WO2021213542A1 - Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network - Google Patents

Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network Download PDF

Info

Publication number
WO2021213542A1
WO2021213542A1 PCT/CN2021/095467 CN2021095467W WO2021213542A1 WO 2021213542 A1 WO2021213542 A1 WO 2021213542A1 CN 2021095467 W CN2021095467 W CN 2021095467W WO 2021213542 A1 WO2021213542 A1 WO 2021213542A1
Authority
WO
WIPO (PCT)
Prior art keywords
logistics
accident
bayesian network
drones
failure
Prior art date
Application number
PCT/CN2021/095467
Other languages
French (fr)
Chinese (zh)
Inventor
韩鹏
Original Assignee
中国民航大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国民航大学 filed Critical 中国民航大学
Publication of WO2021213542A1 publication Critical patent/WO2021213542A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management

Definitions

  • the present disclosure belongs to the technical field of logistics drones, and particularly relates to a method for assessing the failure risk of logistics drones based on a Bayesian network.
  • the Bayesian network-based risk assessment method for logistics drones includes the following steps in order:
  • Step 1 Analyze and obtain the causes of accidental fall of logistics drones and the probability values of each cause from the perspective of system structure;
  • Step 2 From the perspective of operating environment, analyze and obtain the causes of accidental fall of logistics drones and the probability value of each cause;
  • Step 3 From the perspective of human factors, analyze and obtain the inducement of the accidental fall of the logistics drone and the probability value of each inducement;
  • Step 4 Build a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in step 1, step 2 and step 3;
  • Step 5 Construct the conditional probability table of the Bayesian network for the failure assessment of logistics drones obtained in Step 4;
  • Step 6 According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, calculate the accident of logistics drone when different accident causes occur The risk of a fall accident.
  • step 4 the method of constructing a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in steps 1, 2 and 3 is:
  • the four failure modes of control failure, stall, loss of all power and loss of partial power are respectively regarded as the accident nodes, according to the causal relationship between them.
  • Each accident node is connected by a directed edge in the form of a single arrow, where the accident node at the beginning of the arrow is the parent node, and the accident node at the end of the arrow is the child node, thus connecting the failure mode and the cause through a directed acyclic graph,
  • the directed acyclic graph based on the network framework describes the dependence and independent relationship between variables, and constructs a Bayesian network for failure assessment of logistics drones.
  • step 5 the conditional probability in the conditional probability table of the Bayesian network of the logistics drone failure assessment obtained in the construction step 4 refers to the probability of an accident under the condition that another accident has occurred, which can be passed through the logistics
  • the statistical analysis of the UAV operating data and the calculation formula of the joint distribution probability of each accident node are obtained, as shown in formula (1):
  • P is the conditional probability of accidents node
  • X i is the node number of accidents
  • n is the number of all nodes parent node of X i accident
  • step 6 according to the probability value obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of the logistics drone failure assessment obtained in step 5, the logistics when different causes of accidents occur are calculated
  • the method of risk of accidental fall of drones is:
  • step 1, step 2 and step 3 According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, using Bayesian network principle, calculate step 1, step 2 and step The risk of accidental fall of the logistics drone when one of the accident causes or certain causes of accidents occurs in 3.
  • the module greatly reduces the difficulty of knowledge acquisition and the complexity of probabilistic reasoning. It is suitable for uncertain reasoning problems in the risk assessment of logistics UAV systems, and provides an effective method for quantitative assessment of logistics UAV safety risks.
  • Figure 1 is a schematic diagram of the structure of the logistics UAV system.
  • Figure 2 is a schematic diagram of a Bayesian network for failure assessment of logistics drones constructed in the present disclosure.
  • Fig. 3 is a diagram of the probability distribution of occurrence of various events in the communication link failure condition in the present disclosure.
  • Figure 4 shows the probability distribution of each event in the motor fault condition.
  • the Bayesian network-based risk assessment method for logistics drones includes the following steps 1-6 in order:
  • Step 1 Analyze and obtain the causes of accidental fall of logistics drones and the probability values of each cause from the perspective of system structure;
  • the logistics UAV system includes a flight platform system, a ground control system and a mission-related load system composed of a power system, a flight control system, a power system, a communication system, and an airframe structure.
  • the failure modes of each link are analyzed in turn, and the logistics drone accidental fall accident caused by the failure of the system structure is statistically analyzed based on the operating data of the logistics drone.
  • the incentives and the probability values of each incentive are shown in Table 1.
  • Step 2 From the perspective of operating environment, analyze and obtain the causes of accidental fall of logistics drones and the probability value of each cause;
  • the failure of logistics drones caused by operating environment is mainly related to weather conditions during operation, including strong gusts, heavy precipitation and ice accumulation.
  • weather conditions exceed the maximum index for the normal operation of logistics drones, the operation of logistics drones will be greatly affected.
  • Table 2 statistical analysis of the causes of accidental fall of logistics drones caused by the operating environment and the probability of each cause are shown in Table 2.
  • Step 3 From the perspective of human factors, analyze and obtain the inducement of the accidental fall of the logistics drone and the probability value of each inducement;
  • the human factors for the failure of logistics drones are mainly caused by logistics drone operators and ground maintenance personnel. According to the operating data of logistics drones, statistical analysis of the causes of accidental fall of logistics drones caused by the personnel involved in the operation and their job responsibilities and the probability values of each cause are shown in Table 3.
  • Step 4 Build a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in step 1, step 2 and step 3;
  • the four failure modes of control failure, stall, loss of all power and loss of partial power are respectively regarded as the accident nodes, according to the causal relationship between them.
  • Each accident node is connected by a directed edge in the form of a single arrow, where the accident node at the beginning of the arrow is the parent node, and the accident node at the end of the arrow is the child node, thus connecting the failure mode and the cause through a directed acyclic graph,
  • the directed acyclic graph based on the network framework describes the dependence and independent relationship between variables, and constructs a Bayesian network for failure assessment of logistics drones, as shown in Figure 2.
  • Step 5 Construct the conditional probability table of the Bayesian network of logistics drone failure evaluation obtained in Step 4;
  • the Bayesian network for failure assessment of logistics drones constructed in step 4 contains a set of accident nodes, and the causal relationship between the connected accident nodes is represented by a conditional probability table.
  • Conditional probability refers to the probability of occurrence of an accident under the condition that another accident has occurred, and it is used to describe the dependency between accident nodes. Because the relationship between the logistics drone accident nodes is more complicated, and the relationship between the accident nodes is not independent of each other, the conditional probability of the accident node can be calculated through the statistical analysis of the logistics drone operation data and the joint distribution probability calculation formula of each accident node Obtained, as shown in formula (1):
  • P is the conditional probability of accidents node
  • X i is the node number of accidents
  • n is the number of all nodes parent node of X i accident
  • Step 6 According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, calculate the accident of logistics drone when different accident causes occur The risk of a fall accident;
  • step 1, step 2 and step 3 According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, using Bayesian network principle, calculate step 1, step 2 and step The risk of accidental fall of the logistics drone when one of the accident causes or certain causes of accidents occurs in 3.
  • the probability of accidental fall accidents and intermediate events under the two working conditions are calculated respectively.
  • the probability of accidental fall of logistics drones is 21.69%.
  • the probability of occurrence is 99.98% and 92.18%, while the probability of control failure and manual intervention failure is less than 1%;
  • the probability of accidental fall of logistics drones is 22.14%, of which Among the intermediate accidents that cause the bottom accident, the manual intervention failure is the most likely to occur, with a probability of 97.99%.
  • the probability of unmanned aerial vehicle stalls and control failures is also greater. The calculation results of these two working conditions clearly show the probability of the occurrence of various intermediate events in the development process of different logistics drone accident inducements to accidental fall accidents.

