An Autonomous Driving Approach Based on Trajectory Learning Using Deep Neural Networks

Wang Dan,Wang Canye,Wang Yulong,Wang Hang,Pei Feng 한국자동차공학회 2021 International journal of automotive technology Vol.22 No.6

Autonomous driving approaches today are mainly based on perception-planning-action modular pipelines and the End2End paradigm respectively. The End2End paradigm is a strategy that directly maps raw sensor data to vehicle control actions. This strategy is very promising and appealing because complex module design and cumbersome data labeling are avoided. Since this approach lacks a degree of interpretability, safety and practicability. we propose an autonomous driving approach based on trajectory learning using deep neural networks in this paper. In comparison to End2End algorithm, it is found that the trajectory learning algorithm performs better in autonomous driving. As for trajectory learning algorithm, the CNN_Raw-RNN network structure is established, which is verified to be more effective than the original CNN_LSTM network structure. Besides, we propose an autonomous driving architecture of a pilot and copilot combination. The pilot is responsible for trajectory prediction via imitation learning with labeled driving trajectories, while the copilot is a safety module that is employed to verify the effectiveness of the vehicle trajectory by the results of the semantic segmentation auxiliary task. The proposed autonomous driving architecture is verified with a real car on urban roads without manual intervention within 40 km.

A putative plastidial adenine nucleotide transporter, BRITTLE1-3, plays an essential role in regulating chloroplast development in rice (Oryza sativa L.)

Jia Lyu,Yihua Wang,Linglong Liu,Chunming Wang,Yulong Ren,Cheng Peng,Feng Liu,Yunlong Wang,Mei Niu,Di Wang,Ming Zheng,Kunneng Zhou,Shaolu Zhao,Fuqing Wu,Haiyang Wang,Jianmin Wan 한국식물학회 2017 Journal of Plant Biology Vol.60 No.5

Differentiation from proplastids into chloroplasts isa light- and energy-dependent process. How this process isregulated is still poorly understood at the molecular level. We herein report a new putative plastidial adenine nucleotidetransporter, BRITTLE1-3 (referred to as OsBT1-3), encoded bythe rice (Oryza sativa) White Stripe Leaf 2 (WSL2) gene. Loss of OsBT1-3 function results in defective chloroplastbiogenesis, severely reduced photosynthetic efficiency, andfinally a white stripe leaf phenotype in the first four leaves. The expression levels of genes related to chlorophyllbiosynthesis and photosynthesis are drastically reduced,accompanied with over accumulation of reactive oxygenspecies (ROS) in the wsl2 mutant. OsBT1-3 is targeted tothe chloroplasts and it expresses in almost all tissues inplants, especially in young leaves. OsBT1-3 consists of 419amino acids and exhibits features of all mitochondrialcarrier proteins, including a typical transmembrane-spanningdomain and a highly conserved sequence motif designatedas the ‘mitochondrial energy transfer signatures’. Phylogeneticanalysis shows that OsBT1-3 is a putative plastidialadenine nucleotide transporter and is most closely relatedto ZmBT1-2. Together, these observations suggest that thenew putative adenine nucleotide transporter, OsBT1-3,plays an essential role in regulating chloroplast biogenesisand maintenance of ROS homeostasis during rice seedlingde-etiolation.

Weather optimal area-keeping control for underactuated autonomous surface vehicle with input time-delay

Wang Shasha,Tuo Yulong,Wang Dan 대한조선학회 2022 International Journal of Naval Architecture and Oc Vol.14 No.1

The problem of the underactuated Autonomous Surface Vehicle (ASV) area-keeping control under unknown slow time-varying external environment disturbances and input time-delay, an integral timedelay sliding mode control algorithm based on weather optimal intermittent strategy is proposed in this article. Firstly, integral time-delay sliding mode control law is constructed, with controller parameters chosen according to the upper bound of time-delay. In the control law, a geometric updating rule is designed for the virtual suspension point, which enables ASV at a particular small area with a weather optimal heading. And, weather optimal intermittent strategy is developed to extend the area-keeping control method, while reducing ASV energy consumption. Besides, the closed-loop system is globally asymptotically stable through Lyapunov-Krasovskii analysis. Finally, the performance of the proposed algorithm is illustrated through the simulation examples.

