面向智能驾驶的平行视觉感知：基本概念、框架与应用

目的 视觉感知技术是智能车系统中的一项关键技术,但是在复杂挑战下如何有效提高视觉性能已经成为智能驾驶领域的重要研究内容。本文将人工社会（artificial societies）、计算实验（computational experiments）和平行执行（parallel execution）构成的ACP方法引入智能驾驶的视觉感知领域,提出了面向智能驾驶的平行视觉感知,解决了视觉模型合理训练和评估问题,有助于智能驾驶进一步走向实际应用。方法 平行视觉感知通过人工子系统组合来模拟实际驾驶场景,构建人工驾驶场景使之成为智能车视觉感知的“计算实验室”;借助计算实验两种操作模式完成视觉模型训练与评估;最后采用平行执行动态优化视觉模型,保障智能驾驶对复杂挑战的感知与理解长期有效。结果 实验表明,目标检测的训练阶段虚实混合数据最高精度可达60.9%,比单纯用KPC（包括：KITTI（Karlsruhe Institute of Technology and Toyota Technological Institute）,PASCAL VOC（pattern analysis,statistical modelling and computational learning visual object classes）和MS COCO（Microsoft common objects in context））数据和虚拟数据分别高出17.9%和5.3%;在评估阶段相较于基准数据,常规任务（-30°且垂直移动）平均精度下降11.3%,环境任务（雾天）平均精度下降21.0%,困难任务（所有挑战）平均精度下降33.7%。结论 本文为智能驾驶设计和实施了在实际驾驶场景难以甚至无法进行的视觉计算实验,对复杂视觉挑战进行分析和评估,具备加强智能车在行驶过程中感知和理解周围场景的意义。     

The effects of body armour on the power development and agility of police officers

Abstract

A study was conducted in which 11 police officers wore one of three different types of Individual Light Armour Vests (ILAV), or normal station wear, for an entire day while completing power and agility-based tasks including a vertical jump (VJ), agility test, 20?m sprint and counter movement jump (CMJ). Despite all three ILAVs being significantly (p?<?.05) heavier than normal station wear, there were no significant differences between any of the ILAVs in VJ, time to complete the agility test, 20?m sprint time, peak force, velocity, power and jump distance in the CMJ. There was a significantly (p?<?.05) higher mean force produced in the CMJ while wearing all three ILAVs. The ILAV’s investigated do not appear to be heavy enough to significantly affect the power or agility of police officers. The utilisation of ILAVs by police officers does not appear to hinder policing tasks that involve agility or power development.

Practitioner summary: The addition of the extra load of military-styled body armour is known to decrease performance and mobility. When compared to normal station wear, the wearing of three different ILAV types used in policing do not appear to be heavy enough to affect the power or agility of police officers.

Abbreviations: ILAV: individual light armour vest; VJ: vertical jump; CMJ: counter movement jump; N: normal station wear     

Design and analysis of a wall-climbing robot for water wall inspection of thermal power plants

Boiler water wall in thermal power plants is characterized by high-altitude detection requirements. Moreover, the existing water wall-climbing robots are characterized by low obstacle-crossing performance, deviations, and a lack of autonomous crossing pipeline function. In view of this feature, a wall-climbing robot with permanent magnet and electromagnetic hybrid adsorption wheels is proposed. The robot has the function of independent traverse according to its own structural characteristics. Furthermore, transverse movement is proposed by comparing different adsorption modes, moving and driving modes. Robot statics in upward, downward, and transverse crawling are carried out, and nonsliding mechanical and nonoverturning mechanical models are obtained. Robot's dynamics are analyzed by considering the wall movement. The finite element simulation analysis of its main stressed parts is carried out by employing ANSYS, and an optimal structural model is obtained. The gap adsorption permanent magnet model is constructed, and its parametric simulation analysis is carried out using the Ansoft Maxwell module. The influence curve of the gap on the magnetic force is then obtained. Finally, the prototype is developed according to the design model and calculation analysis, and the experimental test is carried out. The experimental results show that the robot meets the expected functions and indexes, providing a basis for the intelligent development of thermal power plants.     

