液压驱动下肢外骨骼机器人关键技术研究

发布时间：2018-03-09 11:50

本文选题：外骨骼　切入点：机器人　出处：《浙江大学》2017年博士论文　论文类型：学位论文

【摘要】：助力型下肢外骨骼机器人是一类可以被人穿戴,提高穿戴者行走耐久性、行走速度、负重能力等体能的智能装备。助力型下肢外骨骼机器人可实现行走、跑、跳等多样化的运动形式,可在战场、抢险救灾、野外施工作业等缺少电力、燃料等能源供应和道路交通的野外工作环境发挥重要作用。因此助力型下肢外骨骼机器人相关研究是国际学术前沿,其难点在于实现样机的高承载能力、高功率密度、长工作时间等特性。本学位论文针对液压驱动助力型下肢外骨骼机器人的关键技术开展研究,选题具有很强的工程应用背景和重要的学术研究价值。本课题开展中做出了以下创新性研究:1.研制了高功率密度长运行时间的外骨骼机器人动力系统。当前外骨骼机器人的动力系统多采用锂电池供应能源,然而锂电池能量密度有限,难以保证机器人大功率、长时间工作。为解决这一难题,本学位论文研制了外骨骼机器人发动机动力系统样机,该样机重约20kg,最大可向机器人提供1.5kW液压能源与160W直流电力,在携带5kg燃料时该样机可以维持1.2kW总功率持续运行4.5小时,满足上述功率及容量的条件下该方案整机(含燃料)重量仅占蓄电池-电机方案中锂电池重量(39kg)的64%。2.提出了一种图形化多目标优化方法。在动力系统研发中提出了弹簧增压油箱实现机器人液压系统闭式回路,避免了空气吸入、大气中颗粒污染物混入及液压泵气蚀。弹簧增压油箱的设计存在重量、空间体积、增压压力等6项优化指标。现有的多目标优化方法均依靠偏好模型来权衡各项性能指标;而弹簧增压油箱的优化问题由于优化目标多、优化目标之间无法比较等因素难以建立偏好模型。本文提出了一种多目标优化方法,采用图形化界面展示优化结果,具有流程清晰、数据展示直观的优点,弥补了以数据为输出的多数现有方法的不足,可以协助设计者方便快捷的获得最终解,尤其具有求解过程不需要设计者提供偏好模型的优点,因此与现有方法相比在实际问题中可以得到更广泛的应用。3.开发了易于功能扩展和改进的控制器,提出了一种用于控制器的多任务动态调度框架。本论文提出的主板-功能模块结构的机器人控制系统具有工作电压低、功耗小的特点,且易于功能扩展和改进。针对控制系统工作中需要对大量任务高速调度执行的特点提出了一种多任务动态调度框架,利用任务间固有的依赖关系维护活动任务表,可以使控制器在不采用实时操作系统的情况下完成多任务的并发执行,避免了实时操作系统任务同步与调度带来的处理器时间损耗与存储空间损耗。本文具体研究内容如下:第一章,在助力型下肢外骨骼机器人相关文献的基础上对其发展现状进行了总结;同时分析了外骨骼机器人应用环境的特征及需求,阐明了本课题的主要研究内容,确定了本文研究的机器人采用发动机驱动液压作动器的总体方案;最后介绍了本课题研究难点。第二章,外骨骼机器人下肢机构设计研究。首先介绍了步态采集装置及获取的步态数据;进而论述下肢机构的需求以及综合考虑了结构强度、承载能力、对穿戴者的适当保护措施等因素的设计实现。主动关节采用液压伺服阀驱动非对称缸,以获取足够的下肢关节驱动力。在下肢机构研发的基础上分析了其运动学模型与静力学模型。第三章,外骨骼机器人动力系统设计研究。本章通过对比基于发动机-燃料方案与基于电池方案展示了前者在功率质量比上的优势。介绍了动力系统原动机模块、液压回路及发电模块等子系统的研发实现及工作特性。提出了基于图形界面的多目标优化方法,并采用此方法完成了增压油箱的多目标优化设计,从而实现动力系统闭式液压回路。建立了动力系统发动机功率控制模型以及液压回路负载流量观测模型。第四章,控制系统硬件及程序设计研究。硬件上采用主板/功能模块结构方案,研制了数据采集、作动器驱动的多个模块,对数据采集模块采样精度做出详细分析,结果表明采样精度满足外骨骼机器人系统需求。在控制系统程序方面分析了其构成与运行方式,提出了多任务动态调度策略实现多任务高效并发执行。第五章,外骨骼机器人物理样机试验研究。试验研究从动力系统特性、下肢空载运动特性、整机承载特性等多个方面验证动力系统的输出能力及运行特性、下肢机械关节的运动性能,并对前文提出的模型进行验证。第六章,对论文研究工作进行总结,突出了研究成果及创新点,并对将来工作的方向和内容做出展望。
[Abstract]:Assist the lower extremity exoskeleton robot is a kind of can be worn, improve wearer walking durability, walking speed, intelligent equipment and other physical loading capability. The power type lower extremity exoskeleton robot can walk, run, jump, exercise form diversification, can be on the battlefield, disaster relief, field work and lack of power fuel, energy supply and road traffic field work environment play an important role. Therefore research power type lower extremity exoskeleton robot is an international academic frontier, it is difficult to achieve high bearing capacity, high power density, long working time and other characteristics. Key technologies are researched in this thesis, the hydraulic driving type lower limbs the bones of the robot, the value of academic research topic has great application background and important. This research made the following research: 1. developed high The exoskeleton robot dynamic system power density long running time. The dynamic system of the exoskeleton robot using lithium battery energy supply, however, the energy density of the lithium battery is limited, it is difficult to ensure the robot with high power and long time work. In order to solve this problem, this paper developed the degree of exoskeleton robot prototype engine power system, the the prototype weighs about 20kg, can provide 1.5kW hydraulic energy and 160W DC power to the robot, carrying 5kg fuel of the prototype can maintain the total power of 1.2kW continuous operation for 4.5 hours, the scheme of the power and meet the capacity conditions (including fuel) weight accounted for lithium battery weight batteries - motor program (39kg) the 64%.2. presents a graphical multi-objective optimization method in power system research. We put forward the spring pressurizing oil tank to achieve the robot close loop hydraulic system, avoid empty The gas inhaled particulate pollutants into the atmosphere. And the hydraulic pump cavitation has weight, design pressure of the spring tank volume, 6 optimization indexes of pressure. Optimization depends on preference model to weigh the performance indexes of the existing multi-objective method; and the pressure of the spring tank because of the optimization goal optimization problem, it is difficult to establish the model of preference the optimization goal cannot be compared between other factors. This paper proposes