WO2021077983A1 - 一种电梯故障判断逻辑验证方法、系统及存储介质 - Google Patents

一种电梯故障判断逻辑验证方法、系统及存储介质 Download PDF

Info

Publication number
WO2021077983A1
WO2021077983A1 PCT/CN2020/117611 CN2020117611W WO2021077983A1 WO 2021077983 A1 WO2021077983 A1 WO 2021077983A1 CN 2020117611 W CN2020117611 W CN 2020117611W WO 2021077983 A1 WO2021077983 A1 WO 2021077983A1
Authority
WO
WIPO (PCT)
Prior art keywords
fault
elevator
maintenance
judgment logic
data
Prior art date
Application number
PCT/CN2020/117611
Other languages
English (en)
French (fr)
Inventor
江荣钿
黄丹燕
李基源
Original Assignee
日立楼宇技术（广州）有限公司
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 日立楼宇技术（广州）有限公司 filed Critical 日立楼宇技术（广州）有限公司
Priority to JP2022517999A priority Critical patent/JP2022549615A/ja
Publication of WO2021077983A1 publication Critical patent/WO2021077983A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B66—HOISTING; LIFTING; HAULING
    • B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
    • B66B5/00—Applications of checking, fault-correcting, or safety devices in elevators
    • B66B5/0006—Monitoring devices or performance analysers
    • B66B5/0037—Performance analysers