PRR11 and SKA2 gene pair is overexpressed and regulated by p53 in breast cancer

Wang, Yitao,Zhang, Chunxue,Mai, Li,Niu, Yulong,Wang, Yingxiong,Bu, Youquan Korean Society for Biochemistry and Molecular Biol 2019 BMB Reports Vol.52 No.2

Our previous study found that two novel cancer-related genes, PRR11 and SKA2, constituted a classic gene pair that was regulated by p53 and NF-Y in lung cancer. However, their role and regulatory mechanism in breast cancer remain elusive. In this study, we found that the expression levels of PRR11 and SKA2 were upregulated and have a negative prognotic value in breast cancer. Loss-of-function experiments showed that RNAi-mediated knockdown of PRR11 and/or SKA2 inhibited proliferation, migration, and invasion of breast cancer cells. Mechanistic experiments revealed that knockdown of PRR11 and/or SKA2 caused dysregulation of several downstream genes, including CDK6, TPM3, and USP12, etc. Luciferase reporter assays demonstrated that wild type p53 significantly repressed the PRR11-SKA2 bidirectional promoter activity, but not NF-Y. Interestingly, NF-Y was only essential for and correlated with the expression of PRR11, but not SKA2. Consistently, adriamycin-induced (ADR) activation of endogenous p53 also caused significant repression of the PRR11 and SKA2 gene pair expression. Notably, breast cancer patients with lower expression levels of either PRR11 or SKA2, along with wild type p53, exhibited better disease-free survival compared to others with p53 mutations and/or higher expression levels of either PRR11 or SKA2. Collectively, our study indicates that the PRR11 and SKA2 transcription unit might be an oncogenic contributor and might serve as a novel diagnostic and therapeutic target in breast cancer.

Determination of the air-kerma in mammography x-rays at NIM

Ji Wang,Siming Guo,Pingzhou Lu,Jinjie Wu,Ying Liu,Ruonan Li,Yulong Wang 한국물리학회 2022 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.81 No.4

Breast cancer has become the top killer of women’s health in the world, and mammography X-ray has become the gold standard in various diagnostic methods, due to its low energy, low dose and high resolution. To meet the quality control requirements of mammography X-rays dose in clinical, the National Institute of Metrology (NIM,China) has established a mammography X-ray reference radiation facility. The radiation quality half-value layer (HVL) range from 0.274 to 0.364 mm (aluminum) meets the requirements of IEC61267, and a new parallel-plate free-air ionization is designed. At NIM, various correction factors were measured and simulated, and fnally, the absolute measurement of the X-ray air kerma rate of the mammography was realized. By analysis, the combined standard uncertainty of the air kerma is 0.44%. Provide traceability services for the use and quality control of mammography in China.

Effect of Grinding Parameters on Industrial Robot Grinding of CFRP and Defect Formation Mechanism

Fangyuan Wang,Shanyong Xuan,Zongyu Chang,Kai Jin,Yulong Gao,Hao Wang,Qiye Song 한국정밀공학회 2024 International Journal of Precision Engineering and Vol.11 No.2

The use of industrial robots for grinding CFRP is a green processing method. This method not only allows in-situ repair to reduce unnecessary waste of resources, but also produces no excessive contaminants. The effect of various process parameters, including grinding directions, the mesh size of grinding heads and rotating speed, on the grinding quality of Carbon Fiber Reinforced Polymers (CFRP) using industrial robots was investigated. The mechanism of grinding defects was also studied. According to the experimental results, the CFRP grinding process is mainly controlled by the rotating speed, number of grinding heads, and grinding direction. In particular, high-speed grinding helps to improve the surface quality of CFRP. In turn, the use of diamond grinding heads with too small or too large particles may reduce surface quality. Grinding quality changes with the grinding direction. In the grinding direction between 0° and 90°, the surface roughness increases with the angle (but drops at 60°), and The same trend is observed in the grinding direction between 90° and 150°, whereby the surface roughness increases with the angle (but drops at 120°). The surface quality of CFRP is thereby improved after grinding in the direction of 0°, 60°, 120° and 180°. Furthermore, the fiber pull-out occurs, when the feed direction and fiber orientation are aligned. Finally, the low-frequency vibration easily causes fiber pull-out defects.

FLOURY ENDOSPERM12 Encoding Alanine Aminotransferase 1 Regulates Carbon and Nitrogen Metabolism in Rice

Mingsheng Zhong,Xi Liu,Feng Liu,Yulong Ren,Yunlong Wang,Jianping Zhu,Xuan Teng,Erchao Duan,Fan Wang,Huan Zhang,Mingming Wu,Yuanyuan Hao,Xiaopin Zhu,Ruonan Jing,Xiuping Guo,Ling Jiang,Yihua Wang,Jianmi 한국식물학회 2019 Journal of Plant Biology Vol.62 No.1

Starch is a major storage substance in cerealgrains, and starch biosynthesis is a complex process. In orderto elucidate regulation of the starch biosynthesis pathway, wescreened a series of rice (Oryza sativa L.) endospermmutants. In this study, we identified a floury white-coreendosperm mutant named floury endosperm12 (flo12). Theflo12 mutant exhibited loosely packed starch granules and alower thousand kernel weight compared to wild type. Semithinsections revealed that compound starch grains (SG) inflo12 interior endosperm cells were developed abnormally. Furthermore, amylose content was decreased, while totalprotein content was significantly increased in flo12 grains. Map-based cloning showed that FLO12 encodes rice alanineaminotransferase 1 (OsAlaAT1). OsAlaAT1 is highly expressedin developing endosperm. Subcellular localization showedthat OsAlaAT1 is localized in the cytosol. Moreover, theexpression of most starch synthesis-related genes wasdecreased, while most of the storage protein coding geneshad elevated expression levels in the flo12 mutant. Inaddition, overexpression of the OsAlaAT1 gene increasedgrain weight. In brief, we demonstrated that OsAlaAT1regulates carbon and nitrogen metabolism, which provides anew insight for the improvement of rice quality and yield.

