【作者】 唐黎明；

【导师】 龙湘云； 伍乾坤；

【作者基本信息】 湖南大学 ， 机械（专业学位）， 2023， 硕士

【摘要】 机械臂作为耐辐射机器人关键执行部件,可适应多种复杂场景中探查、检修、维护等作业需求,在诸多工业场景中被广泛使用。然而,在辐射环境中,由于γ、中子等强放射性射线或粒子作用,机械臂中电路系统极易失效,极大程度上制约了其在涉核场景中的工程应用。目前,中国原子能科学研究院、西南科技大学、新疆理化所等单位针对耐辐射机械臂的开发与应用开展了系统性研究工作,有力的推动了机械臂在核辐射场景中的应用与拓展推广。但是,上述研究工作多是将机械臂关节驱动电路后置并结合结构屏蔽的方式实现其电路系统的抗辐射加固,导致机械臂精度不足、负载受限、体积庞大等。为解决上述问题,亟待开发适用于强核辐射场景的机械臂耐辐射驱动电路。然而,由于国外技术封锁、耐辐射芯片禁运,且国内航天级芯片价格高昂等,直接采用耐辐射器件替换的技术路线难以实现。通过辐射筛选出适用于强核辐射场景的机械臂耐辐射驱动电路的电子元器件,并基于此开展电路集成设计是目前相对成熟的技术方案。但由于机械臂耐辐射驱动电路构成复杂,涉及控制、通讯、传感、功率裂变等多个部分,电路系统中同时存在几十种核心电子元器件,工艺、批次等外界因素导致其在核辐射场景下的性能衰变存在较大差异;同时,由于辐射测试筛选成本高昂、样本匮乏,使得驱动电路系统的整体耐辐射性能变得极为困难。因此,亟需开发小样本条件下的机械臂驱动电路耐γ辐射电离总剂量评估方法并展开实际应用。为此,本文将针对机械臂驱动电路开展电子元器件单元和整体驱动电路系统的耐γ辐射累积电离总剂量评估方法研究,实现了小样本条件下机械臂驱动电路的耐γ辐射电离总剂量评估,可为后续机械臂驱动电路加固设计提供技术支撑和理论指导。本文的主要研究内容如下:(1)发展了一种基于概率盒模型的裕量和不确定性量化方法(P-Box-QMU)。该方法利用概率盒模型融合随机不确定性与认知不确定性,适用于小样本数据。并将概率盒模型(P-Box)与裕量和不确定性量化(QMU)方法框架结合,量化评估对象的裕量和不确定性,计算得到置信因子作为评估标准。(2)提出了基于概率盒模型的QMU方法的驱动电路元器件耐γ辐射电离总剂量评估方法,实现了驱动电路中主要敏感元器件的耐γ辐射电离总剂量评估。首先,分析驱动电路得到其中的敏感元器件,并设计偏置板进行辐射实验获得失效评估参数数据;其次,采用多项式拟合分别得到元器件失效评估参数均值、均方差与累积剂量的函数关系;最后,利用累积剂量对应的均值与均方差构建P-Box,使用P-Box-QMU方法计算得到置信因子,由此推算得到失效累积剂量评估值,并与实验结果对比,验证了该评估方法的有效性。(3)引入贝叶斯网络模型发展了一种驱动电路系统耐γ辐射电离总剂量评估方法,实现了驱动电路系统的耐γ辐射电离总剂量评估。首先,通过分析驱动电路故障来源构建故障树,并将其转化为贝叶斯网络(BN);其次,对BN节点层次间的概率重要度进行了分析,结合主要敏感元器件的单元评估结果,分析计算得到了驱动电路系统的失效累积剂量评估值;最后,针对驱动电路设计偏置板进行钴-60源γ辐射实验,将实验结果与评估结果对比,证明此系统评估方法合理且正确。更多还原

