Elevator fault determination logic verification method, system, and storage medium

Abstract

The present invention discloses an elevator fault logic verification method, system, and storage medium. The method includes acquiring elevator operation data, acquiring corresponding maintenance records based on the elevator operation data, including maintenance times, fault codes, and treatment instructions, performing a data set classification operation on the maintenance records, comparing the results of the data set classification with fault determination logic, optimizing the fault logic based on the comparison results, and performing an elevator fault determination operation based on the optimized result. The verification logic is formed using a large amount of historical fault data, elevator status data, and maintenance data, enabling intelligent identification of the authenticity of faults and verifying the appropriateness of the fault determination logic. This reduces false and missing reports of fault information, significantly reducing the workload of maintenance personnel. The method can be widely applied in the field of elevator control technology.

Description

[Technical Field]

[0001] The present invention relates to the technical field of elevator control, and more particularly to an elevator fault determination logic verification method, system, and storage medium. [Background technology]

[0002] With the development of elevator Internet of Things, it is necessary to realize functions such as collecting, storing, and analyzing elevator operation data. Elevator terminals are equipped with a DTU (Data Transformation Unit) to perform logical processing of the collected base data. Among them, elevator fault generation is also realized by the DTU. Elevator fault generation sources include the following two types: master control fault reporting and fault diagnosis.

[0003] 1) Master control fault reporting refers to the process whereby the elevator master control collects the fault codes reported when a fault occurs in each component and uploads them to the center via the DTU.

[0004] 2) Fault diagnosis refers to the process of using the DTU to collect operating parameters for each component of the elevator, determining that a fault has occurred based on a predetermined logic or threshold, and uploading the results to the center.

[0005] Regarding the second point, the logic for fault detection is based on research and development, construction experience, elevator specifications, operating environment, elevator utilization rate, etc., and the appropriateness of the logic and threshold settings can only be verified after long-term operation. A large amount of fault data is generated during elevator operation, and the large amount of erroneous and duplicated data significantly increases the workload of maintenance personnel. In addition, as the elevator's service life increases, the preset logic may no longer apply, leading to erroneous or missed reports of fault information and potentially affecting the operation of the components. Summary of the Invention [Problem to be solved by the invention]

[0006] In order to solve the above-mentioned technical problems, the present invention aims to provide an elevator fault determination logic verification method, system, and storage medium, which can reverse-verify the accuracy of elevator faults based on the elevator's operating status and construction maintenance work. [Means for solving the problem]

[0007] The first technical solution adopted in the present invention is an elevator fault judgment logic verification method, which includes the steps of: acquiring elevator operation data; acquiring corresponding maintenance records, including maintenance time, fault code, and processing description, based on the elevator operation data; performing a data set classification operation on the maintenance records; comparing based on the data set classification result and fault judgment logic; optimizing the fault judgment logic based on the comparison result; and performing an elevator fault judgment operation based on the optimized result.

[0008] Further, the elevator operation data includes state parameters, Failure Data and shutdown maintenance data, wherein the status parameters include current, voltage, speed, load, mileage, and temperature, Failure Data includes a fault type, a fault time, an elevator type and an elevator number, and the out-of-service maintenance data includes a maintenance processing time, an elevator number and a fault code.

[0009] Furthermore, the step of performing a data set classification operation on the maintenance records specifically includes: filtering the failure data and operational shutdown maintenance data based on the maintenance records; and classifying the filtered results, wherein the classification includes actual failures, falsely reported failures, and human failures.

[0010] Furthermore, the step of comparing based on the result of the data set classification and the fault determination logic specifically includes: selecting the state parameters based on the falsely reported fault; and based on the selection result, Feature Engineering and establishing Feature Engineering and verifying the feature value and a fault determination logic.

[0011] Furthermore, the step of optimizing the elevator fault logic based on the comparison result specifically includes: improving a feature dimension of the fault determination logic based on a result of verifying the feature value and the fault determination logic; correcting a threshold of the fault determination logic based on a result of verifying the feature value and the fault determination logic; and correcting an offset of the fault determination logic based on a result of verifying the feature value and the fault determination logic.

[0012] Furthermore, the step of classifying the filtered result specifically includes: obtaining the state parameters based on the processing description; and Failure Data and determining:

[0013] Furthermore, based on the elevator state parameters, Failure Data The step of determining specifically includes: identifying a process description based on NLP; and classifying the state parameters based on the identification result, where the process description includes a fault situation, a fault expression, a process, and a process result.

[0014] The second technical solution adopted in the present invention is an elevator fault judgment logic verification system, including: a data acquisition unit for acquiring elevator operation data; a maintenance record acquisition unit for acquiring corresponding maintenance records, including maintenance time, fault code and processing description, based on the elevator operation data; a dataset classification unit for performing a dataset classification operation on the maintenance records; a fault logic judgment unit for comparing based on the dataset classification result and the fault judgment logic; a fault logic optimization unit for optimizing the fault judgment logic based on the comparison result; and an execution unit for executing the elevator fault judgment operation based on the optimized result.