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Quality & Reliability (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Game Theory and Decision Science (AREA)
  • Educational Administration (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

Disclosed is a logistics unmanned aerial vehicle risk assessment method based on a Bayesian network. The method comprises the steps of: from the perspective of a system structure, an operating environment and a human factor, analyzing and obtaining causal factors of an accidental crash accident of a logistics unmanned aerial vehicle and probability values of the causal factors; constructing a logistics unmanned aerial vehicle failure assessment Bayesian network; and constructing a conditional probability table for the logistics unmanned aerial vehicle failure evaluation Bayesian network: calculating the risk of an accidental crash accident occurring in the logistics unmanned aerial vehicle when different accident causal causes occur, etc.

Description

基于贝叶斯网络的物流无人机失效风险评估方法A method for evaluating the failure risk of logistics drones based on Bayesian networks
相关申请的引用References to related applications
本公开要求于2020年11月9日向中华人民共和国国家知识产权局提交的申请号为202011239492.1、名称为“一种基于贝叶斯网络的物流无人机失效风险评估方法”的发明专利申请的全部权益，并通过引用的方式将其全部内容并入本文。This disclosure requires all of the invention patent applications filed with the State Intellectual Property Office of the People’s Republic of China on November 9, 2020, with the application number 202011239492.1, titled "A method for assessing the failure risk of logistics drones based on the Bayesian network" Rights and interests, and incorporate its entire content into this article by reference.
技术领域Technical field
本公开属于物流无人机技术领域，特别是涉及基于贝叶斯网络的物流无人机失效风险评估方法。The present disclosure belongs to the technical field of logistics drones, and particularly relates to a method for assessing the failure risk of logistics drones based on a Bayesian network.
背景技术Background technique
目前评估无人机失效风险的技术较少，尤其是针对物流无人机展开失效风险评估的技术。由于无人机的失效风险与其结构特点和运行环境密切相关，而城市物流无人机结构复杂，各系统故障都可能导致无人机整体失效坠毁。同时，在城市空中物流运输中，强风、暴雨、强电磁干扰都会诱发无人机系统意外故障，造成坠机伤人。现有的无人机失效风险评估方法并未较好地针对物流无人机的结构特点、运行环境和保障人员工作特性进行建模，同时由于缺乏对物流无人机失效诱因的全面提取，无法建立风险定量化评估模型。At present, there are few technologies for evaluating the failure risk of drones, especially for logistics drones. As the failure risk of UAVs is closely related to its structural characteristics and operating environment, and the structure of urban logistics UAVs is complex, the failure of each system may cause the overall failure of the UAV to crash. At the same time, in urban air logistics and transportation, strong winds, heavy rains, and strong electromagnetic interference can all induce unexpected failures of the UAV system, causing crashes and injuries. Existing drone failure risk assessment methods do not properly model the structural characteristics, operating environment, and support personnel's work characteristics of logistics drones. At the same time, due to the lack of comprehensive extraction of the failure incentives of logistics drones, they cannot Establish a quantitative risk assessment model.
公开内容Public content
本公开提供的基于贝叶斯网络的物流无人机风险评估方法包括按顺序进行的下列步骤：The Bayesian network-based risk assessment method for logistics drones provided in the present disclosure includes the following steps in order:
步骤1：从系统结构角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 1: Analyze and obtain the causes of accidental fall of logistics drones and the probability values of each cause from the perspective of system structure;
步骤2：从运行环境角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 2: From the perspective of operating environment, analyze and obtain the causes of accidental fall of logistics drones and the probability value of each cause;
步骤3：从人为因素角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 3: From the perspective of human factors, analyze and obtain the inducement of the accidental fall of the logistics drone and the probability value of each inducement;
步骤4：基于步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因构建物流无人机失效评估贝叶斯网络；Step 4: Build a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in step 1, step 2 and step 3;
步骤5：构建步骤4获得的物流无人机失效评估贝叶斯网络的条件概率表；以及Step 5: Construct the conditional probability table of the Bayesian network for the failure assessment of logistics drones obtained in Step 4; and
步骤6：依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，计算出不同事故诱因发生时物流无人机发生意外坠落事故的风险。Step 6: According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, calculate the accident of logistics drone when different accident causes occur The risk of a fall accident.