Can Autonomous Vehicles Be Safe and Trustworthy? Effects of Appearance and Autonomy of Unmanned Driving Systems

Although autonomous vehicles are increasingly becoming a reality, eliminating human intervention from driving may imply significant safety and trust-related concerns. To address this issue from a psychological perspective, this study applies layers of anthropomorphic cues to an artificial driving agent and explicates the process in which these cues promote positive evaluations and perceptions of an unmanned driving system. In a between-subjects factorial experiment (N = 89) consisting of three unmanned driving scenarios, participants interacted with an artificial driving agent with different levels of anthropomorphic cues induced by the variations in appearance (human-like vs. gadget-like) and autonomy (high vs. low) of the agent. The results indicated that human-like appearance and high autonomy were more effective in eliciting positive perceptions of the agent. In addition, a mediation analysis revealed that the greater level of anthropomorphism induced by human-like appearance and high autonomy in the agent evoked the feelings of social presence, which in turn positively affected the perceived intelligence and safety of and trust in the agent, suggesting that the extent to which users perceive the driving agent as intelligent, safe, and trustworthy is largely determined by the feelings of social presence experienced during their interaction.     

Digitalizing a Fuzzy Observer-Based Output-Feedback Control: Intelligent Digital Redesign Approach

This paper concerns an intelligent digital redesign (IDR) technique for a Takagi-Sugeno fuzzy observer-based output-feedback control (FOBOFC) system. The term IDR involves converting an existing analog control into an equivalent digital counterpart in the sense of state-matching. The IDR problem is herein viewed as a minimization problem of the norm distances between nonlinearly interpolated linear operators to be matched. Its constructive condition with global rather than local state-matching is formulated in terms of bilinear matrix inequalities. The main features of the proposed method are that the state estimation error in the plant dynamics is considered in the IDR condition that plays a crucial role in the performance improvement; the stability property is preserved by the proposed IDR algorithm; the separation principle is shown when the premise variables are measurable; finally, the IDR condition for a more general FOBOFC-an estimated premise variable case is conducted. A numerical example is demonstrated to visualize the feasibility of the developed methodology     

智能系统全生命周期安全测试理论与方法

人工智能技术在公共、国防安全领域得到了广泛应用,然而智能系统的安全性面临极大挑战。如何有效、全面、深入地对智能系统进行安全测试成为解决当前智能系统安全问题的重要途径。近年来,国内外高度重视智能系统的安全性问题,开展了大量的安全测试理论方法研究并出台了大量相关政策文件。针对智能系统面临的安全问题,本文详细阐述了面向智能系统全生命周期的安全测试理论与方法。首先,本文说明了智能系统的特点、安全内涵及安全机理;接着,本文结合模型训练、模型推理、模型部署三个主要生命周期阶段,详细地阐述了智能系统面临的安全性挑战和测试理论方法;最后,本文从标准、平台等角度阐明了构建智能系统安全测试支撑体系的方式,分析了自动驾驶典型场景下的智能系统安全测试案例,并给出了未来展望。开展智能系统安全测试理论和方法体系的建设,可有效规避系统潜在风险和质量缺陷,是实现人工智能算法可解释、可信赖的基本路径,对于确保人工智能技术安全、可靠、可控具有重要意义。     

Microprocessor—controlled driving unit for artificial hearts

A driving unit for artificial hearts has been produced. This unit uses sodium solution for the transmission of force between the power source and the artificial ventricle instead of the commonly used compressed air and thus avoids the danger of an air embolism. The power source of the hydraulic driving unit is an electromagnet; its armature drives a rolling membrane pump (safety chamber) which moves the transmission fluid (sodium solution) to the artificial ventricle (ellipsoid heart). The incompressible connection between the safety chamber and the ellipsoid heart allows direct control of the membrane motion in the ellipsoid heart and of the volume of pumped blood by measuring the armature stroke. The stroke is measured continuously with an optical position sensor; preprogrammable positions determine the systole end and diastole end respectively, so that the membrane need not strike the housing and is not exposed to additional stress. This volume-controlled mode is characterized by automatic selfregulation according to Starling's law.The microcomputer is programmed in a two-level technique. In the interrupt level the solenoid current is calculated according to the operating mode and armature position every 10 ms which is sufficient to simulate continuous operation. In the main-program level every operational parameter (maximum force, systole end, diastole end; systole duration and beat frequency for the frequency-fixing mode) is displayed on the CRT and can be changed via keyboard entry without any interruption in pumping.The durations of systole and diastole, respectively, are determined by the pumping force, the friction in the hydraulic transmission system and the preload and afterload respectively. The preload and afterload can be calculated from a time-force relation without invasive measurements.The driving unit has shown its haemodynamic efficiency in several in vivo experiments and has been running for 8 months in an in vitro durability test.     