[0015] The third technical solution adopted in the present invention is an elevator fault judgment logic verification system, including at least one processor and at least one memory for storing at least one program, When the at least one program is executed by at least one processor, the elevator fault determination logic verification system causes the at least one processor to realize the elevator fault determination logic verification method.

[0016] The fourth technical solution adopted in the present invention is a storage medium storing executable commands, which is used to execute the elevator fault determination logic verification method when the processor-executable commands are executed by a processor. [Effects of the Invention]

[0017] The beneficial effects of the present invention are as follows: 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 faults, verifies the appropriateness of the fault judgment logic, and reduces erroneous reporting and omission of fault information, thereby significantly reducing the workload of maintenance personnel and having wide applicability. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 1 is a flowchart of an elevator fault determination logic verification method according to a specific embodiment of the present invention. [Figure 2] FIG. 2 is a detailed flowchart of the elevator fault determination logic verification method in a specific embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0019] The present invention will be described in more detail below in conjunction with the drawings and specific examples.

[0020] As shown in FIG. 1, in an embodiment of the present invention, there is provided an elevator fault determination logic verification method, which includes the following steps:

[0021] S101: Elevator operation data is acquired.

[0022] S102: According to the elevator operation data, obtain corresponding maintenance records, including maintenance time, fault code and processing description.

[0023] Specifically, in a specific embodiment of the present invention, a large amount of elevator operation data is collected, and based on the fault data and operation stop maintenance data in the operation data, the corresponding fault maintenance record is obtained from the operation data, so that the maintenance personnel accepts the work order, checks and handles the fault on site, and after the processing is completed, fills in the processing process description using the PDA, and receives the maintenance record data including the elevator number, maintenance time, fault code and processing description from the center.

[0024] S103: A data set classification operation is performed on the maintenance records.

[0025] Specifically, in this embodiment, the fault history data, corresponding processing data, and operational shutdown data are loaded. These three types of data are associated with the elevator number and fault code. Based on the fault code and processing description in the maintenance record, the specific cause of the elevator failure or operational shutdown maintenance is determined, and all the fault data are classified into Actual Faults Dataset_T, False Reported Faults Dataset_F, and Man-Made Faults Dataset_M based on the cause.

[0026] S104: Comparing the data sets based on the classification results and fault determination logic.

[0027] Specifically, the fault data classified in the previous step S103 is divided into an actual fault Dataset_T, a falsely reported fault Dataset_F, and a human-caused fault Dataset_M. The operation data of each of the falsely reported fault Dataset_F is selected and loaded, and the corresponding Feature Engineering and obtain the corresponding characteristic parameters. The obtained characteristic parameters are compared with the parameter thresholds preset in the fault determination logic. The fault determination logic is implemented in the hardware terminal and is a part of the parameters (such as current, stress, count, time, etc.) collected by the sensors, and determines whether the elevator has a fault based on these parameters and logic.

[0028] S105: Optimize the fault logic based on the comparison result.

[0029] Specifically, if the feature parameter values clearly do not reach the preset thresholds or if only a few feature parameters exceed the thresholds, the judgment logic needs to be reconsidered. The judgment logic and thresholds for various elevator fault types are not all the same. The same fault may not be the same for different elevator models. For example, the dimensions of a door fault typically include the door motor current, door opening time, and door closing time. For example, if the current value exceeds a predetermined current (threshold 1) and a predetermined number of times (threshold 2), the elevator reports a fault indicating that the doors are not opening and closing smoothly. However, if maintenance personnel on-site discover that the excessively high door motor current is caused by excessive weight on the outer door of the hall, the current threshold should be set higher to reduce the occurrence of false reports. If a feature analysis of an actual fault reveals that the features causing such a fault do not match the features used in the judgment logic, it is necessary to consider whether to add or remove features from the judgment logic. Although the characteristics of various elevator types or various types of faults are different, if it is discovered that a certain fault is frequently misreported, the actual report data of this type of fault can be analyzed to determine the characteristics that cause the actual fault, and based on the results of machine learning, the logic, dimensions, or thresholds of the fault judgment program can be reverse-verified to determine whether they are appropriate.

[0030] S106: Based on the optimized result, execute elevator fault logic judgment.

[0031] In a further preferred embodiment, the elevator operation data includes state parameters: Failure Data and shutdown maintenance data, wherein the status parameters include current, voltage, speed, load, mileage, and temperature, Failure Data includes a fault type, a fault time, an elevator type and an elevator number, and the out-of-service maintenance data includes a maintenance processing time, an elevator number and a fault code.