【Abstract】 As a key executive component of radiation-resistant robot,robot arm can adapt to exploration,overhaul,maintenance and other operational requirements in a variety of complex scenes,and is widely used in many industrial scenes.However,in the radiation environment,due to the action of gamma,neutron and other strong radioactive rays or particles,the circuit system in the manipulator is easy to fail,which greatly restricts its engineering application in nuclear scenarios.At present,China Institute of Atomic Energy,Southwest University of Science and Technology,Xinjiang Institute of Physics and Chemistry and other units have carried out systematic research on the development and application of radiation-resistant robotic arms,which has strongly promoted the application and expansion of the robotic arms in nuclear radiation scenarios.However,most of the above research work is to realize the radiation-resistant reinforcement of the circuit system by repositioning the drive circuit of the joint of the manipulator and combining with structural shielding,which leads to the insufficient precision,load limitation and large volume of the manipulator.In order to solve the above problems,it is urgent to develop the radiation-resistant drive circuit for the manipulator which is suitable for the strong nuclear radiation scene.However,due to the blockade of foreign technology,the embargo of radiation-resistant chips,and the high price of domestic space-grade chips,it is difficult to directly replace the technical route of radiation-resistant devices.At present,it is a relatively mature technical scheme to screen out the electronic components of radiation-resistant drive circuit of mechanical arm suitable for strong nuclear radiation scene through radiation,and carry out circuit integrated design based on this.However,due to the complex composition of the radiation-resistant drive circuit of the manipulator,involving control,communication,sensing,power fission and other parts,there are dozens of core electronic components in the circuit system,and external factors such as process and batch lead to great differences in its performance decay under nuclear radiation scenarios.At the same time,due to the high cost of radiation testing and screening and the lack of samples,it is very difficult to drive the overall radiation resistance performance of the circuit system.Therefore,it is urgent to develop a method to evaluate the total dose of γ-radiation ionization resistance of the manipulator drive circuit under the condition of small samples and carry out practical application.Therefore,this paper will carry out the research on the evaluation method of the cumulative ionization dose of γ-radiation resistance of the electronic component unit and the whole drive circuit system for the mechanical arm drive circuit,and achieve the evaluation of the total dose of γ-radiation ionization resistance of the mechanical arm drive circuit under the condition of small samples,which can provide technical support and theoretical guidance for the subsequent reinforcement design of the mechanical arm drive circuit.The main research content of this paper is as follows:(1)A quantitative method of margin and uncertainty based on probability box model(PBox-QMU)is developed.This method uses probability box model to fuse stochastic uncertainty and cognitive uncertainty,and is suitable for small sample data.The probability box model(PBox)was combined with the margin and uncertainty quantification(QMU)method framework to quantify the margin and uncertainty of the evaluation object,and the confidence factor was calculated as the evaluation standard.(2)A probability box model based QMU method is proposed to evaluate the total dose ofγ-radiation ionization resistance of the driving circuit components,realizing the evaluation of the total dose of γ-radiation ionization resistance of the main sensitive components in the driving circuit.Firstly,the sensitive components were obtained by analyzing the driving circuit,and the bias plate was designed for radiation experiment to obtain the failure evaluation parameter data.Secondly,polynomial fitting was used to obtain the functional relationship between mean value,mean square error and cumulative dose of component failure evaluation parameters.Finally,the mean and mean square error corresponding to the cumulative dose were used to construct a P-Box,and the confidence factor was calculated using the P-Box-qmu method.Then the evaluation value of the failure cumulative dose was calculated,and compared with the experimental results,the effectiveness of the evaluation method was verified.(3)A method to evaluate the total radiation ionizing dose of drive circuit system was developed by using Bayesian network model.Firstly,the fault tree is constructed by analyzing the fault source of the driver circuit and converted into Bayesian network(BN).Secondly,the probabilistic importance between BN nodes is analyzed.Combined with the unit evaluation results of the main sensitive components,the failure cumulative dose evaluation value of the drive circuit system is obtained.Finally,the experiment γ radiation by Co-60 was carried out for the drive circuit design bias plate,and the experimental results were compared with the evaluation results,which proved that the system evaluation method was reasonable and correct.更多还原