In order to calculate the phase mission reliability, the length of each phase (i.e. the time duration for completing each phase) is required and the values assumed in this work are shown in Table 1. T
In order to calculate the phase mission reliability, the length of each phase (i.e. the time duration for completing each phase) is required and the values assumed in this work are shown in Table 1. These values taken are for demonstration purposes, where the total time duration to complete the whole mission is 0.51 h. It is worth mentioning that the assumed values in Table 1 are based on the consultation with an AGV operator. These values would be different for different applications. Therefore, the time duration should be modified correspondingly when considering different AGV applications.
4 AGV reliability model generation
4.1 Subsystem level reliability models
Initially, a detailed FMECA analysis of the AGV (with an extract illustrated in Table 2) was performed in order to obtain a detailed understanding of the vehicle. Eight subsystems were identified for analysis. In Table 2, the subsystem, the laser navigation system (LNS), is shown where columns labelled S, F, D and RPN refer to the severity ranking, frequency ranking, detectability ranking (all with ratings 1–5) and risk priority number, respectively (calculated as the multiplicity of all three rankings). Similar tables were developed for all AGV subsystems.
The understanding gained from the FMECA was then used to construct fault trees describing the failure of each subsystem
As an example the failure of the laser navigation system (LNS) is shown in Fig. 3. This subsystem failure can be broken down to be caused by four basic events: laser
emitter failure; laser sensor failure; GPS failure and signal transmitter failure. In total, eight subsystem level fault trees have been constructed, varying in size from just 1 gate and 5 events to 3 gates and 11 events. Following this understanding of the interrelationships between failures within the AGV system, this information can be used to establish the likelihood/frequency of AGV subsystem failure. A quantitative analysis of the fault trees has yielded the frequency of failure of each subsystem as shown in Table 3. The data used for the basic events in the subsystem fault trees was
based on RAC FMD-97 and expert knowledge
摘要:自动导引车(AGV)由于其高效率和低成本的特点而被广泛用于仓库和自动生产线中物料的智能运输和分配。这些车辆沿着预定的路线行进,以在没有操作员的监督的情况下提供期望的任务。这方面的努力主要集中在这些AGV的路线优化和交通管理上。然而,这些车辆的健康管理及其最佳任务配置却很少受到关注。 为了保证它们的附加价值,以AGV传输系统为例,本文研究了评估自动导引车可靠性问题的能力。在故障模式效应和临界性分析(FMECA)之后,通过故障树分析(FTA)分析AGV系统的可靠性,并且使用Petri网(PN)方法评估车辆任务可靠性。通过执行分析,可以分析任务失败的可接受性,因此可以检查AGV系统的服务能力和潜在利润,并在性能不可接受的情况下更改任务。PN方法可以很容易地扩展到有能力处理车队AGV任务的可靠性评估。
关键词: 自动引导车辆 可靠性 Petri网故障树分析
1 介绍
为了在仓库和生产设施中智能运输和分配材料,近年来越来越多地使用自动导引车(AGV)。这些车辆沿着预定的路线行驶,以在没有船上操作员的监督的情况下提供各种任务。随着AGV系统越来越大和越来越复杂,通过研究AGV的设计和控制方面,提高AGV系统的效率和降低运营成本自然成为首要任务,通过识别新的流程布局,包括工作站布局和开发先进的交通管理策略,包括车辆路线和任务分配。 Trenkle介绍了分散式受控AGV系统的安全要求和安全功能。 确定了三个主要危险,即与一个人发生碰撞,倾斜和坠落。 分析了AGV的速度,制动距离和检测区域要求以及平均时间对危险故障和性能的影响。 关于故障响应,Ebben为AGV的特殊案例研究开发了一种失效控制管理方法,这是一种地下交通系统(考虑装载和卸载的AGV)。 在AGV的可靠性建模领域,Fazlollahtabar 创建了一个模型,以最大限度地提高AGV的可靠性,并将维修成本降到最低,Tavana和Fazlollahtabar将AGV的可靠性建模为成本函数,以优化时间和成本目标。 但是,充分了解AGV如何失效以及导致此类故障的原因仍然需要。Duran等人在这方面取得了一些进展。试图通过使用故障树分析(FTA)和贝叶斯信念网络(BN)的组合方法在AGV上识别光检测和测距(LIDAR)系统和基于相机的计算机视觉系统(CV)的基本故障模式)。 在工作中,人为伤害,财产损失和车辆损坏被定义为故障树中的最高事件。 但是,这项研究并未涵盖自动导引车中包含的所有组件和组件。 使用Petri网(PN)进行建模已成为评估系统或任务可靠性的常用工具和热门研究课题。 例如,Wu提出了一个扩展的面向对象Petri网模型,用于分析2015年导致共因故障的分阶段任务的可靠性。另一方面,从工业应用的角度来看,乐和安德鲁斯提出了一个风力涡轮机资产模型,以研究基于PN的不同机翼隧道部件的退化,维护和检查过程。 然而,迄今为止,Petri网(PN)方法仅被用作调查AGV系统的路线规划和控制策略的数学工具。 例如,罗和倪设计了一个使用Petri网的可编程逻辑控制器(PLC),以防止AGV系统中车辆的碰撞。Nishi和Maeno提出了一种选择的方法,模拟半导体制造车间AGV的路由规划。