[0032] Specifically, in an embodiment of the present invention, the state parameters during elevator operation are: Failure Data and outage maintenance data are acquired, and the three data sets are associated based on elevator number and fault code, which is convenient for the analysis work in the steps described below.

[0033] 2, in another preferred embodiment, the step S103 of performing a data set classification operation on the maintenance records specifically includes: filtering the failure data and operational shutdown maintenance data based on the maintenance records in S1031; and classifying the filtered results in S1032. The classification includes actual failures, falsely reported failures, and human failures.

[0034] Specifically, in this embodiment of the present invention, a dataset associated with elevator numbers and fault codes is loaded. Records of downtimes and records of downtimes exceeding two minutes (i.e., maintenance personnel are unable to arrive at the site and complete the repair within two minutes) are filtered out, allowing these faults to be considered actual faults. These actual faults are stored in a dataset Dataset_T, and the fault handling descriptions are analyzed. The authenticity of the fault (actual fault or falsely reported fault) is confirmed by analyzing the contents of the descriptions using NLP. The fault is then determined to be either an equipment fault or a human fault (e.g., a fault description might be that sand from decorations got into the rails, preventing the doors from opening and closing. After the sand was removed, the fault was eliminated. Based on the fault description, such a fault may be determined to be a human fault). All actual fault and human fault records are filtered out to distinguish them, and the actual faults at this stage are stored in a dataset Dataset_T.

[0035] As shown in FIG. 2 , in a further preferred embodiment, the step S104 of comparing based on the result of the data set classification and the fault determination logic specifically includes: S1041 of selecting the state parameters based on the falsely reported fault; and S1042 of selecting the state parameters based on the result of the selection. Feature Engineering S1042 which establishes the Feature Engineering and S1044 of verifying the feature value and the fault determination logic.

[0036] Specifically, in an embodiment of the present invention, the dataset Dataset_T is loaded, and the related state parameters of all the actual faults are selected based on the related relationships. Based on the state parameters of all the actual faults, Feature Engineering Establishing Feature Engineering The feature value with the highest weight (the strongest correlation) is extracted from the above, the generated feature value is compared with the feature parameter in the preset fault determination logic, and the result is fed back. Feature Engineering The establishment process of is a general-purpose technical direction, Feature Engineering In this method, a target label is set using a large amount of sampling data, and a machine learning algorithm is used to analyze and extract the features in the data that cause the target label, and obtain the weight of each feature.

[0037] In a further preferred embodiment, the step S105 of optimizing the elevator fault logic based on the comparison result specifically includes: S1051: improving a feature dimension of the fault judgment logic based on a result of verifying the feature value and the fault judgment logic; S1052: correcting a threshold value of the fault judgment logic based on a result of verifying the feature value and the fault judgment logic; and S1053: correcting an offset of the fault judgment logic based on a result of verifying the feature value and the fault judgment logic.

[0038] Specifically, based on the misreported faults classified in S1032 and the verification results in the previous step S1044, two types of faults, Dataset_F for misreported faults and Dataset_T for actual faults, are refined and optimized. For misreported faults with high occurrence frequencies, the corresponding fault determination logic must be modified and even reconstructed. For actual faults, the logic must be carefully refined and optimized based on the verification results in step S1044. That is, the fault determination logic must approach the results of the feature analysis and related analysis. If a feature value does not appear in the fault determination logic or if there is one more feature value, the feature dimension must be revised. If the specific value of the feature parameter for the misreported fault is higher or lower than the value of the feature parameter in the fault determination logic, the threshold value of the feature parameter must be reset.

[0039] In a further preferred embodiment, the step of classifying the filtered result specifically includes: obtaining the state parameters based on the processing description; and classifying the elevator state parameters based on the elevator state parameters. Failure Data This includes determining:

[0040] Specifically, all Failure Data After the data is linked to the maintenance records based on the above, elevator fault discrimination is performed to distinguish between actual faults, human faults, and falsely reported faults.

[0041] In a further preferred embodiment, the elevator status parameter is Failure Data The step of determining specifically includes: performing a classification operation on the process description based on NLP; and classifying the state parameters based on the classification result, where the process description includes a fault situation, a fault expression, a process, and a process result.

[0042] Specifically, after handling an elevator malfunction, maintenance personnel must enter malfunction handling process information. This information is manually entered to explain the malfunction situation, description, handling process, and handling result. Based on the explanatory information in the maintenance record, an NLP (natural language analysis) classifier is trained to determine whether the malfunction corresponding to the maintenance record is attributable to the equipment itself or to a human being based on the cause description in the maintenance record. NLP (natural language analysis) is then used to identify, analyze, classify, and determine the type of malfunction for the maintenance record, thereby determining whether each record is accurate and valid (whether it is a human-caused malfunction).