在步骤4中，所述的基于步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因构建物流无人机失效评估贝叶斯网络的方法是：In step 4, the method of constructing a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in steps 1, 2 and 3 is:
根据步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因，分别将控制失效、失速、失去全部动力和失去部分动力四项失效模式作为事故节点，按照各自之间的因果关系将各个事故节点通过有向边以单箭头形式连接，其中箭头起始端的事故节点是父节点，箭头末端的事故节点是子节点，由此将失效模式与诱因通过有向无环图连接起来，基于网络框架的有向无环图描述变量间的依赖和独立关系，构建成物流无人机失效评估贝叶斯网络。According to the inducement of the accidental fall of the logistics drone obtained in step 1, step 2 and step 3, the four failure modes of control failure, stall, loss of all power and loss of partial power are respectively regarded as the accident nodes, according to the causal relationship between them. Each accident node is connected by a directed edge in the form of a single arrow, where the accident node at the beginning of the arrow is the parent node, and the accident node at the end of the arrow is the child node, thus connecting the failure mode and the cause through a directed acyclic graph, The directed acyclic graph based on the network framework describes the dependence and independent relationship between variables, and constructs a Bayesian network for failure assessment of logistics drones.
在步骤5中，所述的构建步骤4获得的物流无人机失效评估贝叶斯网络的条件概率表中的条件概率是指一个事故在另外一个事故已经发生条件下的发生概率，可通过物流无人机运行数据的统计分析和各事故节点的联合分布概率计算公式获得，如式(1)所示：In step 5, the conditional probability in the conditional probability table of the Bayesian network of the logistics drone failure assessment obtained in the construction step 4 refers to the probability of an accident under the condition that another accident has occurred, which can be passed through the logistics The statistical analysis of the UAV operating data and the calculation formula of the joint distribution probability of each accident node are obtained, as shown in formula (1):
Figure PCTCN2021095467-appb-000001
Figure PCTCN2021095467-appb-000001
式中：P为事故节点的条件概率；X i为事故节点编号；n为事故节点X i的所有父节点数量；∏为累计相乘函数；π(X i)为事故节点X i的所有父节点。 Where: P is the conditional probability of accidents node; X i is the node number of accidents; n is the number of all nodes parent node of X i accident; [pi cumulative multiplication function; [pi] (X i) of the accident all parent node X i node.
在步骤6中，所述的依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，计算出不同事故诱因发生时物流无人机发生意外坠落事故的风险的方法是：In step 6, according to the probability value obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of the logistics drone failure assessment obtained in step 5, the logistics when different causes of accidents occur are calculated The method of risk of accidental fall of drones is:
依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，使用贝叶斯网络原理，计算出步骤1、步骤2和步骤3中某一项事故诱因或某几项事故诱因发生时，物流无人机发生意外坠落事故的风险。According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, using Bayesian network principle, calculate step 1, step 2 and step The risk of accidental fall of the logistics drone when one of the accident causes or certain causes of accidents occurs in 3.
本公开提供的基于贝叶斯网络的物流无人机失效风险评估方法的某些实施方案具有如下有益效果：Some implementations of the Bayesian network-based logistics drone failure risk assessment method provided in the present disclosure have the following beneficial effects:
针对物流无人机的系统结构、运行场景和人为因素的特点，全面地分析了造成物流无人机失效的各类诱因，并利用贝叶斯网络将复杂的联合概率分布分解成一系列相对简单的模块，大大降低了知识获取的难度和概率推理的复杂性，适用于物流无人机系统风险评估中不确定性的推理问题，为物流无人机安全风险定量化评估提供了有效方法。In view of the characteristics of the logistics drone's system structure, operating scenarios and human factors, it comprehensively analyzes the various incentives that cause the logistics drone to fail, and uses Bayesian networks to decompose the complex joint probability distribution into a series of relatively simple The module greatly reduces the difficulty of knowledge acquisition and the complexity of probabilistic reasoning. It is suitable for uncertain reasoning problems in the risk assessment of logistics UAV systems, and provides an effective method for quantitative assessment of logistics UAV safety risks.
附图说明Description of the drawings
图1为物流无人机系统结构组成示意图。Figure 1 is a schematic diagram of the structure of the logistics UAV system.
图2为本公开构建的物流无人机失效评估贝叶斯网络示意图。Figure 2 is a schematic diagram of a Bayesian network for failure assessment of logistics drones constructed in the present disclosure.
图3为本公开中通信链路故障工况各事件发生概率分布图。Fig. 3 is a diagram of the probability distribution of occurrence of various events in the communication link failure condition in the present disclosure.
图4为电机故障工况各事件发生概率分布图。Figure 4 shows the probability distribution of each event in the motor fault condition.
具体实施方式Detailed ways
为使本公开的目的、技术方案及优点更加清楚明白，以下根据附图并列举实施例，对本公开做进一步详细说明。In order to make the objectives, technical solutions, and advantages of the present disclosure clearer, the present disclosure will be further described in detail below with reference to the accompanying drawings and examples.