智能汽车人机协同控制的研究现状与展望

随着人工智能、互联网技术、通信技术、计算机技术的快速发展,以电动化、智能化及网联化为基础的智能汽车成为汽车行业发展的一大趋势.按照汽车智能化、自动化的发展进程,美国汽车工程师协会将智能汽车的发展分为手动驾驶、驾驶辅助、部分自动化、有条件自动化、高度自动化和完全自动化6个级别,虽然不同层次、不同功能的汽车智能化技术正迅猛发展,但是真正意义上的全工况自动驾驶在短期内很难实现.因此,在未来很长一段时期内,智能汽车必然面对人机协同控制的局面,本文详细介绍了智能汽车人机协同控制中驾驶员建模及人机驾驶权动态优化控制的国内外研究现状,同时简要介绍了智能汽车测试与评价的国内外研究现状,提炼了共性问题,并对人机协同控制的发展趋势给出了一些观点.     

Motion sickness in passenger vehicles during test track operations

Abstract

As automation transforms drivers into passengers, the deployment of automated vehicles (AVs) has the potential to greatly increase the incidence of motion sickness. A study was conducted to quantify motion sickness response of front-seat passengers performing ecologically relevant passenger activities during conditions consistent with driving on public roadways. Fifty-two adults with a large range of self-reported levels of motion sickness susceptibility and age participated in data collection on a closed test track in a passenger sedan. Motion sickness ratings increased with task vs. no-task and moderate vs. low acceleration test conditions. Increased motion sickness susceptibility was associated with higher motion sickness ratings. In comparison to older participants (age?>?60), younger participants (age?<?60) experienced increased motion sickness. This is the first in-vehicle study that systematically compared normative passenger activities and acceleration magnitudes typical of normative driving conditions on motion sickness response for a large, diverse sample of passengers, enabling the exploration of the effects of covariates.

Practitioner summary: The data demonstrate that a relatively large range of motion sickness response can be expected to result from passengers performing visual tasks in passenger vehicles. Measurement and modelling efforts should seek to elucidate relationships among the factors contributing to motion sickness for the purpose of informing and prioritising future countermeasures for automated vehicles (AVs).

Abbreviations: AV(S): automated vehicles; BMI: body mass index; BVP: blood volume pulse; EDA: electrodermal activity; FMS: fast motion sickness scale; GPS: global positioning system; IMU: inertial measurement unit; ISO: International Organization for Standardization; MISC: misery scale; MSDV: motion sickness dose value; NDS: naturalistic driving study; SAE: Society of Automotive Engineers International; UMTRI: The University of Michigan Transportation Research Institute

Key Aspect of Research: Motion sickness may be an important barrier to widespread adoption of automated vehicles @UMTRI.     

Intelligent stock trading system with price trend prediction and reversal recognition using dual-module neural networks

This article presents an intelligent stock trading system that can generate timely stock trading suggestions according to the prediction of short-term trends of price movement using dual-module neural networks(dual net). Retrospective technical indicators extracted from raw price and volume time series data gathered from the market are used as independent variables for neural modeling. Both neural network modules of thedual net learn the correlation between the trends of price movement and the retrospective technical indicators by use of a modified back-propagation learning algorithm. Reinforcing the temporary correlation between the neural weights and the training patterns, dual modules of neural networks are respectively trained on a short-term and a long-term moving-window of training patterns. An adaptive reversal recognition mechanism that can self-tune thresholds for identification of the timing for buying or selling stocks has also been developed in our system. It is shown that the proposeddual net architecture generalizes better than one single-module neural network. According to the features of acceptable rate of returns and consistent quality of trading suggestions shown in the performance evaluation, an intelligent stock trading system with price trend prediction and reversal recognition can be realized using the proposed dual-module neural networks.     