[0043] In an embodiment of the present invention, a data acquisition unit for acquiring elevator operation data; a maintenance record obtaining unit for obtaining corresponding maintenance records according to the elevator operation data, the maintenance records including maintenance time, fault code and processing description; a dataset classification unit for performing a dataset classification operation on the maintenance records; a fault logic judgment unit for comparing the result of the data set classification and the fault judgment logic; a fault logic optimization unit for optimizing the fault determination logic based on the comparison result; An elevator fault judgment logic verification system is further provided, which includes: an execution unit for executing an elevator fault judgment operation based on the optimized result.

[0044] The present invention provides an elevator fault determination logic verification method, including at least one processor and at least one memory for storing at least one program, The present invention further provides an elevator fault determination logic verification system that, when the at least one program is executed by at least one processor, causes the at least one processor to realize the elevator fault determination logic verification method.

[0045] The content of the above method embodiments may all be applied to this system embodiment, and the functions specifically realized by this system embodiment are the same as those of the above method embodiments, and the beneficial effects achieved by this system embodiment are the same as those of the above method embodiments.

[0046] In addition, an embodiment of the present invention further provides a storage medium having processor-executable commands stored therein, the storage medium being used to execute the elevator fault determination logic verification method when the processor-executable commands are executed by a processor.

[0047] Compared with the prior art, the elevator fault determination logic verification method, system, and storage medium according to the present invention have the following advantages:

[0048] 1) The present invention uses the elevator's operating status and construction maintenance work to reverse-verify the accuracy of elevator faults, thereby estimating whether the fault judgment logic is appropriate and valid, and reducing erroneous and missed reports of elevator fault information.

[0049] 2) The method according to the present invention can realize real-time updating of fault determination logic, thereby reducing the workload of maintenance personnel and research and development personnel.

[0050] The step numbers in the above method examples are given for ease of explanation, and the order of the steps is not particularly limited, and the order of execution of each step in the examples may be adaptively adjusted based on the understanding of a person skilled in the art.

[0051] Although the preferred embodiments of the present invention have been described in detail above, the present invention is not limited to these embodiments. Those skilled in the art may make various equivalent modifications or substitutions without departing from the spirit of the present invention. All of these equivalent modifications or substitutions should be included within the scope of the claims.

Claims

1. Obtaining elevator operation data including status parameters, fault data, and out-of-service maintenance data, wherein the fault data includes a fault type, a fault time, an elevator type, and an elevator number, and the out-of-service maintenance data includes a maintenance processing time, an elevator number, and a fault code; obtaining an elevator maintenance record, including maintenance time, fault code, and treatment description, based on the elevator operation data; a step of performing a data set classification operation on the maintenance records, the step of performing a data set classification operation on the maintenance records including: filtering fault data and operational stop maintenance data in the elevator operation data based on the maintenance records; and classifying the filtered result, wherein the faults after the classification process include actual faults, falsely reported faults, and human-caused faults; comparing the results of the data set classification based on the fault determination logic; optimizing the fault determination logic based on the comparison result; and performing an elevator fault determination operation based on the optimized result; According to the maintenance records, the fault data and the operation stop maintenance data in the elevator operation data are filtered, and the filtered results are classified, and the faults after the classification process include actual faults, falsely reported faults, and human faults. loading the fault data and the corresponding out-of-operation maintenance data and the maintenance records, the fault data, the maintenance records and the out-of-operation maintenance data being associated with the elevator number and the fault code, determining the cause of causing the elevator fault or performing out-of-operation maintenance based on the fault code and the processing description in the maintenance records, and classifying the fault data into actual faults, falsely reported faults and human faults based on the cause; The fault determination logic is executed on a hardware terminal and determines whether a fault has occurred in the elevator based on parameters collected by sensors, and the fault determination logic is a logic and threshold realized based on research and development, construction experience, elevator specifications, usage environment, and elevator utilization rate; The step of comparing based on the result of the dataset classification and the fault determination logic includes: loading the operation data of the false-reported fault, establishing corresponding feature engineering to obtain corresponding feature values, and comparing the obtained feature values with parameter thresholds preset in the fault determination logic; and optimizing the elevator fault determination logic based on the comparison result includes modifying a threshold value of the fault determination logic, and if a specific value of a feature value in a false-reported fault is higher or lower than the preset parameter threshold value in the fault determination logic, the preset parameter threshold value needs to be reset.

2. at least one processor; at least one memory for storing at least one program, 2. An elevator fault determination logic verification system, wherein, when the at least one program is executed by the at least one processor, the at least one processor implements the elevator fault determination logic verification method according to claim 1.

3. A storage medium having processor-executable instructions stored thereon, 10. A storage medium used to perform the elevator fault determination logic verification method of claim 1 when processor-executable commands are executed by a processor.

Images