本公开提供的基于贝叶斯网络的物流无人机风险评估方法包括按顺序进行的下列步骤1-6：The Bayesian network-based risk assessment method for logistics drones provided in the present disclosure includes the following steps 1-6 in order:
步骤1：从系统结构角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 1: Analyze and obtain the causes of accidental fall of logistics drones and the probability values of each cause from the perspective of system structure;
由于物流无人机系统结构复杂，各系统故障都会导致物流无人机发生意外坠落事故。物流无人机系统包括由动力系统、飞控系统、电力系统、通信系统和机体结构组成的飞行平台系统、地面控制系统和任务相关的载荷系统。按照图1所示的物流无人机系统的设备零部件构成，依次分析各环节发生故障的形式，根据物流无人机的运行数据，统计分析因系统结构故障造成的物流无人机意外坠落事故的诱因及各诱因的概率值，如表1所示。Due to the complex structure of the logistics drone system, the failure of each system will cause the logistics drone to accidentally fall. The logistics UAV system includes a flight platform system, a ground control system and a mission-related load system composed of a power system, a flight control system, a power system, a communication system, and an airframe structure. According to the equipment components of the logistics drone system shown in Figure 1, the failure modes of each link are analyzed in turn, and the logistics drone accidental fall accident caused by the failure of the system structure is statistically analyzed based on the operating data of the logistics drone. The incentives and the probability values of each incentive are shown in Table 1.
表1、系统结构角度的物流无人机意外坠落事故诱因及相关描述Table 1. Reasons and related descriptions of accidental fall of logistics drones from the perspective of system structure
Figure PCTCN2021095467-appb-000002
Figure PCTCN2021095467-appb-000002
Figure PCTCN2021095467-appb-000003
Figure PCTCN2021095467-appb-000003
步骤2：从运行环境角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 2: From the perspective of operating environment, analyze and obtain the causes of accidental fall of logistics drones and the probability value of each cause;
运行环境原因导致的物流无人机故障主要与运行时的天气条件有关，包括强阵风、强降水和积冰。当天气条件超出物流无人机正常运行的最大指标时，物流无人机的运行将受到较大的影响。根据物流无人机的运行数据，统计分析因运行环境造成的物流无人机意外坠落事故的诱因及各诱因的概率值，如表2所示。The failure of logistics drones caused by operating environment is mainly related to weather conditions during operation, including strong gusts, heavy precipitation and ice accumulation. When weather conditions exceed the maximum index for the normal operation of logistics drones, the operation of logistics drones will be greatly affected. According to the operating data of logistics drones, statistical analysis of the causes of accidental fall of logistics drones caused by the operating environment and the probability of each cause are shown in Table 2.
表2、运行环境角度的物流无人机意外坠落事故诱因及相关描述Table 2. Reasons and related descriptions of accidental fall of logistics drones from the perspective of operating environment
Figure PCTCN2021095467-appb-000004
Figure PCTCN2021095467-appb-000004
Figure PCTCN2021095467-appb-000005
Figure PCTCN2021095467-appb-000005
步骤3：从人为因素角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 3: From the perspective of human factors, analyze and obtain the inducement of the accidental fall of the logistics drone and the probability value of each inducement;
物流无人机失效的人为因素主要由物流无人机操控人员及地面检修人员产生。根据物流无人机的运行数据，统计分析因运行中所涉及的人员及其工作职责造成的物流无人机意外坠落事故的诱因及各诱因的概率值，如表3所示。The human factors for the failure of logistics drones are mainly caused by logistics drone operators and ground maintenance personnel. According to the operating data of logistics drones, statistical analysis of the causes of accidental fall of logistics drones caused by the personnel involved in the operation and their job responsibilities and the probability values of each cause are shown in Table 3.
表3、人为因素角度的物流无人机意外坠落事故诱因及相关描述Table 3. Reasons and related descriptions of accidental fall of logistics drones from the perspective of human factors
Figure PCTCN2021095467-appb-000006
Figure PCTCN2021095467-appb-000006
步骤4：基于步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因构建物流无人机失效评估贝叶斯网络；Step 4: Build a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in step 1, step 2 and step 3;
根据步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因，分别将控制失效、失速、失去全部动力和失去部分动力四项失效模式作为事故节点，按照各自之间的因果关系将各个事故节点通过有向边以单箭头形式连接，其中箭头起始端的事故节点是父节点，箭头末端的事故节点是子节点，由此将失效 模式与诱因通过有向无环图连接起来，基于网络框架的有向无环图描述变量间的依赖和独立关系，构建成物流无人机失效评估贝叶斯网络，如图2所示。According to the inducement of the accidental fall of the logistics drone obtained in step 1, step 2 and step 3, the four failure modes of control failure, stall, loss of all power and loss of partial power are respectively regarded as the accident nodes, according to the causal relationship between them. Each accident node is connected by a directed edge in the form of a single arrow, where the accident node at the beginning of the arrow is the parent node, and the accident node at the end of the arrow is the child node, thus connecting the failure mode and the cause through a directed acyclic graph, The directed acyclic graph based on the network framework describes the dependence and independent relationship between variables, and constructs a Bayesian network for failure assessment of logistics drones, as shown in Figure 2.