Definitions

  • the invention relates to the technical field of elevator control, in particular to an elevator fault judgment logic verification method, system and storage medium.
  • Elevator failure sources include two types of main control failure reporting and failure diagnosis:
  • the main control failure report is the failure code that the elevator main control collects when each component fails, and uploads it to the center through DTU;
  • DTU collects the operating parameters of each component of the elevator, judges and generates faults according to the established logic or threshold, and uploads it to the center.
  • the second fault judgment logic is implemented based on R&D, engineering experience, elevator specifications, use environment, elevator utilization, etc. Whether its logic and threshold settings are reasonable can only be verified after a long period of operation. A large amount of fault data will be generated during the operation of the elevator, but a large number of false alarms and duplicate data have increased the workload of maintenance personnel. And as the service life of the elevator increases, the originally set logic may no longer be used. Applicable, resulting in misreporting or missing fault information, affecting the operation of components.
  • the purpose of the present invention is to provide an elevator fault judgment logic verification method, system and storage medium, so that the running state of the elevator and the engineering maintenance work can reversely verify the accuracy of the elevator fault.
  • the first technical solution adopted by the present invention is: an elevator fault judgment logic verification method, including the following steps: obtaining elevator operation data; obtaining corresponding maintenance records according to the elevator operation data, the maintenance records including maintenance time, Fault code and processing description; perform data set classification operations on the maintenance records; compare the results of the data set classification with the fault judgment logic; optimize the fault judgment logic according to the comparison results; execute according to the optimized results Elevator failure judgment operation.
  • the elevator operation data includes status parameters, fault records, and elevator shutdown maintenance data; the status parameters include current, voltage, speed, load, mileage, and temperature; the fault records include fault type, fault time, and elevator type And the elevator number; the maintenance data for stopping the elevator includes the maintenance processing time, elevator number and fault code.
  • the step of performing a data set categorization operation on the maintenance record specifically includes: filtering the failure data and the elevator shutdown maintenance data according to the maintenance record; classifying the filtered results; and the classification Including real faults, false alarms, and man-made faults.
  • the step of comparing the result of classification according to the data set with the fault judgment logic specifically includes: screening the state parameters according to the false alarm failure; establishing a feature project according to the screening result; according to the feature The engineering obtains the characteristic value with high weight; the characteristic value is verified with the fault judgment logic.
  • the step of optimizing the elevator fault logic according to the comparison result specifically includes: improving the characteristic dimension of the fault judgment logic according to the verification result of the characteristic value and the fault judgment logic; and according to the characteristic
  • the threshold value of the fault judgment logic is modified as a result of verifying the value and the fault judgment logic; and the deviation of the fault judgment logic is corrected according to the result of verifying the characteristic value and the fault judgment logic.
  • the step of classifying the filtered results specifically includes: obtaining the state parameter according to the processing description; and judging the fault record according to the elevator state parameter.
  • the step of judging the fault record according to the elevator status parameter specifically includes: identifying the processing description according to the NLP; classifying the status parameter according to the identification result; and the processing
  • the description includes the fault condition, fault performance, processing process and processing result.
  • an elevator fault judgment logic verification system includes: a data acquisition unit for acquiring elevator operation data; a maintenance record acquisition unit; for acquiring corresponding maintenance based on the elevator operation data Record, the maintenance record includes maintenance time, failure code and processing description; a data set classification unit, used to perform data set classification operations on the maintenance record; a failure logic judgment unit, used to classify the result of the data set Compare with the fault judgment logic; the fault logic optimization unit is used to optimize the fault logic according to the comparison result; the execution unit is used to perform the elevator fault judgment operation according to the optimized result.
  • an elevator fault judgment logic verification system includes: at least one processor; at least one memory for storing at least one program; when the at least one program is executed by at least one processor , Enabling the at least one processor to implement the logic verification method for elevator failure judgment.
  • the fourth technical solution adopted by the present invention is: a storage medium in which executable instructions are stored, and the processor-executable instructions are used to execute the elevator fault judgment when executed by the processor Logic verification method.
  • the present invention uses a large amount of historical fault data, elevator status data, and maintenance data to form a verification logic, realizes intelligent identification of the authenticity of the fault, and verifies the rationality of the fault judgment logic, reducing The misreporting and omission of fault information can greatly reduce the workload of maintenance personnel and have a wide range of applicability.
  • Figure 1 is a flow chart of the steps of a logic verification method for elevator fault judgment in a specific embodiment of the present invention
  • Fig. 2 is a detailed flow chart of a method for verifying elevator fault judgment logic according to a specific embodiment of the present invention.
  • an embodiment of the present invention provides a method for verifying elevator fault judgment logic, which includes the following steps:
  • the specific embodiment of the present invention collects a large amount of elevator operation data, and obtains the maintenance record of the corresponding failure from the operation data according to the failure data in the operation data and the maintenance data of stopping the elevator, and the maintenance personnel receive the work order, and on-site Confirm and deal with the fault.
  • the process description is backfilled through the PDA.
  • the center receives the maintenance record data, including the elevator number, maintenance time, fault code and processing description.
  • the embodiment of the present invention loads fault history data, corresponding processing data, and elevator shutdown data.
  • the three types of data are associated with elevator numbers and fault codes. According to the fault code and processing description in the maintenance record, it is determined that it specifically caused the elevator failure or
  • the reasons for the maintenance of the elevators are divided into three categories according to the reasons: real fault Dataset_T, false fault Dataset_F and man-made fault Dataset_M.
  • the fault data after the classification and processing in the previous step S103 is divided into true fault Dataset_T, false alarm fault Dataset_F, and man-made fault Dataset_M, select and load various operating data of the false alarm fault Dataset_F among them, and establish the corresponding feature project to obtain the corresponding Compare the characteristic parameters obtained with the parameter thresholds preset in the fault judgment logic.
  • the fault judgment logic runs on the hardware terminal and is based on some parameters collected by the sensor (such as current, stress, counting, and timing). Etc.), based on these parameters and logic, determine whether the elevator is faulty.
  • the judgment logic needs to be re-examined.
  • the judgment logic and threshold value of various elevator fault types are different, and similar faults Different types of elevators may also be different.
  • the dimension of door failure is usually to collect door motor current, door opening time, door closing time, etc.
  • the elevator For example, if the current value exceeds the preset current (threshold value 1) and exceeds the preset number of times (threshold value 2), the elevator The failure of the elevator door opening and closing will be reported, but if the maintenance personnel find that the reason for the excessive current of the door machine is caused by the excessive weight of the door outside the hall, the current threshold must be set a little higher to reduce the occurrence of false alarms; If you analyze the characteristics of real faults and find that the characteristics that cause such faults are inconsistent with the characteristics used in the judgment logic, you must consider whether the judgment logic increases or decreases the characteristics.