面向智能驾驶测试的仿真场景构建技术综述

随着汽车智能化程度的不断提高,智能汽车通过环境传感器与周边行驶环境的信息交互与互联更为密切,需应对的行驶环境状况也越来越复杂,包括行驶道路、周边交通和气象条件等诸多因素,具有较强的不确定性、难以重复、不可预测和不可穷尽。限于研发周期和成本、工况复杂多样性,特别是安全因素的考虑,传统的开放道路测试试验或基于封闭试验场的测试难以满足智能驾驶系统可靠性与鲁棒性的测试要求。因此,借助数字虚拟技术的仿真测试成为智能驾驶测试验证一种新的手段,仿真场景的构建作为模拟仿真的重要组成部分,是实现智能驾驶测试中大样本、极限边界小概率样本测试验证的关键技术,这对提升智能驾驶系统的压力和加速测评水平显得尤为重要。面向智能驾驶测试的仿真场景构建技术已成为当前汽车智能化新的研究课题和世界性的研究热点,作为一种新兴技术仍面临许多挑战。本文提出了面向智能驾驶测试的仿真场景构建方法,系统阐述了国内外研究工作的进展与现状,包括场景自动构建方法和交通仿真建模方法,重点分析一些值得深入研究的问题并围绕场景构建技术的发展趋势进行了讨论分析,最后介绍了团队相关研究在2020中国智能驾驶挑战赛仿真赛和世界智能驾驶挑战赛的仿真场景应用情况。     

A microcomputer system for collecting and processing physiological information about the masticatory system.

A software system has been written for IBM PC, XT, AT and compatible computers to be used for data collection, analysis and display. The system supports the sampling and processing of data for jaw movement tracks, myoelectrical activities of masticatory muscles, the temporomandibular joint (TMJ) and occlusal sounds and bite force, etc. The package includes the following functions: calculating displacement, velocity, curvature, and curvature center of jaw movement trace, analyzing myoelectrical signals in amplitude integration, root mean square and power spectrum, processing TMJ sounds and occlusal sounds and bite force, analysing jaw movement traces and myoelectrical activities during mastication simultaneously, determining the maxillomandibular relations, etc. The program also provides versatile formatting capability for video, printing and plotting of data, and graph creation. The most of the above programs have flexibility and adaptability to other physiological signal processes.     

Fuzzy filter for nonlinear sampled-data systems: Intelligent digital redesign approach

This paper presents a fuzzy filter design method for nonlinear sampled-data systems using an intelligent digital redesign (IDR) technique. Based on a Takagi–Sugeno (T–S) fuzzy model, discretized closed-loop systems with pre-designed analog fuzzy and digital fuzzy filters are presented. An IDR problem is given to guarantee both state-matching condition and asymptotic stability. Sufficient conditions for solving the IDR problem are proposed and are derived in terms of linear matrix inequalities (LMIs). Finally, a simulation example is given to show the effectiveness of the proposed method.     

Robust digital implementation of fuzzy control for uncertain systems and its application to active magnetic bearing system

In this paper, we propose the robust digital control for active magnetic bearing (AMB) systems. For achieving the robust stability, we deal with the uncertainties of the given system based on the Takagi-Sugeno (T-S) fuzzy model. Also, in order to solve the digital implementation for real plants, this paper presents a robust intelligent digital redesign (IDR) method. The term IDR involves converting an analog controller into an equivalent digital one in the sense of state-matching. The uncertainties in the plant dynamics is shown in the IDR condition by virtue of the pade and inverse-pade approximation method. Also, the robust stability property is preserved by the proposed method. The sufficient conditions for robust controller are obtained in terms of solutions to linear matrix inequalities (LMIs). Finally, simulation results for two AMB systems are demonstrated to visualize the feasibility of the proposed method.     

Sampled-data Control of Fuzzy Systems Based on the Intelligent Digital Redesign Technique: An Input-delay Approach

International Journal of Control, Automation and Systems - In this paper, a novel intelligent digital redesign (IDR) method for a Takagi-Sugeno fuzzy system is proposed based on the guaranteed cost...     

The MOLDY short-range molecular dynamics package

We describe a parallelised version of the MOLDY molecular dynamics program. This Fortran code is aimed at systems which may be described by short-range potentials and specifically those which may be addressed with the embedded atom method. This includes a wide range of transition metals and alloys. MOLDY provides a range of options in terms of the molecular dynamics ensemble used and the boundary conditions which may be applied. A number of standard potentials are provided, and the modular structure of the code allows new potentials to be added easily. The code is parallelised using OpenMP and can therefore be run on shared memory systems, including modern multicore processors. Particular attention is paid to the updates required in the main force loop, where synchronisation is often required in OpenMP implementations of molecular dynamics. We examine the performance of the parallel code in detail and give some examples of applications to realistic problems, including the dynamic compression of copper and carbon migration in an iron–carbon alloy.