步骤5：构建步骤4获得的物流无人机失效评估贝叶斯网络的条件概率表；Step 5: Construct the conditional probability table of the Bayesian network of logistics drone failure evaluation obtained in Step 4;
步骤4所构建的物流无人机失效评估贝叶斯网络包含一组事故节点，将相连的事故节点间的因果关系用条件概率表表示。条件概率是指一个事故在另外一个事故已经发生条件下的发生概率，用来刻画事故节点间的依赖关系。由于物流无人机事故节点间的关系较为复杂，事故节点间的关系并非相互独立的，因此事故节点的条件概率可通过物流无人机运行数据的统计分析和各事故节点的联合分布概率计算公式获得，如式(1)所示：The Bayesian network for failure assessment of logistics drones constructed in step 4 contains a set of accident nodes, and the causal relationship between the connected accident nodes is represented by a conditional probability table. Conditional probability refers to the probability of occurrence of an accident under the condition that another accident has occurred, and it is used to describe the dependency between accident nodes. Because the relationship between the logistics drone accident nodes is more complicated, and the relationship between the accident nodes is not independent of each other, the conditional probability of the accident node can be calculated through the statistical analysis of the logistics drone operation data and the joint distribution probability calculation formula of each accident node Obtained, as shown in formula (1):
Figure PCTCN2021095467-appb-000007
Figure PCTCN2021095467-appb-000007
式中：P为事故节点的条件概率；X i为事故节点编号；n为事故节点X i的所有父节点数量；∏为累计相乘函数；π(X i)为事故节点X i的所有父节点。 Where: P is the conditional probability of accidents node; X i is the node number of accidents; n is the number of all nodes parent node of X i accident; [pi cumulative multiplication function; [pi] (X i) of the accident all parent node X i node.
以人工干预失败、电池故障和失去全部动力之间的条件概率为例，所建立的条件概率表如表4所示。Taking the conditional probability between manual intervention failure, battery failure and loss of all power as an example, the established conditional probability table is shown in Table 4.
表4、人工干预失败、电池故障和失去全部动力之间的条件概率Table 4. Conditional probability between manual intervention failure, battery failure and loss of all power
Figure PCTCN2021095467-appb-000008
Figure PCTCN2021095467-appb-000008
步骤6：依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，计算出不同事故诱因发生时物流无人机发生意外坠落事故的风险；Step 6: According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, calculate the accident of logistics drone when different accident causes occur The risk of a fall accident;
依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，使用贝叶斯网络原理，计算出步骤1、步骤2和步骤3中某一项事故诱因或某几项事故诱因发生时，物流无人机发生意外坠落事故的风险。According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, using Bayesian network principle, calculate step 1, step 2 and step The risk of accidental fall of the logistics drone when one of the accident causes or certain causes of accidents occurs in 3.
以电机故障和通信链路故障两种工况为例，分别计算两种工况下意外坠落事故及各中间事件发生概率。如图3和图4所示，在电机故障工况下，物流无人机意外坠落事故发生概率为21.69％，其中导致底事故发生的中间事故中概率最大的是桨叶失效和无人机失去部分动力，发生概率分别为99.98％和92.18％，而控制失效和人工干预失败发生概率低于1％；在通信链路故障工况下，物流无人机意外坠落事故发生概率为22.14％，其中导致底事故发生的中间事故中概率最大的是人工干预失败，发生概率为97.99％，其次无人机失速和控制失效发生概率也较大。这两种工况的计算结果清晰地展示了不同物流无人机事故诱因向意外坠落事故底事件发展过程中，各中间事件发生概率。Taking two working conditions of motor failure and communication link failure as examples, the probability of accidental fall accidents and intermediate events under the two working conditions are calculated respectively. As shown in Figure 3 and Figure 4, under the condition of motor failure, the probability of accidental fall of logistics drones is 21.69%. Among the intermediate accidents that lead to bottom accidents, the highest probability is blade failure and drone loss. Part of the power, the probability of occurrence is 99.98% and 92.18%, while the probability of control failure and manual intervention failure is less than 1%; under the communication link failure condition, the probability of accidental fall of logistics drones is 22.14%, of which Among the intermediate accidents that cause the bottom accident, the manual intervention failure is the most likely to occur, with a probability of 97.99%. Secondly, the probability of unmanned aerial vehicle stalls and control failures is also greater. The calculation results of these two working conditions clearly show the probability of the occurrence of various intermediate events in the development process of different logistics drone accident inducements to accidental fall accidents.