  • each ladder type or each type of fault are different, but If a certain type of fault is frequently falsely reported, then analyze the characteristics of the true fault based on the true report data of this type of fault, and use the results of machine learning to reversely verify whether the logic, dimension, or threshold of the fault judgment program is reasonable.
  • the elevator operation data includes status parameters, fault records, and elevator shutdown maintenance data;
  • the status parameters include current, voltage, speed, load, mileage, and temperature;
  • the fault records include fault types, Failure time, elevator type, and elevator number;
  • the maintenance data for stopping elevators includes maintenance processing time, elevator number, and failure code.
  • the embodiment of the present invention acquires status parameters, fault records, and maintenance data during elevator operation, and associates the three data sets according to the elevator number and the fault code, so as to facilitate the analysis work in the subsequent steps.
  • the step S103 of performing a data set classification operation on the maintenance record specifically includes: S1031, performing the fault data and the shutdown maintenance data on the maintenance record according to the maintenance record. Filtering; S1032, classify the filtered results; the classification includes real faults, false alarms, and man-made faults.
  • the embodiment of the present invention loads the data set associated with the elevator number and the fault code; filters the records that have the elevator stop time and the elevator stop time exceeds 2 minutes (2 minutes maintenance personnel cannot rush to the scene and complete the processing) It can be considered that these faults are real faults, save these real faults as data set Dataset_T, analyze the fault processing description, use nlp to analyze the description content, confirm the authenticity of the fault (real fault, false alarm fault), and judge the fault Responsibilities (Is it equipment failure or man-made fault, such as fault description: renovation causes sand to enter the track, which results in blockage of opening and closing the door.
  • the fault After cleaning the sand, the fault is eliminated; such faults can be judged as man-made faults according to the fault description), and all true faults Filter and distinguish the records of man-made faults, and store the real faults of this link in the data set Dataset_T.
  • the step S104 of comparing the result of the data set classification with the fault judgment logic specifically includes: S1041, screening the status parameters according to the false alarm; S1042 Establish a feature project according to the screening result; S1043, obtain a high-weight feature value according to the feature project; S1044, verify the feature value and the fault judgment logic.
  • the embodiment of the present invention loads the data set of Dataset_T, filters out all relevant state parameters of real faults according to the correlation relationship, creates a feature project based on the state parameters of all real faults, and extracts the high weight of the feature project (the strongest correlation relationship) ), the generated characteristic value is compared with the characteristic parameter in the preset fault judgment logic, and the result is fed back.
  • the establishment process of the feature engineering is a general technical direction. In feature engineering, a large amount of sampled data is used to set target tags, and machine learning algorithms are used to analyze the features in the data that lead to the target tags, and obtain each feature the weight of.
  • the step S105 of optimizing the elevator fault logic according to the comparison result specifically includes: S1051, verifying the fault judgment logic according to the result of verification of the characteristic value and the fault judgment logic Improve the feature dimension of the fault judgment logic; S1052, modify the threshold value of the fault judgment logic according to the result of the verification of the characteristic value and the fault judgment logic; S1053, compare the verification result of the fault judgment logic according to the characteristic value and the fault judgment logic Correct the deviation of the fault judgment logic.
  • the two fault types of false alarm fault Dataset_F and real fault Dataset_T are respectively improved and optimized, and for the higher frequency of false alarms
  • the corresponding fault judgment logic needs to be corrected or even reconstructed; for real faults, it is carefully improved and optimized based on the verification results of step S1044, that is, the fault judgment logic should be closer to the results of feature analysis and correlation analysis, if the fault judgment logic If a feature value does not appear or is more than a feature value, the feature dimension needs to be modified; if there is a false alarm, the specific value of the feature parameter in the fault is significantly higher or lower than the feature parameter in the fault judgment logic Value, you need to reset the threshold of the characteristic parameter.
  • the step of classifying the filtered results specifically includes: obtaining the status parameters according to the processing description, and judging the fault records according to the elevator status parameters .
  • the elevator fault judgment is performed, and the faults that actually occur, the faults caused by man-made faults, and the faults falsely reported are distinguished.
  • the step of judging the fault record according to the elevator status parameter specifically includes: identifying the processing description according to NLP; and returning the status parameter according to the identification result.
  • the processing description includes fault conditions, fault performance, processing procedures, and processing results.
  • maintenance personnel need to backfill information about the failure processing process after handling the elevator failure.
  • This information is manually entered to describe the failure status, performance, processing process, processing result, etc., and train an NLP based on the description information in the maintenance record.
  • the natural language analysis) classifier can determine whether the fault corresponding to the repair record is the responsibility of the equipment itself or human responsibility based on the description of the cause in the repair record, and use NLP (Natural Language Analysis) to identify, analyze and classify the repair record, and determine the fault Category, to determine whether each record is accurate and effective (whether it is a human fault).
  • NLP Natural Language Analysis
  • the embodiment of the present invention also provides an elevator fault judgment logic verification system, which includes:
  • Maintenance record acquisition unit used to acquire corresponding maintenance records according to the elevator operation data, the maintenance records including maintenance time, fault code and processing description;
  • the data set classification unit is used to perform a data set classification operation on the maintenance record
  • the fault logic judgment unit is used to compare the result of the data set classification with the fault judgment logic
  • a fault logic optimization unit for optimizing the fault logic according to the comparison result
  • the execution unit is used to execute the elevator fault judgment operation according to the optimized result.
  • the present invention also provides an elevator fault judgment logic verification system, which includes:
  • At least one processor At least one processor
  • At least one memory for storing at least one program
  • the at least one processor implements the method for verifying the elevator fault judgment logic.
  • an embodiment of the present invention also provides a storage medium in which instructions executable by a processor are stored, and the instructions executable by the processor are used to execute the elevator fault judgment logic when executed by the processor. Authentication method.
  • an elevator fault judgment logic verification method, system and storage medium of the present invention have the following advantages:
  • the present invention uses the operating status of the elevator and the engineering maintenance work to reversely verify the accuracy of the elevator failure to deduce whether the failure judgment logic is reasonable and effective, and reduces the false alarms and omissions of the elevator failure information;
  • the method provided according to the present invention can realize the real-time update of the fault judgment logic, which reduces the workload of maintenance personnel and R&D personnel.
  • step numbers in the above method embodiments they are set only for ease of elaboration and description, and the order between the steps is not limited in any way.
  • the execution order of the steps in the embodiments can be performed according to the understanding of those skilled in the art. Adaptive adjustment.