a multi-objective optimization method, using the graphical interface to display the results of optimization, has advantages of clear flow, data display, to make up for the lack of data in most existing methods of output, can help designers to easily and quickly get the final solution, especially with the solving process does not require designers to provide advantages of preference model,.3. can be applied more widely developed easily in practical problems so compared with the existing methods The extension and improvement of the proposed controller, a controller for a multi task dynamic scheduling framework. The control system function module structure of the robot is proposed in this paper, the motherboard with low working voltage, small power consumption, and easy extension and improvement. In order to propose a multi task dynamic scheduling framework for a large number of features high speed task scheduling execution control system work, the use of inter task dependencies inherent in the maintenance task list, concurrency controller can be multi task in a real-time operating system under the condition of execution, to avoid the loss and loss of storage space, processor time synchronization and scheduling bring real-time operating system. In this paper, the specific research task the contents are as follows: the first chapter, based on the power type lower extremity exoskeleton robot on the related literature to summarize the status of the development and analysis; The characteristics and demands of the exoskeleton robot application environment, expounds the main research contents of this topic, determines the research robot using engine driven hydraulic actuator for the overall scheme; finally introduced the research difficulties. In the second chapter, study on Mechanism Design of limb exoskeleton robot. First introduced the gait acquisition device and the gait data acquisition; and then discusses the legal mechanism of demand and considering the structural strength, bearing capacity, design factors on the wearer's proper protective measures. The active joint adopts hydraulic servo valve driven asymmetric cylinder, to obtain sufficient driving force. Based on lower limb joint leg mechanism on the development of the analysis the kinematics model and static model. In the third chapter, the exoskeleton robot dynamic system design research. This chapter through the comparison based on the engine with fuel scheme based on battery In the former scheme shows the advantage of power quality than on power system. This paper introduces the prime mover module, power module and so on research and development of hydraulic circuit and the realization of the system and working characteristics is proposed. The multi-objective optimization method based on graphical interface, and using this method to complete the multi-objective optimization design of pressure tank, so as to realize the close power system hydraulic power system is established. The power of the engine control model and hydraulic circuit load flow observation model. In the fourth chapter, the design of hardware and program control system. The hardware adopts the motherboard / function module structure, developed a data acquisition module, a plurality of actuator drive, make a detailed analysis of the data acquisition module, sampling precision and the results show that the accuracy meets the needs of the exoskeleton robot system. The control system program to analyze the composition and operation mode, put forward multi tasks Service dynamic scheduling strategy to achieve multi task efficient concurrent execution. In the fifth chapter, experimental study on exoskeleton robot prototype. Experimental study on dynamic characteristics from the system, the motion characteristics of the bearing load of lower extremity, output capacity and operating characteristics of many aspects of performance verification of the dynamic system, the motion performance of lower extremity mechanical joint, and the proposed model verify. The sixth chapter, the thesis summarizes the research work, the research results and innovation, and the contents and directions for future work is prospected.

【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】