Claims (4)

  1. 基于贝叶斯网络的物流无人机风险评估方法，其包括按顺序进行的下列步骤：The Bayesian network-based risk assessment method for logistics drones includes the following steps in order:
    步骤1：从系统结构角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 1: Analyze and obtain the causes of accidental fall of logistics drones and the probability values of each cause from the perspective of system structure;
    步骤2：从运行环境角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 2: From the perspective of operating environment, analyze and obtain the causes of accidental fall of logistics drones and the probability value of each cause;
    步骤3：从人为因素角度分析并获得物流无人机意外坠落事故的诱因及各诱因的概率值；Step 3: From the perspective of human factors, analyze and obtain the inducement of the accidental fall of the logistics drone and the probability value of each inducement;
    步骤4：基于步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因构建物流无人机失效评估贝叶斯网络；Step 4: Build a Bayesian network for failure assessment of logistics drones based on the triggers of accidental fall of logistics drones obtained in step 1, step 2 and step 3;
    步骤5：构建步骤4获得的物流无人机失效评估贝叶斯网络的条件概率表；以及Step 5: Construct the conditional probability table of the Bayesian network for the failure assessment of logistics drones obtained in Step 4; and
    步骤6：依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，计算出不同事故诱因发生时物流无人机发生意外坠落事故的风险。Step 6: According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, calculate the accident of logistics drone when different accident causes occur The risk of a fall accident.
  2. 如权利要求1所述的基于贝叶斯网络的物流无人机风险评估方法，其中：在步骤4中，所述的基于步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因构建物流无人机失效评估贝叶斯网络的方法是：The method for risk assessment of logistics drones based on the Bayesian network according to claim 1, wherein: in step 4, the accidental fall of logistics drones based on the accidental fall of logistics drones obtained in step 1, step 2 and step 3 Incentives to build a Bayesian network for failure assessment of logistics drones are:
    根据步骤1、步骤2和步骤3获得的物流无人机意外坠落事故的诱因，分别将控制失效、失速、失去全部动力和失去部分动力四项失效模式作为事故节点，按照各自之间的因果关系将各个事故节点通过有向边以单箭头形式连接，其中箭头起始端的事故节点是父节点，箭头末端的事故节点是子节点，由此将失效 模式与诱因通过有向无环图连接起来，基于网络框架的有向无环图描述变量间的依赖和独立关系，构建成物流无人机失效评估贝叶斯网络。According to the inducement of the accidental fall of the logistics drone obtained in step 1, step 2 and step 3, the four failure modes of control failure, stall, loss of all power and loss of partial power are respectively regarded as the accident nodes, according to the causal relationship between them. Each accident node is connected by a directed edge in the form of a single arrow, where the accident node at the beginning of the arrow is the parent node, and the accident node at the end of the arrow is the child node, thus connecting the failure mode and the cause through a directed acyclic graph, The directed acyclic graph based on the network framework describes the dependence and independent relationship between variables, and constructs a Bayesian network for failure assessment of logistics drones.
  3. 如权利要求1所述的基于贝叶斯网络的物流无人机风险评估方法，其中：在步骤5中，所述的构建步骤4获得的物流无人机失效评估贝叶斯网络的条件概率表中的条件概率是指一个事故在另外一个事故已经发生条件下的发生概率，可通过物流无人机运行数据的统计分析和各事故节点的联合分布概率计算公式获得，如式(1)所示：The method for risk assessment of logistics drones based on the Bayesian network according to claim 1, wherein: in step 5, the conditional probability table of the Bayesian network of logistics drone failure assessment obtained in step 4 is constructed The conditional probability in refers to the occurrence probability of an accident under the condition that another accident has occurred. It can be obtained through the statistical analysis of logistics drone operation data and the joint distribution probability calculation formula of each accident node, as shown in formula (1) :
    Figure PCTCN2021095467-appb-100001
    Figure PCTCN2021095467-appb-100001
    式中：P为事故节点的条件概率；X i为事故节点编号；n为事故节点X i的所有父节点数量；Π为累计相乘函数；π(X i)为事故节点X i的所有父节点。 Where: P is the conditional probability of accidents node; X i is the node number of accidents; n is the number of all nodes parent node of X i accident; [pi cumulative multiplication function; [pi] (X i) of the accident all parent node X i node.
  4. 如权利要求1所述的基于贝叶斯网络的物流无人机风险评估方法，其中：在步骤6中，所述的依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，计算出不同事故诱因发生时物流无人机发生意外坠落事故的风险的方法是：The method for risk assessment of logistics drones based on the Bayesian network according to claim 1, wherein: in step 6, the probability value obtained in step 1, step 2 and step 3 and the logistics obtained in step 5 The conditional probability table of the Bayesian network diagram of UAV failure assessment, and the method to calculate the risk of accidental fall accidents of logistics UAVs when different accident factors occur is:
    依据步骤1、步骤2和步骤3获得的概率值和步骤5获得的物流无人机失效评估贝叶斯网络图的条件概率表，使用贝叶斯网络原理，计算出步骤1、步骤2和步骤3中某一项事故诱因或某几项事故诱因发生时，物流无人机发生意外坠落事故的风险。According to the probability values obtained in step 1, step 2 and step 3 and the conditional probability table of the Bayesian network diagram of logistics drone failure assessment obtained in step 5, using Bayesian network principle, calculate step 1, step 2 and step The risk of accidental fall of the logistics drone when one of the accident causes or certain causes of accidents occurs in 3.
PCT/CN2021/095467 2020-11-09 2021-05-24 Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network WO2021213542A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011239492.1A CN112330161A (en) 2020-11-09 2020-11-09 Logistics unmanned aerial vehicle failure risk assessment method based on Bayesian network
CN202011239492.1 2020-11-09