Landscapes

  • Indicating And Signalling Devices For Elevators (AREA)
  • Maintenance And Inspection Apparatuses For Elevators (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

一种电梯故障逻辑验证方法、系统及存储介质，方法包括：获取电梯运行数据；根据电梯运行数据获取对应的维修记录，维修记录包括维修时间、故障代号以及处理描述；对维修记录进行数据集归类操作；根据数据集归类的结果和故障判断逻辑进行对比；根据对比结果对故障逻辑进行优化；根据优化后的结果执行电梯故障判断操作。利用大量的历史故障数据及电梯的状态数据、维修保养数据来形成一个验证逻辑，对故障的真实性实现智能识别，并验证故障判读逻辑的合理性，减少了故障信息误报、漏报的情况，从而极大地减少维保人员的工作量，可广泛应用于电梯控制技术领域。

Description

一种电梯故障判断逻辑验证方法、系统及存储介质 技术领域
本发明涉及电梯控制技术领域，尤其是一种电梯故障判断逻辑验证方法、系统及存储介质。
背景技术
随着电梯物联网发展，需要实现电梯运行数据收集、存储及分析等功能，电梯终端设有DTU(数据转换单元)，将收集到的基础数据进行逻辑处理。其中电梯的故障也是通过DTU实现生成。电梯故障来源包括主控故障发报和故障诊断两类：
1)主控故障发报是电梯主控收集到各个部件出现故障时报的故障码，并通过DTU上传到中心；
2)故障诊断时DTU收集电梯各个部件的运行参数，按照制定的逻辑或阈值，判断生成故障，上传到中心。
第二点的故障判断逻辑，是以研发、工程经验及电梯规格、使用环境、电梯使用率等为基础实现的，其逻辑和阈值设置是否合理，需要经过长时间的运行才能验证。电梯运行过程中会产生大量的故障数据，但大量的误报、重复数据给维保人员增加了巨大的工作量，而且随着电梯使用寿命的增加，原先设定好的逻辑有可能就不再适用，导致故障信息误报、漏报，影响部件的运行。
发明内容
为解决上述技术问题，本发明的目的在于：提供一种电梯故障判断逻辑验证方法、系统及存储介质，可使得电梯的运行状态和工程维保工作能够反向验证电梯故障的准确性。
本发明所采取的第一种技术方案是：一种电梯故障判断逻辑验证方法，包括以下步骤：获取电梯运行数据；根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；对所述维修记录进行数据集归类操作；根据数据集归类的结果和故障判断逻辑进行对比；根据对比结果对所述故障判断逻辑进行优化；根据优化后的结果执行电梯故障判断操作。
进一步，所述电梯运行数据包括状态参数、故障记录和停梯维保数据；所述状态参数包括电流、电压、速度、载重、里程以及温度；所述故障记录包括故障类型、故障时间、电梯类型及电梯号；所述停梯维保数据包括维保处理时间、电梯号及故障代号。
进一步，对所述维修记录进行数据集归类操作这一步骤具体包括：根据所述维修记录对所述故障数据和停梯维保数据进行过滤；将过滤后的结果进行分类处理；所述分类包括真实故障、误报故障以及人为故障。
进一步，所述根据所述数据集归类的结果和故障判断逻辑进行对比这一步骤具体包括：.根据所述误报故障筛选所述状态参数；根据筛选的结果建立特征工程；根据所述特征工程获取高权重的特征值；将所述特征值与所述故障判断逻辑进行验证。
进一步，所述根据对比结果对所述电梯故障逻辑进行优化这一步骤具体包括：根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的特征维度进行改进；根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的阈值进行修改；根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的偏离进行修正。
进一步，将过滤后的结果进行分类处理这一步骤具体包括：根据所述处理描述获取所述状态参数；根据所述电梯状态参数对所述故障记录进行判责。
进一步，所述根据所述电梯状态参数对所述故障记录进行判责这一步骤具体包括：根据NLP对处理描述进行识别操作；根据所述识别结果对所述状态参数进行归类；所述处理描述包括故障状况、故障表现、处理过程和处理结果。
本发明所采取的第二种技术方案是：一种电梯故障判断逻辑验证系统包括：数据获取单元，用于获取电梯运行数据；维修记录获取单元；用于根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；数据集归类单元，用于对所述维修记录进行数据集归类操作；故障逻辑判断单元，用于根据数据集归类的结果和故障判断逻辑进行对比；故障逻辑优化单元，用于根据对比结果对所述故障逻辑进行优化；执行单元，用于根据优化后的结果执行电梯故障判断操作。
本发明所采取的第三种技术方案是：一种电梯故障判断逻辑验证系统包括：至少一个处理器；至少一个存储器，用于存储至少一个程序；当所述至少一个程序被至少一个处理器执行，使得所述至少一个处理器实现所述的一种电梯故障判断逻辑验证方法。
本发明所采取的第四种技术方案是：一种存储介质，其中存储有可执行的指令，所述处理器可执行的指令在由处理器执行时用于执行所述的一种电梯故障判断逻辑验证方法。
本发明的有益效果是：本发明利用大量的历史故障数据及电梯的状态数据、维修保养数据来形成一个验证逻辑，对故障的真实性实现智能识别，并验证故障判读逻辑的合理性，减少了故障信息误报、漏报的情况，从而极大地减少维保人员的工作量，并且具有广泛的适用性。
附图说明
图1为本发明具体实施例的一种电梯故障判断逻辑验证方法步骤流程图；
图2为本发明具体实施例的一种电梯故障判断逻辑验证方法的详细步骤流程图。
具体实施方式
下面结合说明书附图和具体的实施例对本发明进行进一步的详细说明。
如图1所示，本发明实施例提供了一种电梯故障判断逻辑验证方法，包括以下步骤：
S101、获取电梯运行数据；
S102、根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；
具体的，本发明的具体实施例采集大量电梯运行数据，并根据运行数据中的故障数据，以及停梯维保数据从运行数据中获取相应的故障的维修记录，维保人员接收工单，现场确认并处理故障，处理完毕，通过PDA回填处理过程描述，中心接收维修记录数据，包括电梯号、维修时间、故障代号以及处理描述。
S103、对所述维修记录进行数据集归类操作；
具体的，本发明实施例加载故障历史数据及对应的处理数据、停梯数据，三类数据以电梯号、故障代号关联，根据维修记录中的故障代号以及处理描述，判断其具体导致电梯故障或者进行停梯维保的原因所在，根据其原因将所有的故障数据划分为三类：真实故障Dataset_T、误报故障Dataset_F以及人为故障Dataset_M。
S104、根据数据集归类的结果和故障判断逻辑进行对比；
具体的，经过上一步骤S103分类处理后的故障数据分为真实故障Dataset_T、误报故障Dataset_F以及人为故障Dataset_M，选择加载其中的误报故障Dataset_F的各项运行数据，建立对应的特征工程得到相应的特征参数，将获得的特征参数与故障判断逻辑中预设的参数阈值进行比对，其中故障判断逻辑是运行在硬件终端的，依据传感器采集到的一些参数(如电流、应力、计数、计时等)，依据这些参数和逻辑，判断电梯是否故障。
