CN113071966A

CN113071966A - Elevator fault prediction method, device, equipment and storage medium

Info

Publication number: CN113071966A
Application number: CN202110451780.1A
Authority: CN
Inventors: 巴豪; 程纪华
Original assignee: Ping An International Smart City Technology Co Ltd
Current assignee: Ping An International Smart City Technology Co Ltd
Priority date: 2021-04-26
Filing date: 2021-04-26
Publication date: 2021-07-06

Abstract

The invention relates to the technical field of data processing, and discloses a method, a device, equipment and a storage medium for elevator fault prediction. Through the analysis to elevator trouble data, discern the risk factor of elevator, establish risk assessment system and risk assessment mathematical model, utilize the model to carry out real-time control to elevator operation data to realize prediction and control to the trouble, with the emergence probability that reduces the trouble, avoid the production of trouble even, improved the safety in utilization of elevator greatly, reduce the potential safety hazard of elevator, simultaneously, use the model to monitor, make the result precision that obtains can be higher. In addition, the invention also relates to a block chain technology, and the related information of elevator fault prediction can be stored in the block chain.

Description

Elevator fault prediction method, device, equipment and storage medium

Technical Field

The invention relates to the technical field of data processing, in particular to an elevator fault prediction method, device, equipment and storage medium.

Background

With the rapid development of society and the improvement of living standard and material culture of people, the elevator is not only a production device, but also an essential tool for people in work and life, and is a transportation tool frequently used by people like a transportation tool such as a bus, a subway and the like. The national requirements on the elevator safety supervision mode, safety technology evaluation, technology detection and emergency rescue technology are higher, however, the existing elevator risk evaluation standard is fuzzy, and the evaluation accuracy of each risk factor is poor.

At present, a plurality of methods for elevator risk assessment are available, but the existing methods are basically assessed according to the experience of experts, especially on the setting of risk index weights, a fault detection rule is obtained by learning fault detection results of the experts, and corresponding faults are matched by comparing fault expressions generated by the elevator, but such a method is not intelligent, the fault determination precision is low, meanwhile, real-time monitoring and advanced prejudgment of the faults cannot be realized, and the probability of fault occurrence is difficult to reduce.

Disclosure of Invention

The invention mainly aims to solve the technical problems that the existing elevator fault evaluation precision is low and real-time monitoring and prediction cannot be realized.

The invention provides an elevator fault prediction method, which is applied to an elevator control system and comprises the following steps: acquiring real-time operation data of an elevator; carrying out risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator; taking the at least two risk factors as query indexes, and querying a risk assessment system and a risk analysis strategy matched with the at least two risk factors from a risk assessment system relation table constructed in advance by a big data analysis technology; selecting a corresponding risk assessment mathematical model from a preset model library according to the risk assessment system; inputting the at least two risk factors into the risk assessment mathematical model to calculate a risk coefficient of the fault occurrence to obtain an assessment result; and analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

Optionally, in a first implementation manner of the first aspect of the present invention, the performing risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator includes: extracting time information and phase marking information in the real-time operation data, and determining the operation phase type of the elevator according to the time information and the phase marking information; determining a corresponding risk factor system from a preset risk factor library according to the operation stage type; and inputting the risk factor system and the real-time operation data into the risk factor mining model, extracting all risk factors in the real-time operation data by using the risk factor system as a reference through the risk factor mining model, and outputting at least two risk factors.

Optionally, in a second implementation manner of the first aspect of the present invention, the querying, with the at least two risk factors as query indexes, a risk assessment system and a risk analysis policy that are matched with the at least two risk factors from a risk assessment system relationship table that is constructed in advance by a big data analysis technology includes: screening out a corresponding relation record which takes the operation stage type as a main node from a pre-constructed risk assessment system relation table, wherein the corresponding relation record is a corresponding relation branch comprising risk factors, a risk assessment system and a risk analysis strategy; matching the corresponding relationship records item by taking the at least two risk factors as query indexes, and calculating the matching degree of each corresponding relationship record and the at least two risk factors; and screening out a corresponding relation record with the highest matching degree as a record corresponding to the at least two risk factors, and extracting a risk evaluation system and a risk analysis strategy in the record.

