，国家对新能源汽 车与智能网联汽车从科技研发、产业发展、应用示范和市场推动等方面进行了全 维度的支持，中国成为新能源汽车领域发展最活跃的国家之一。虽然该产业在中 国呈现出全面发展的良好局面，但回溯到核心技术层面，该领域需要突破若干核 心关键技术，包括动力系统技术、关键零部件、感知决策、车联网及系统集成。 在此背景下，进行新能源汽车与智能网联汽车科技创新至关重要。
2018, 7(6):2-8.
DOI: 10.12146/j.issn.2095-3135.20180717003
Abstract:
To control the pore size of anode in thin film solid oxide fuel cells, a simple polymer injection method was introduced to replace conventional anode functional layer method in this paper. Firstly, the liquid polystyrene was pressed and injected to the porous nickel oxide yttria-stabilized zirconia (NiO-YSZ) substrate. Then, a 300-nm-thick film was deposited on a composite substrate by sputtering procedure. Finally,by removing polystyrene from the pore of substrate through pyrolysis, a free-standing thin film on the porous substrate could be obtained. The polymer-injected NiO-YSZ substrate has more controlled pores. Moreover,with the proposed method, throughput of large scale thin film solid oxide fuel cell can be improved.
To control the pore size of anode in thin film solid oxide fuel cells, a simple polymer injection method was introduced to replace conventional anode functional layer method in this paper. Firstly, the liquid polystyrene was pressed and injected to the porous nickel oxide yttria-stabilized zirconia (NiO-YSZ) substrate. Then, a 300-nm-thick film was deposited on a composite substrate by sputtering procedure. Finally,by removing polystyrene from the pore of substrate through pyrolysis, a free-standing thin film on the porous substrate could be obtained. The polymer-injected NiO-YSZ substrate has more controlled pores. Moreover,with the proposed method, throughput of large scale thin film solid oxide fuel cell can be improved.
2018, 7(6):9-18.
DOI: 10.12146/j.issn.2095-3135.20180710001
Abstract:
The parking guidance system (PGS) can alleviate the disordered parking problem in the peak time, and reduce the time to find parking space. But existing PGS techniques are quite dependent to the realtime data and historical data, its performance will be greatly degraded when the data is insufficient. To solve this problem, a data repairing based PGS method was presented. First, by mining the spatial data around the parking lots, a spatial similarity metric of parking lot was proposed. Then, the possibility of parking data similarity was calculated when the parking lots had spatial similarity. If the conditional probability was large enough, the known data of parking lots would be used as the learning samples. Finally, the reparative data could be generated by recurrent generative adversarial networks. Experimental results show that when the parking lots have high spatial similarity, the data of parking lots have high similarity probability also. The data generated by the recurrent generative adversarial networks also have the same distribution with real data. By the proposed method, a large number of reasonable data can be generated efficiently, and the PGS performance can be improved while only few parking data is available.
The parking guidance system (PGS) can alleviate the disordered parking problem in the peak time, and reduce the time to find parking space. But existing PGS techniques are quite dependent to the realtime data and historical data, its performance will be greatly degraded when the data is insufficient. To solve this problem, a data repairing based PGS method was presented. First, by mining the spatial data around the parking lots, a spatial similarity metric of parking lot was proposed. Then, the possibility of parking data similarity was calculated when the parking lots had spatial similarity. If the conditional probability was large enough, the known data of parking lots would be used as the learning samples. Finally, the reparative data could be generated by recurrent generative adversarial networks. Experimental results show that when the parking lots have high spatial similarity, the data of parking lots have high similarity probability also. The data generated by the recurrent generative adversarial networks also have the same distribution with real data. By the proposed method, a large number of reasonable data can be generated efficiently, and the PGS performance can be improved while only few parking data is available.
2018, 7(6):19-30.
DOI: 10.12146/j.issn.2095-3135.20180717002
Abstract:
Current artificial intelligent (AI) engines are usually designed for specific supervised learning algorithms, which have been widely used in computer vision and natural language processing domains etc. However, very few AI engines have been designed to support on-chip reinforcement learning algorithms, which is the foremost algorithm kernel for decision-making subsystem of many autonomous systems. In this work, a coarse-grained reconfigurable array like AI computing engine has been designed for the deployments of both supervised and reinforcement learning through on-chip configuration, action Random Access Memory (RAM) and reward RAM. Logic synthesis at the design frequency of 200 MHz based on 65 nm CMOS technology reveals the physical statistics of the proposed engine of 0.32 mm2 in silicon area, 15.46 mW in power consumption. The proposed on-chip AI engine facilitates the implementation of end-to-end perceptual and decision-making networks, which has great potentials in applications like autonomous driving, robotics and unmanned aerial vehicle.