Publications (1)

Publication Number Publication Date
WO2021213542A1 true WO2021213542A1 (en) 2021-10-28

Family

ID=74316621

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/095467 WO2021213542A1 (en) 2020-11-09 2021-05-24 Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network

Country Status (2)

Country Link
CN (1) CN112330161A (en)
WO (1) WO2021213542A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114363340A (en) * 2022-01-12 2022-04-15 东南大学 Unmanned aerial vehicle cluster failure control method and system and storage medium
CN115876409A (en) * 2023-02-16 2023-03-31 广州中工水务信息科技有限公司 Sewage pipeline leakage monitoring and analyzing system and method
CN116523384A (en) * 2023-04-11 2023-08-01 中国人民解放军陆军工程大学 Unmanned aerial vehicle efficiency determining method and system based on independent overruling and combined overruling

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112330161A (en) * 2020-11-09 2021-02-05 中国民航大学 Logistics unmanned aerial vehicle failure risk assessment method based on Bayesian network
CN113222358A (en) * 2021-04-23 2021-08-06 北京国信云服科技有限公司 Logistics risk assessment model method, device, equipment and medium based on dynamic Bayesian network

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101917297A (en) * 2010-08-30 2010-12-15 烽火通信科技股份有限公司 Method and system for diagnosing faults of core network based on Bayesian network
CN104238545A (en) * 2014-07-10 2014-12-24 中国石油大学(北京) Fault diagnosis and pre-warning system in oil refining production process and establishment method thereof
US20200184324A1 (en) * 2018-04-17 2020-06-11 Hrl Laboratories, Llc Network composition module for a bayesian neuromorphic compiler
US10685064B2 (en) * 2014-04-09 2020-06-16 Introspective Systems LLC Executable graph framework for the management of complex systems
CN111680882A (en) * 2020-05-15 2020-09-18 北京航空航天大学 Multi-level risk assessment method for unmanned aerial vehicle system under probability coupling relation
CN112330161A (en) * 2020-11-09 2021-02-05 中国民航大学 Logistics unmanned aerial vehicle failure risk assessment method based on Bayesian network