Optionally, in a third implementation manner of the first aspect of the present invention, the inputting the at least two risk factors into the risk assessment mathematical model to calculate a risk coefficient of a fault occurrence, and obtaining an assessment result includes: determining the weight corresponding to each risk factor in the at least two risk factors according to the operation stage type; calculating the danger level of the corresponding risk factor according to the weight; sequencing the at least two risk factors from high to low according to the risk levels to obtain a risk factor sequence; calculating a risk assessment value and a risk prediction result corresponding to each risk factor in the risk factor sequence by using a risk assessment mathematical model; and comprehensively calculating all risk evaluation values and risk prediction results to obtain evaluation results.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the risk analysis policy includes a failure prediction calculation rule and an expert risk evaluation rule, and analyzing the evaluation result according to the risk analysis policy to obtain the failure prediction result of the elevator includes: selecting the first N risk factors from the risk factor sequence as calculation factors, and calculating fault risk factors existing in the real-time operation data according to a fault probability calculation formula specified in the fault prediction calculation rule; mining and evaluating the fault by utilizing the expert risk evaluation rule according to the fault risk factor to obtain an artificial fault mining result; comparing the evaluation result with the artificial fault mining result to obtain a comparison result; and adjusting the evaluation result based on the comparison result to obtain a fault prediction result of the elevator.

Optionally, in a fifth implementation manner of the first aspect of the present invention, before the acquiring the real-time operation data of the elevator, the method further includes: obtaining elevator fault record data generated in different operation stages from a maintenance log table of the elevator control system; extracting fault features and fault labels in the elevator fault record data, and classifying the fault features by using a clustering algorithm to obtain risk factor sets generating faults in different operation stages; and establishing an incidence relation between the risk factor set fault and the fault label to obtain a risk evaluation system and a risk factor system.

Optionally, in a sixth implementation manner of the first aspect of the present invention, after the establishing an association relationship between the risk factor set fault and the fault label to obtain a risk assessment system and a risk factor system, the method further includes: analyzing risk factors and fault labels in the risk assessment system by using a neural network, and performing deep learning on the association relation to obtain a risk assessment mathematical model prototype; dividing elevator fault record data generated in different operation stages into a training set and a verification set, sequentially inputting the training set and the verification set into the risk assessment mathematical model prototype to perform training and verification cyclic operation, and obtaining the risk assessment mathematical model when a verification result meets a preset convergence value.

A second aspect of the present invention provides the elevator failure prediction apparatus, including: the acquisition module is used for acquiring real-time operation data of the elevator; the mining module is used for mining risk factors of the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator; the query module is used for taking the at least two risk factors as query indexes and querying a risk assessment system and a risk analysis strategy which are matched with the at least two risk factors from a risk assessment system relation table which is constructed in advance through a big data analysis technology; the selection module is used for selecting a corresponding risk assessment mathematical model from a preset model library according to the risk assessment system; the evaluation module is used for inputting the at least two risk factors into the risk evaluation mathematical model to calculate the risk coefficient of the fault occurrence so as to obtain an evaluation result; and the prediction module is used for analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

Optionally, in a first implementation manner of the second aspect of the present invention, the mining module includes: the first extraction unit is used for extracting time information and phase mark information in the real-time running data and determining the running phase type of the elevator according to the time information and the phase mark information; the determining unit is used for determining a corresponding risk factor system from a preset risk factor library according to the operation stage type; and the mining unit is used for inputting the risk factor system and the real-time operation data into the risk factor mining model, extracting all risk factors in the real-time operation data by using the risk factor mining model and taking the risk factor system as a reference, and outputting at least two risk factors.