Current artificial intelligent (AI) engines are usually designed for specific supervised learning algorithms, which have been widely used in computer vision and natural language processing domains etc. However, very few AI engines have been designed to support on-chip reinforcement learning algorithms, which is the foremost algorithm kernel for decision-making subsystem of many autonomous systems. In this work, a coarse-grained reconfigurable array like AI computing engine has been designed for the deployments of both supervised and reinforcement learning through on-chip configuration, action Random Access Memory (RAM) and reward RAM. Logic synthesis at the design frequency of 200 MHz based on 65 nm CMOS technology reveals the physical statistics of the proposed engine of 0.32 mm2 in silicon area, 15.46 mW in power consumption. The proposed on-chip AI engine facilitates the implementation of end-to-end perceptual and decision-making networks, which has great potentials in applications like autonomous driving, robotics and unmanned aerial vehicle.
2018, 7(6):31-38.
DOI: 10.12146/j.issn.2095-3135.20180806001
Abstract:
Accurate estimation of charging state of the power lithium battery is an important function in the battery management system of electric vehicle. In this paper, based on the second-order resistor-capacitance equivalent circuit model, an accurate charging state estimation of power lithium battery was investigated. State space expression was established firstly, and the parameters of equivalent circuit model were identified by the least square method. The relationship between open circuit voltage and residual charge was fitted by polynomial fitting method. By the usage of cubature Kalman filter, the state of charge of lithium battery was estimated at the same time. In the experiment, a digital signal processor-based charge and discharge platform was constructed. And the experimental results show that, the cubature Kalman filtering algorithm can achieve real-time online estimation, and the maximum error is less than 2%, which has high estimation accuracy.
Accurate estimation of charging state of the power lithium battery is an important function in the battery management system of electric vehicle. In this paper, based on the second-order resistor-capacitance equivalent circuit model, an accurate charging state estimation of power lithium battery was investigated. State space expression was established firstly, and the parameters of equivalent circuit model were identified by the least square method. The relationship between open circuit voltage and residual charge was fitted by polynomial fitting method. By the usage of cubature Kalman filter, the state of charge of lithium battery was estimated at the same time. In the experiment, a digital signal processor-based charge and discharge platform was constructed. And the experimental results show that, the cubature Kalman filtering algorithm can achieve real-time online estimation, and the maximum error is less than 2%, which has high estimation accuracy.
2018, 7(6):39-48.
DOI: 10.12146/j.issn.2095-3135.20180710002
Abstract:
Prediction of available parking spaces is the critical technique in the intelligent parking guidance system. The prediction technology based on neural network can achieve high accuracy in short-term prediction. And existing techniques can reach an average absolute prediction error of about 10. However, with the increase of prediction steps or time-span, the prediction accuracy will decrease dramatically. To solve this problem, a prediction method that can keep the characteristics of data changes in the long-span is introduced in this paper. The method uses the fuzzy information granulation to obtain the feature data sets. Then, a long shortterm memory network is trained to predict the future feature data sets. Finally, an interpolation procedure is applied to reconstruct the curve of the parking space. The simulation results show that the proposed method can achieve better prediction accuracy and higher computation efficiency when compared with traditional prediction methods.
Prediction of available parking spaces is the critical technique in the intelligent parking guidance system. The prediction technology based on neural network can achieve high accuracy in short-term prediction. And existing techniques can reach an average absolute prediction error of about 10. However, with the increase of prediction steps or time-span, the prediction accuracy will decrease dramatically. To solve this problem, a prediction method that can keep the characteristics of data changes in the long-span is introduced in this paper. The method uses the fuzzy information granulation to obtain the feature data sets. Then, a long shortterm memory network is trained to predict the future feature data sets. Finally, an interpolation procedure is applied to reconstruct the curve of the parking space. The simulation results show that the proposed method can achieve better prediction accuracy and higher computation efficiency when compared with traditional prediction methods.
2018, 7(6):49-59.
DOI: 10.12146/j.issn.2095-3135.20180629001
Abstract:
With the improvement of integrated circuit manufacturing technology, the size of electronic components is shrinking accordingly. And that makes the memory components more susceptible to working environment. To solve this problem, this paper presents a memory fault tolerance method based on modular principal component analysis (PCA). Main features of the data were obtained via modular PCA firstly. Then, the feature data is averaged to obtain the best available estimate of the original data. This best available estimate can be used to make fault-tolerant replacements for any faults in the data, minimizing the sum of the squared errors of the fault-tolerant replaced data and the original data. Finally, using the reconstructed block data, fault-tolerant replacement of the erroneous data in the original data block can be performed. The experimental results show that the picture data can keep a peak signal to noise ratio of more than 30 dB under 0.003 5 error rate. In comparison with conventional error correcting code approach, the execution time can be reduced about 40%, and the memory occupancy can be reduced about 12%.
With the improvement of integrated circuit manufacturing technology, the size of electronic components is shrinking accordingly. And that makes the memory components more susceptible to working environment. To solve this problem, this paper presents a memory fault tolerance method based on modular principal component analysis (PCA). Main features of the data were obtained via modular PCA firstly. Then, the feature data is averaged to obtain the best available estimate of the original data. This best available estimate can be used to make fault-tolerant replacements for any faults in the data, minimizing the sum of the squared errors of the fault-tolerant replaced data and the original data. Finally, using the reconstructed block data, fault-tolerant replacement of the erroneous data in the original data block can be performed. The experimental results show that the picture data can keep a peak signal to noise ratio of more than 30 dB under 0.003 5 error rate. In comparison with conventional error correcting code approach, the execution time can be reduced about 40%, and the memory occupancy can be reduced about 12%.