Robust Adaptive Dynamic Surface Control Based on Structural Reliability for a Turret-moored Floating Production Storage and Offloading Vessel

Yulong Tuo,Yuanhui Wang,Simon X. Yang,Mohammad Biglarbegian,Mingyu Fu 제어·로봇·시스템학회 2018 International Journal of Control, Automation, and Vol.16 No.4

For floating production storage and offloading (FPSO) vessels, a dynamic positioning controller is necessary because using only a mooring system is not possible to keep the ship within a predefined region. Position control of the FPSO vessel is extremely challenging due to model uncertainties and unknown control coefficients. This paper develops a new robust adaptive positioning controller consisting of several components: adaptive law, dynamic surface control (DSC) technology, sigmoid tracking differentiator (STD), Nussbaum gain function, and structural reliability index. Model uncertainties can be estimated by the adaptive law derived from the Lyapunov theory. The DSC technology is used to eliminate repeated differentiation by introducing first-order filtering of the virtual control. The chattering-free STD with the characteristics of global fast convergence can estimate the derivatives of model uncertainties that are difficult to calculate directly. Therefore, the DSC and STD techniques make the proposed controller simpler to compute and easier to implement in engineering practice. Most of the traditional controllers require the information about the control coefficients to guarantee the stability of the closed-loop system while the Nussbaum gain function can remove the requirement for a priori knowledge of the sign of control coefficients. The capacity of the mooring system can be fully utilized to position the FPSO vessel by adjusting the structural reliability index on the premise of ensuring the safety of mooring lines, and hence less control effort is needed for the positioning controller. Simulations using two sets of system parameters demonstrate the proposed controller’s effectiveness. In addition, a qualitative comparison with the adaptive backstepping controller shows that our proposed controller is computationally more efficient and does not require a priori knowledge of the sign of control coefficients. A quantitative comparison with robust adaptive controller without the structural reliability shows that less control effort is needed using our proposed controller.

Optimal Resource Allocation for Anti-terrorism in Protecting Overpass Bridge Based on AHP Risk Assessment Model

Yulong Li,Tao Wang,Xinyi Song,Guijun Li 대한토목학회 2016 KSCE JOURNAL OF CIVIL ENGINEERING Vol.20 No.1

Terrorist attacks on infrastructures are serious incidents in many countries. Measures to prevent terrorist attacks and minimize the damage are critical in the design of infrastructures. This paper develops an Analytic Hierarchy Process (AHP) based risk assessment model to identify the key components of the overpass bridge in terrorist attacks, and prioritizes a variety of possible attack means with multiple criteria. The AHP risk assessment model can be used to not only set the priority of protected components, but also to allocate resources to the most critical protecting methods within a limited budget. An interchange overpass bridge in Beijing, China is analysed to validate the model. The results show that the method is suitable for the optimal resource allocation in the process of the overpass bridge anti-terrorism design, the formulation of a defence strategy and the protection resource allocation scheme for the overpass bridge against a terrorist attack. The method can also be used for dynamic allocating the resources for the overpass bridge under the early warning of terrorist attacks.

Evaluation of the susceptibility of coal to spontaneous combustion by a TG profile subtraction method

Yulong Zhang,Sheng Xue,Junfeng Wang,Yue Wu,Zhengfeng Li,Liping Chang 한국화학공학회 2016 Korean Journal of Chemical Engineering Vol.33 No.3

−It is imperative to have an in-depth understanding of the intrinsic reaction between coal and oxygen during low-temperature oxidation, as the reaction is the main source responsible for the self-heating and spontaneous combustion of coal. As low-temperature oxidation of coal involves a series of physical and chemical process and many parallel reactions, it is difficult to directly investigate the intrinsic reaction between coal and oxygen by conventional analytical method. Thermogravimetric analysis (TGA) was used to investigate the intrinsic reaction between coal and oxygen based on the mass change. By means of the subtraction analysis method of TGA, the TG-subtraction curves were obtained by subtracting the TG-N2 curves from the TG-air curves. The results indicate that a TG-subtraction curve can better reflect the intrinsic reaction of coal oxidation than a TG-air curve by eliminating the influence of evaporation of water and thermal decomposition of inherent oxygen-containing groups. In terms of the rate of mass increase, the intrinsic reactions can be divided into three stages: slow oxidation stage, advanced oxidation stage and rapid oxidation stage. The activation energy at each of the stages, obtained by Coats and Redfern’s model, can be used to as a technical parameter to evaluate the proneness of coal spontaneous combustion. The optimum experiment conditions were also developed to study low-temperature coal oxidation with the subtraction method of TGA.