S105、根据对比结果对所述故障逻辑进行优化；
具体的，若特征参数值明显未达到预设的阈值，或存在仅有个别特征参数超过阈值的情况，就需要重新检讨判断逻辑，各种电梯故障类型其判断逻辑和阈值都不一样，同类故障在不同型号的电梯也可能不一样，例如门故障的维度通常是采集门机电流，开门时长，关门时长等，例如电流值超出预设电流(阈值1)超过预设次数(阈值2)，电梯会报梯门开关不顺畅的故障，但如果维修人员现场发现引起门机电流过高的原因是厅外门过重导致的，须将电 流阈值设置得稍高一点，减少误报情况的发生；如果对真实故障进行特征分析，发现导致此类故障的特征与判断逻辑里用到的特征不一致，就要考虑判断逻辑是否增加或减少特征，每种梯型或每种故障的特征都不同，只是如果发现某类故障频繁误报，然后针对这类型故障的真实发报数据分析其导致真实故障的特征，根据机器学习得到的结果反向验证故障判断程序的逻辑、维度或阈值是否合理。
S106、根据优化后的结果执行电梯故障逻辑判断。
作为进一步的优选实施方式，所述电梯运行数据包括状态参数、故障记录和停梯维保数据；所述状态参数包括电流、电压、速度、载重、里程以及温度；所述故障记录包括故障类型、故障时间、电梯类型及电梯号；所述停梯维保数据包括维保处理时间、电梯号、及故障代号。
具体的，本发明实施例获取电梯运行期间的状态参数、故障记录以及停梯保修数据，根据电梯号和故障代号将三个数据集进行关联，便于后面步骤的分析工作。
如图2所示，作为进一步的优选实施方式，对所述维修记录进行数据集归类操作这一步骤S103具体包括：S1031、根据所述维修记录对所述故障数据和停梯维保数据进行过滤；S1032、将过滤后的结果进行分类处理；所述分类包括真实故障、误报故障以及人为故障。
具体的，本发明实施例加载以电梯号和故障代号关联后的数据集；过滤有停梯时间、而且停梯时间超过2分钟的记录(2分钟维保人员不可能赶到现场并处理完成)，可以认为这些故障是真实的故障，把这些真实故障存为数据集Dataset_T，分析故障处理描述，利用nlp对描述内容分析，确认故障的真实性(真实故障、误报故障)，并对故障判责(是设备故障还是人为故障，例如故障描述：装修导致沙子进入轨道，导致开关门阻滞，清理沙子后，故障排除；根据故障描述可将此类故障判定为人为故障)，将所有真实故障和人为故障的记录进行过滤区分，并把该环节的真实故障存入数据集Dataset_T。
如图2所示，作为进一步的优选实施方式，根据所述数据集归类的结果和故障判断逻辑进行对比这一步骤S104具体包括：S1041、根据所述误报故障筛选所述状态参数；S1042根据筛选的结果建立特征工程；S1043、根据所述特征工程获取高权重的特征值；S1044、将所述特征值与所述故障判断逻辑进行验证。
具体的、本发明实施例加载Dataset_T的数据集，并根据关联关系筛选出所有真实故障的相关状态参数，根据所有的真实故障的状态参数创建特征工程，提取出特征工程中高权重(相关关系最强)的特征值，产生的特征值与预设的故障判断逻辑中的特征参数进行对比，反馈结果。所述特征工程的建立过程为通用的技术方向，在特征工程中，通过大量的采样数据， 设置目标标签，通过机器学习算法，会把数据中导致目标标签的特征分析出来，并得到每个特征的权重。
作为进一步的优选实施方式，所述根据所述对比结果对所述电梯故障逻辑进行优化这一步骤S105具体包括：S1051、根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的特征维度进行改进；S1052、根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的阈值进行修改；S1053、根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的偏离进行修正。
具体的，根据S1032故障分类处理后得到的误报故障以及上一步骤S1044的验证结果，针对误报故障Dataset_F和真实故障Dataset_T两种故障类型分别进行改进和优化，针对出现频率较高的误报故障，需将对应的故障判断逻辑进行修正，甚至重构；针对真实故障，其根据步骤S1044的验证结果精心改进优化，即故障判断逻辑要向特征分析和关联分析的结果靠拢，若故障判断逻辑中没有出现某一项特征值或多出某一项特征值，则需要对特征维度进修改；若出现误报故障中的特征参数的具体数值明显高于或低于故障判断逻辑中特征参数的值，则需要对特征参数的阈值进行重新设定。
作为进一步的优选实施方式，所述将所述过滤后的结果进行分类处理这一步骤具体包括：根据所述处理描述获取所述状态参数，根据所述电梯状态参数对所述故障记录进行判责。
具体的，根据所有的故障记录关联维修记录之后，进行电梯故障判责，区分真实发生的故障、人为造成的故障以及误报故障。
作为进一步的优选实施方式，所述根据所述电梯状态参数对所述故障记录进行判责这一步骤具体包括：根据NLP对处理描述进行识别操作；根据所述识别结果对所述状态参数进行归类；所述处理描述包括故障状况、故障表现、处理过程和处理结果。
具体的，维保人员处理电梯故障后需要回填故障处理过程信息，这些信息是手工录入，是对故障状况、表现、处理过程、处理结果等进行描述，根据维修记录中的描述信息训练一个NLP(自然语言分析)分类器，能够基于维修记录中的原因描述判定该维修记录对应的故障是否是设备本身责任还是人为责任，利用NLP(自然语言分析)对维修记录进行识别、分析归类，判断故障类别，判定每一条记录是否准确有效(是否为人为故障)。
本发明实施例还提供了一种电梯故障判断逻辑验证系统，包括：
数据获取单元，用于获取电梯运行数据；
维修记录获取单元；用于根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；
数据集归类单元，用于对所述维修记录进行数据集归类操作；
故障逻辑判断单元，用于根据数据集归类的结果和故障判断逻辑进行对比；
故障逻辑优化单元，用于根据对比结果对所述故障逻辑进行优化；
执行单元，用于根据优化后的结果执行电梯故障判断操作。
本发明还提供一种电梯故障判断逻辑验证系统，包括：
至少一个处理器；
至少一个存储器，用于存储至少一个程序；
当所述至少一个程序被至少一个处理器执行，使得所述至少一个处理器实现所述的一种电梯故障判断逻辑验证方法。
上述方法实施例中的内容均适用于本系统实施例中，本系统实施例所具体实现的功能与上述方法实施例相同，并且达到的有益效果与上述方法实施例所达到的有益效果也相同。
此外，本发明实施例还提供了一种存储介质，其中存储有处理器可执行的指令，所述处理器可执行的指令在由处理器执行时用于执行所述的一种电梯故障判断逻辑验证方法。
相对于现有技术，本发明一种电梯故障判断逻辑验证方法、系统及存储介质具有以下优点：
1)、本发明利用电梯的运行状态和工程维保工作，反向验证电梯故障的准确性，以推导出故障判断逻辑是否合理有效，减少了电梯故障信息的误报以及漏报；
2)、根据本发明提供的方法可实现故障判断逻辑的实时更新，减少了维保人员以及研发人员的工作量。
对于上述方法实施例中的步骤编号，其仅为了便于阐述说明而设置，对步骤之间的顺序不做任何限定，实施例中的各步骤的执行顺序均可根据本领域技术人员的理解来进行适应性调整。
以上是对本发明的较佳实施进行了具体说明，但本发明并不限于所述实施例，熟悉本领域的技术人员在不违背本发明精神的前提下还可做作出种种的等同变形或替换，这些等同的变形或替换均包含在本申请权利要求所限定的范围内。