Program summary

Program title: MOLDYCatalogue identifier: AEJU_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEJU_v1_0.htmlProgram obtainable from: CPC Program Library, Queen?s University, Belfast, N. IrelandLicensing provisions: GNU General Public License version 2No. of lines in distributed program, including test data, etc.: 382 881No. of bytes in distributed program, including test data, etc.: 6 705 242Distribution format: tar.gzProgramming language: Fortran 95/OpenMPComputer: AnyOperating system: AnyHas the code been vectorised or parallelized?: Yes. OpenMP is required for parallel executionRAM: 100 MB or moreClassification: 7.7Nature of problem: Moldy addresses the problem of many atoms (of order 10⁶) interacting via a classical interatomic potential on a timescale of microseconds. It is designed for problems where statistics must be gathered over a number of equivalent runs, such as measuring thermodynamic properities, diffusion, radiation damage, fracture, twinning deformation, nucleation and growth of phase transitions, sputtering etc. In the vast majority of materials, the interactions are non-pairwise, and the code must be able to deal with many-body forces.Solution method: Molecular dynamics involves integrating Newton?s equations of motion. MOLDY uses verlet (for good energy conservation) or predictor–corrector (for accurate trajectories) algorithms. It is parallelised using open MP. It also includes a static minimisation routine to find the lowest energy structure. Boundary conditions for surfaces, clusters, grain boundaries, thermostat (Nose), barostat (Parrinello–Rahman), and externally applied strain are provided. The initial configuration can be either a repeated unit cell or have all atoms given explictly. Initial velocities are generated internally, but it is also possible to specify the velocity of a particular atom. A wide range of interatomic force models are implemented, including embedded atom, Morse or Lennard-Jones. Thus the program is especially well suited to calculations of metals.Restrictions: The code is designed for short-ranged potentials, and there is no Ewald sum. Thus for long range interactions where all particles interact with all others, the order-N scaling will fail. Different interatomic potential forms require recompilation of the code.Additional comments: There is a set of associated open-source analysis software for postprocessing and visualisation. This includes local crystal structure recognition and identification of topological defects.Running time: A set of test modules for running time are provided. The code scales as order N. The parallelisation shows near-linear scaling with number of processors in a shared memory environment. A typical run of a few tens of nanometers for a few nanoseconds will run on a timescale of days on a multiprocessor desktop.     

Errors and driver support systems

Recent technological developments seem to pave the way to sophisticated electronic co-driver systems that may help automobile drivers to cope with an ever increasing information load, to avoid certain errors, and to recover from others. GIDS—which stands for Generic Intelligent Driver Support—is a research project (under the EEC DRIVE Programme) to study the feasibility of an adaptive co-driver system. The conceptualization of a GIDS system requires close attention to performance errors as they may occur in certain subtasks of the driving task. One important issue that should be considered in some detail is that GIDS may eliminate errors as well as introduce them. Should various types of errors be represented formally and, if so, how they should be represented in order that GIDS can detect and cope with behavioural errors that drivers are likely to make under certain conditions? The requirements imposed by the project's goal to actually implement various driver support functions into a GIDS system is imposing tight constraints on error definition and identification. Some of the requirements will be discussed in terms of Soar. Soar is an intelligent computer architecture which is the embodiment of the theory of human problem solving formulated by Newell and Simon (1972). To the extent that the driving task is representable in Soar, the error theory that is required for any type of GIDS system to function must also be representable in Soar.     

移动机器人路径规划技术综述

智能移动机器人路径规划问题一直是机器人研究的核心内容之一.将移动机器人路径规划方法概括为:基于模版匹配路径规划技术、基于人工势场路径规划技术、基于地图构建路径规划技术和基于人工智能的路径规划技术.分别对这几种方法进行总结与评价,最后展望了移动机器人路径规划的未来研究方向.     

Observer-based sampled-data control for nonlinear systems: Robust intelligent digital redesign approach

This paper suggests a solution for a robust digital implementation of the output-feedback control system by utilizing intelligent digital redesign (IDR). The general IDR method involves designing a digital fuzzy controller which is comparable to a pre-designed analog one via state-matching. However, under containing the parametric uncertainties, the new problem about the structural property of the uncertainties is inevitably occurred in the procedure of discretization. For solving that, we use the bilinear and inverse-bilinear approximation method and the concerned IDR is viewed as a convex minimization problem of the norm distances between linear operators to be matched. Also, we consider the state estimation error in the plant dynamics so that we attempt to improve the control performance. The robust stability property is preserved by the proposed IDR method and its condition is represented in terms of linear matrix inequalities (LMIs). The simulation results for permanent magnet synchronous motor (PMSM) system are demonstrated to visualize the feasibility of the proposed method.