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108664400B (en) * 2018-05-11 2020-08-04 北京航空航天大学 Construction method of fault diagnosis technology based on dependent Bayesian network
CN109657880A (en) * 2019-01-09 2019-04-19 中交信息技术国家工程实验室有限公司 A kind of Collision Accidents of Ships grade prediction technique and system based on Bayesian network
CN110991855A (en) * 2019-11-28 2020-04-10 西南交通大学 Bayesian network-based high-speed railway risk assessment method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101917297A (en) * 2010-08-30 2010-12-15 烽火通信科技股份有限公司 Method and system for diagnosing faults of core network based on Bayesian network
US10685064B2 (en) * 2014-04-09 2020-06-16 Introspective Systems LLC Executable graph framework for the management of complex systems
CN104238545A (en) * 2014-07-10 2014-12-24 中国石油大学(北京) Fault diagnosis and pre-warning system in oil refining production process and establishment method thereof
US20200184324A1 (en) * 2018-04-17 2020-06-11 Hrl Laboratories, Llc Network composition module for a bayesian neuromorphic compiler
CN111680882A (en) * 2020-05-15 2020-09-18 北京航空航天大学 Multi-level risk assessment method for unmanned aerial vehicle system under probability coupling relation
CN112330161A (en) * 2020-11-09 2021-02-05 中国民航大学 Logistics unmanned aerial vehicle failure risk assessment method based on Bayesian network

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114363340A (en) * 2022-01-12 2022-04-15 东南大学 Unmanned aerial vehicle cluster failure control method and system and storage medium
CN114363340B (en) * 2022-01-12 2023-12-26 东南大学 Unmanned aerial vehicle cluster failure control method, system and storage medium
CN115876409A (en) * 2023-02-16 2023-03-31 广州中工水务信息科技有限公司 Sewage pipeline leakage monitoring and analyzing system and method
CN116523384A (en) * 2023-04-11 2023-08-01 中国人民解放军陆军工程大学 Unmanned aerial vehicle efficiency determining method and system based on independent overruling and combined overruling
CN116523384B (en) * 2023-04-11 2024-05-14 中国人民解放军陆军工程大学 Unmanned aerial vehicle efficiency determining method and system based on independent overruling and combined overruling

Also Published As

Publication number Publication date
CN112330161A (en) 2021-02-05

Similar Documents

Publication Publication Date Title
WO2021213542A1 (en) Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network
Ni et al. Civil aviation safety evaluation based on deep belief network and principal component analysis
Guo et al. A novel hybrid method for flight departure delay prediction using Random Forest Regression and Maximal Information Coefficient
US8768534B2 (en) Method and apparatus for using electronic flight bag (EFB) to enable flight operations quality assurance (FOQA)
CN109377029B (en) Airplane operation safety risk evaluation method and device, electronic equipment and storage medium
DE112016005185T5 (en) Very short-term air pollution forecast
Wong et al. Data-driven analysis of resilience in airline networks
Xiao et al. Analysis and assessment of risks to public safety from unmanned aerial vehicles using fault tree analysis and Bayesian network
Luxhøj et al. A ConOps derived UAS safety risk model
Wagner et al. Statistical methods for modeling the risk of runway excursions
Xie et al. Fault diagnosis of the satellite power system based on the Bayesian network
CN103617337B (en) A kind of aircraft maintenance interval aided analysis method
Wang et al. Bayesian network assessment method for civil aviation safety based on flight delays
CN115223401B (en) Airport flight area potential conflict assessment method
Acarbay et al. Risk mitigation in unstabilized approach with fuzzy Bayesian bow-tie analysis
Khattak et al. Explainable Boosting Machine for Predicting Wind Shear-Induced Aircraft Go-around based on Pilot Reports
Lee A Predictive Ensemble Model to Minimize Bird Strike Occurrences on Aircrafts at US Airports
Jharko et al. On creating safety control systems for high operation risk plants
Rentian et al. Cause Analysis of Consumer‑Grade UAV Accidents Based on Grounded Theory‑Bayesian Network.
Zhang et al. Failure Mode and Effects Analysis of UAV Power System Based on Generalized Dempster-Shafer Structures
CN113449238A (en) Unmanned aerial vehicle safe operation method and device, electronic equipment and storage medium
Pan et al. Adoption of a bayesian network for the operational reliability analysis of aircraft systems
SZABO et al. Aviation safety investment assessment utilizing return on investment and bayesian networks
Ezzeddine et al. Efficient self-control device based on inspection model for Pitot sensors system
Campbell et al. Benchmark Problem for Autonomous Urban Air Mobility

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21793704

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21793704

Country of ref document: EP

Kind code of ref document: A1