Claims (10)

  1. 一种电梯故障判断逻辑验证方法，其特征在于，包括以下步骤：
    获取电梯运行数据；
    根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；
    对所述维修记录进行数据集归类操作；
    根据数据集归类的结果和故障判断逻辑进行对比；
    根据对比结果对所述故障判断逻辑进行优化；
    根据优化后的结果执行电梯故障判断操作。
  2. 根据权利要求1所述的一种电梯故障判断逻辑验证方法，其特征在于：所述电梯运行数据包括状态参数、故障记录和停梯维保数据；
    所述状态参数包括电流、电压、速度、载重、里程以及温度；
    所述故障记录包括故障类型、故障时间、电梯类型及电梯号；
    所述停梯维保数据包括维保处理时间、电梯号及故障代号。
  3. 根据权利要求1所述的一种电梯故障判断逻辑验证方法，其特征在于：所述对所述维修记录进行数据集归类操作这一步骤具体包括：
    根据所述维修记录对所述故障数据和停梯维保数据进行过滤；
    将过滤后的结果进行分类处理；所述分类包括真实故障、误报故障以及人为故障。
  4. 根据权利要求3所述的一种电梯故障判断逻辑验证方法，其特征在于：所述根据数据集归类的结果和故障判断逻辑进行对比这一步骤具体包括：
    根据所述误报故障筛选所述状态参数；
    根据筛选的结果建立特征工程；
    根据所述特征工程获取高权重的特征值；
    将所述特征值与故障判断逻辑进行验证。
  5. 根据权利要求4所述的一种电梯故障判断逻辑验证方法，其特征在于，所述根据对比结果对所述电梯故障逻辑进行优化这一步骤具体包括：
    根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的特征维度进行改进；
    根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的阈值进行修改；
    根据所述特征值与所述故障判断逻辑进行验证的结果对故障判断逻辑的偏离进行修正。
  6. 根据权利要求1至3任一项所述的一种电梯故障判断逻辑验证方法，其特征在于：所述 将过滤后的结果进行分类处理这一步骤具体包括：
    根据所述处理描述获取所述状态参数；
    根据所述电梯状态参数对所述故障记录进行判责。
  7. 根据权利要求6所述的一种电梯故障判断逻辑验证方法，其特征在于：所述根据所述电梯状态参数对所述故障记录进行判责这一步骤具体包括：
    根据NLP对处理描述进行识别操作；
    根据所述识别结果对所述状态参数进行归类；
    所述处理描述包括故障状况、故障表现、处理过程和处理结果。
  8. 一种电梯故障判断逻辑验证系统，其特征在于，包括：
    数据获取单元，用于获取电梯运行数据；
    维修记录获取单元；用于根据所述电梯运行数据获取对应的维修记录，所述维修记录包括维修时间、故障代号以及处理描述；
    数据集归类单元，用于对所述维修记录进行数据集归类操作；
    故障逻辑判断单元，用于根据数据集归类的结果和故障判断逻辑进行对比；
    故障逻辑优化单元，用于根据对比结果对所述故障判断逻辑进行优化；
    执行单元，用于根据优化后的结果执行电梯故障判断操作。
  9. 一种电梯故障判断逻辑验证系统，其特征在于，包括：
    至少一个处理器；
    至少一个存储器，用于存储至少一个程序；
    当所述至少一个程序被至少一个处理器执行，使得所述至少一个处理器实现如权利要求1-7任一项所述的一种电梯故障判断逻辑验证方法。
  10. 一种存储介质，其中存储有可执行的指令，其特征在于，所述处理器可执行的指令在由处理器执行时用于执行如权利要求1-7任一项所述的一种电梯故障判断逻辑验证方法。
PCT/CN2020/117611 2019-10-23 2020-09-25 一种电梯故障判断逻辑验证方法、系统及存储介质 WO2021077983A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2022517999A JP2022549615A (ja) 2019-10-23 2020-09-25 エレベーター故障判定論理検証方法、システムおよび記憶媒体