2018, 7(6):60-68.
DOI: 10.12146/j.issn.2095-3135.20180723001
Abstract:
The interior permanent magnet synchronous machines have advantages of high power density, high reliability, field weakening performance etc. However, subject to the nonlinear characteristics of motor parameters, accurate estimation of the electromagnetic motor torque is very difficult. In this paper, a convolutional neural network based electromagnetic torque estimation method, i.e., a torque observer is investigated. Training data of the convolutional neural network are collected from the simulations of a high fidelity nonlinear interior permanent magnet synchronous machine by the means of finite element analysis. Then, a control scheme is adopted to control the interior permanent magnet synchronous machines with the proposed torque observer. In order to reduce the torque estimation error, different parameters and structures of the convolutional neural network are compared. Experimental results show that the proposed convolutional neural network based torque observer can estimate the electromagnetic torque accurately.
The interior permanent magnet synchronous machines have advantages of high power density, high reliability, field weakening performance etc. However, subject to the nonlinear characteristics of motor parameters, accurate estimation of the electromagnetic motor torque is very difficult. In this paper, a convolutional neural network based electromagnetic torque estimation method, i.e., a torque observer is investigated. Training data of the convolutional neural network are collected from the simulations of a high fidelity nonlinear interior permanent magnet synchronous machine by the means of finite element analysis. Then, a control scheme is adopted to control the interior permanent magnet synchronous machines with the proposed torque observer. In order to reduce the torque estimation error, different parameters and structures of the convolutional neural network are compared. Experimental results show that the proposed convolutional neural network based torque observer can estimate the electromagnetic torque accurately.
2018, 7(6):69-80.
DOI: 10.12146/j.issn.2095-3135.20180710003
Abstract:
Traditional differential global positioning system usually demands high-definition map to realize automatic path tracking. However, the high-definition map occupies large storage space and requires highperformance on-board computer or large communication bandwidth. In this paper, a low-cost vector mapbased navigation method for autonomous vehicle is introduced. By recording the vector map offline, the method initializes an optimal global route by giving any starting and ending position on the map. During runtime, the in-vehicle computer filters the real-time positioning data from differential global positioning system and tracks the planned path according to geometric rules. While any obstacles are detected, the system can adjust the local path automatically. In addition, the proposed method can calculate the angle of steering wheel, levels of throttle, brake pedals in real-time, and transmit these actuation commands through controller area network bus interface. Feasibility and accuracy of the proposed navigation method is verified on our autonomous vehicle testing platform.
Traditional differential global positioning system usually demands high-definition map to realize automatic path tracking. However, the high-definition map occupies large storage space and requires highperformance on-board computer or large communication bandwidth. In this paper, a low-cost vector mapbased navigation method for autonomous vehicle is introduced. By recording the vector map offline, the method initializes an optimal global route by giving any starting and ending position on the map. During runtime, the in-vehicle computer filters the real-time positioning data from differential global positioning system and tracks the planned path according to geometric rules. While any obstacles are detected, the system can adjust the local path automatically. In addition, the proposed method can calculate the angle of steering wheel, levels of throttle, brake pedals in real-time, and transmit these actuation commands through controller area network bus interface. Feasibility and accuracy of the proposed navigation method is verified on our autonomous vehicle testing platform.
2018, 7(6):81-87.
DOI: 10.12146/j.issn.2095-3135.20180724001
Abstract:
The anode and the cathode of the thin film solid oxide fuel cells are usually fabricated as the mixed-ionic-electronic-conducting (MIEC) porous electrode in order to increase the triple phase boundaries where the electrochemical reactions occur. In this research, Ni, which is the catalyst of the hydrogen oxidation reaction in the anode, and scandia-stabilized zirconia, which presents a high oxygen ion conductivity, are mixed and finally fabricated as the MIEC anode. Experimental results show that triple phase boundaries can be increased and the polarization resistance can be decreased in comparison with the usage of pure Ni anode. However, the increasing of triple phase boundaries also causes the improvement of ohmic resistance, and the overall performance of fuel cell is still decreased.
The anode and the cathode of the thin film solid oxide fuel cells are usually fabricated as the mixed-ionic-electronic-conducting (MIEC) porous electrode in order to increase the triple phase boundaries where the electrochemical reactions occur. In this research, Ni, which is the catalyst of the hydrogen oxidation reaction in the anode, and scandia-stabilized zirconia, which presents a high oxygen ion conductivity, are mixed and finally fabricated as the MIEC anode. Experimental results show that triple phase boundaries can be increased and the polarization resistance can be decreased in comparison with the usage of pure Ni anode. However, the increasing of triple phase boundaries also causes the improvement of ohmic resistance, and the overall performance of fuel cell is still decreased.
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。