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911010901.8A CN110861987B (zh) 2019-10-23 2019-10-23 一种电梯故障判断逻辑验证方法、系统及存储介质
CN201911010901.8 2019-10-23

Publications (1)

Publication Number Publication Date
WO2021077983A1 true WO2021077983A1 (zh) 2021-04-29

Family

ID=69653102

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/117611 WO2021077983A1 (zh) 2019-10-23 2020-09-25 一种电梯故障判断逻辑验证方法、系统及存储介质

Country Status (3)

Country Link
JP (1) JP2022549615A (zh)
CN (1) CN110861987B (zh)
WO (1) WO2021077983A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113682912A (zh) * 2021-09-09 2021-11-23 重庆伊士顿电梯有限责任公司 一种具有故障预警的智能电梯安全监控系统
CN113743782A (zh) * 2021-09-03 2021-12-03 重庆市特种设备检测研究院 一种基于统筹法的电梯检验工序优化方法及优化系统
CN114283502A (zh) * 2021-12-08 2022-04-05 福建省特种设备检验研究院泉州分院 一种特种设备传感器节点数据分析方法
CN115196456A (zh) * 2022-07-14 2022-10-18 慧川电梯科技有限公司 一种适用于不同品牌电梯的故障判断方法及装置
CN115973125A (zh) * 2023-02-15 2023-04-18 慧铁科技有限公司 一种处理铁路货车脱轨自动制动装置故障的方法
CN117270664A (zh) * 2023-11-23 2023-12-22 深圳市蓝鲸智联科技股份有限公司 一种基于汽车智能存储芯片复位系统

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110861987B (zh) * 2019-10-23 2021-06-08 日立楼宇技术（广州）有限公司 一种电梯故障判断逻辑验证方法、系统及存储介质
CN112723069B (zh) * 2020-12-16 2023-06-09 长沙慧联智能科技有限公司 基于tof视觉检测的电梯门运行状态监测方法及系统
CN113343176B (zh) * 2021-08-05 2022-06-10 北京磁浮有限公司 基于模糊综合评价的电梯设备故障预警方法
CN113469580A (zh) * 2021-08-06 2021-10-01 广东省林业科学研究院 生态修复设备系统的能耗管理方法
CN113837407A (zh) * 2021-10-27 2021-12-24 北京恒远国创科技有限公司 一种电梯维保管理方法、计算机设备及存储介质
CN115028036A (zh) * 2022-05-06 2022-09-09 北京中铁电梯工程有限公司 一种基于大数据的电梯管理方法
CN115101187B (zh) * 2022-07-14 2022-11-15 西南医科大学附属医院 一种基于大数据的麻醉机运行故障预测系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104408656A (zh) * 2014-10-29 2015-03-11 中国建设银行股份有限公司 动态调整流控阈值的方法及系统
JP6321075B2 (ja) * 2016-05-27 2018-05-09 東芝エレベータ株式会社 エレベータ制御装置
CN108320040A (zh) * 2017-01-17 2018-07-24 国网重庆市电力公司 基于贝叶斯网络优化算法的采集终端故障预测方法及系统
CN110334728A (zh) * 2019-05-06 2019-10-15 中国联合网络通信集团有限公司 一种面向工业互联网的故障预警方法及装置
CN110861987A (zh) * 2019-10-23 2020-03-06 日立楼宇技术（广州）有限公司 一种电梯故障判断逻辑验证方法、系统及存储介质

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3573559B2 (ja) * 1996-03-04 2004-10-06 株式会社日立製作所 保守支援システム
JP2009053938A (ja) * 2007-08-27 2009-03-12 Toshiba Corp 複数モデルに基づく設備診断システム及びその設備診断方法
CN104714175A (zh) * 2013-12-12 2015-06-17 北京有色金属研究总院 电池系统故障诊断方法及系统
CN105775943A (zh) * 2016-04-12 2016-07-20 广州深度数据科技有限公司 一种数据驱动的电梯零部件预警系统及方法
CN106199251B (zh) * 2016-06-24 2018-10-16 广东电网有限责任公司佛山供电局 一种基于自适应建模分析的配电网故障预警系统及方法
WO2018123037A1 (ja) * 2016-12-28 2018-07-05 三菱電機ビルテクノサービス株式会社 エレベーターの遠隔監視装置
CN108217364B (zh) * 2018-01-05 2019-06-04 日立楼宇技术（广州）有限公司 终端分配模型建立、目标电梯调试终端确定方法及装置
EP3650388A1 (en) * 2018-11-06 2020-05-13 KONE Corporation A method and a system for detecting a malfunction of an elevator system
CN109885951A (zh) * 2019-02-28 2019-06-14 中科云创(厦门)科技有限公司 设备故障诊断方法及装置
CN110058133B (zh) * 2019-04-15 2021-03-02 杭州拓深科技有限公司 一种基于反馈机制的电气线路故障电弧误报优化方法
CN109987473B (zh) * 2019-04-16 2021-06-15 日立楼宇技术（广州）有限公司 一种控制电梯报错的方法、装置、设备和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104408656A (zh) * 2014-10-29 2015-03-11 中国建设银行股份有限公司 动态调整流控阈值的方法及系统
JP6321075B2 (ja) * 2016-05-27 2018-05-09 東芝エレベータ株式会社 エレベータ制御装置
CN108320040A (zh) * 2017-01-17 2018-07-24 国网重庆市电力公司 基于贝叶斯网络优化算法的采集终端故障预测方法及系统
CN110334728A (zh) * 2019-05-06 2019-10-15 中国联合网络通信集团有限公司 一种面向工业互联网的故障预警方法及装置
CN110861987A (zh) * 2019-10-23 2020-03-06 日立楼宇技术（广州）有限公司 一种电梯故障判断逻辑验证方法、系统及存储介质

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113743782A (zh) * 2021-09-03 2021-12-03 重庆市特种设备检测研究院 一种基于统筹法的电梯检验工序优化方法及优化系统
CN113682912A (zh) * 2021-09-09 2021-11-23 重庆伊士顿电梯有限责任公司 一种具有故障预警的智能电梯安全监控系统
CN114283502A (zh) * 2021-12-08 2022-04-05 福建省特种设备检验研究院泉州分院 一种特种设备传感器节点数据分析方法
CN114283502B (zh) * 2021-12-08 2023-06-23 福建省特种设备检验研究院泉州分院 一种特种设备传感器节点数据分析方法
CN115196456A (zh) * 2022-07-14 2022-10-18 慧川电梯科技有限公司 一种适用于不同品牌电梯的故障判断方法及装置
CN115196456B (zh) * 2022-07-14 2023-08-18 慧川电梯科技有限公司 一种适用于不同品牌电梯的故障判断方法及装置
CN115973125A (zh) * 2023-02-15 2023-04-18 慧铁科技有限公司 一种处理铁路货车脱轨自动制动装置故障的方法
CN117270664A (zh) * 2023-11-23 2023-12-22 深圳市蓝鲸智联科技股份有限公司 一种基于汽车智能存储芯片复位系统

Also Published As

Publication number Publication date
CN110861987A (zh) 2020-03-06
CN110861987B (zh) 2021-06-08
JP2022549615A (ja) 2022-11-28

Similar Documents

Publication Publication Date Title
WO2021077983A1 (zh) 一种电梯故障判断逻辑验证方法、系统及存储介质
US20150346066A1 (en) Asset Condition Monitoring
US8248228B2 (en) Method and device for optimizing the alarm configuration
EP3105644B1 (en) Method of identifying anomalies
KR102247181B1 (ko) Xai에 기초하여 생성된 학습데이터를 이용한 이상행위탐지모델 생성방법 및 장치
US9146800B2 (en) Method for detecting anomalies in a time series data with trajectory and stochastic components
CN108763966B (zh) 一种尾气检测作弊监管系统及方法
CN116991947B (zh) 一种自动化数据同步方法及系统
CN117235743B (zh) 一种基于安全风险的智慧电力管理方法及系统
TW202101141A (zh) 設備異常偵測方法及系統
CN113723827A (zh) 地铁机电设备运行风险诊断及运维管控方法与系统
CN111651760B (zh) 一种设备安全状态综合分析的方法及计算机可读存储介质
CN110727669B (zh) 一种电力系统传感器数据清理装置及清理方法
CN117113135A (zh) 一种可对异常数据整理分类的碳排放异常监测分析系统
CN110399926B (zh) 一种路灯故障诊断方法及装置
CN117435908A (zh) 一种用于旋转机械的多种故障特征提取方法
WO2021038079A1 (de) Verfahren und vorrichtung zum analysieren eines ablaufprozesses
CN105814546B (zh) 用于辅助对算法链的检验和验证的方法和系统
CN108268017A (zh) 制罐生产线效率的评估方法和优化方法及其评估装置
CN110543830A (zh) 动作检测方法、装置、存储介质
CN117095242B (zh) 一种基于机器视觉的智能建筑垃圾分类方法及系统
Pipe et al. An automated data-driven toolset for predictive analytics
CN114637793B (zh) 一种基于大数据分析的装备故障频发区域定位方法
CN110599030B (zh) 一种基于正反特征库加权匹配的电力风险预警方法
CN117783795A (zh) 边缘分析的换流变阀侧套管绝缘状态综合分析方法及系统

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: 20878672

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2022517999

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20878672

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 15/06/2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20878672

Country of ref document: EP

Kind code of ref document: A1