Optionally, in a second implementation manner of the second aspect of the present invention, the query module includes: the screening unit is used for screening out a corresponding relation record which takes the operation stage type as a main node from a pre-constructed risk assessment system relation table, wherein the corresponding relation record is a corresponding relation branch comprising risk factors, a risk assessment system and a risk analysis strategy; the query unit is used for matching the corresponding relationship records one by taking the at least two risk factors as query indexes, and calculating the matching degree of each corresponding relationship record and the at least two risk factors; and the second extraction unit is used for screening out a corresponding relation record with the highest matching degree as a record corresponding to the at least two risk factors, and extracting a risk evaluation system and a risk analysis strategy in the record.

Optionally, in a third implementation manner of the second aspect of the present invention, the evaluation module includes: the weight calculation unit is used for determining the weight corresponding to each risk factor in the at least two risk factors according to the operation stage type; the grade calculation unit is used for calculating the danger grade of the corresponding risk factor according to the weight; the sorting unit is used for sorting the at least two risk factors from high to low according to the risk levels to obtain a risk factor sequence; the risk calculation unit is used for calculating a risk evaluation value and a risk prediction result corresponding to each risk factor in the risk factor sequence by using a risk evaluation mathematical model; and the evaluation unit is used for comprehensively calculating all risk evaluation values and risk prediction results to obtain evaluation results.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the risk calculating unit is specifically configured to: selecting the first N risk factors from the risk factor sequence as calculation factors, and calculating fault risk factors existing in the real-time operation data according to a fault probability calculation formula specified in the fault prediction calculation rule; mining and evaluating the fault by utilizing the expert risk evaluation rule according to the fault risk factor to obtain an artificial fault mining result; comparing the evaluation result with the artificial fault mining result to obtain a comparison result; and adjusting the evaluation result based on the comparison result to obtain a fault prediction result of the elevator.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the elevator fault prediction apparatus further includes an analysis module, where the analysis module is specifically configured to: obtaining elevator fault record data generated in different operation stages from a maintenance log table of the elevator control system; extracting fault features and fault labels in the elevator fault record data, and classifying the fault features by using a clustering algorithm to obtain risk factor sets generating faults in different operation stages; and establishing an incidence relation between the risk factor set fault and the fault label to obtain a risk evaluation system and a risk factor system.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the elevator failure prediction apparatus further includes a verification module, where the verification module is specifically configured to: analyzing risk factors and fault labels in the risk assessment system by using a neural network, and performing deep learning on the association relation to obtain a risk assessment mathematical model prototype; dividing elevator fault record data generated in different operation stages into a training set and a verification set, sequentially inputting the training set and the verification set into the risk assessment mathematical model prototype to perform training and verification cyclic operation, and obtaining the risk assessment mathematical model when a verification result meets a preset convergence value.

A third aspect of the present invention provides an elevator failure prediction apparatus comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the elevator failure prediction device to perform the elevator failure prediction method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute the above-described elevator failure prediction method.

According to the technical scheme, the risk factors of the elevator are identified through analyzing the elevator fault data, a risk evaluation system and a risk evaluation mathematical model are established, the model is used for monitoring the elevator operation data in real time, so that the fault is predicted and monitored, the fault occurrence probability is reduced, even the fault is avoided, the use safety of the elevator is greatly improved, the potential safety hazard of the elevator is reduced, and meanwhile, the model is used for monitoring, so that the obtained result precision is higher.

Drawings

Fig. 1 is a schematic diagram of a first embodiment of an elevator fault prediction method in an embodiment of the invention;

fig. 2 is a schematic diagram of a second embodiment of the elevator fault prediction method in the embodiment of the invention;

fig. 3 is a schematic diagram of a third embodiment of the elevator fault prediction method in the embodiment of the invention;

fig. 4 is a schematic diagram of a fourth embodiment of the elevator fault prediction method in the embodiment of the invention;

fig. 5 is a schematic view of an embodiment of an elevator failure prediction apparatus according to an embodiment of the present invention;

fig. 6 is a schematic view of another embodiment of the elevator failure prediction apparatus according to the embodiment of the present invention;

fig. 7 is a schematic diagram of an embodiment of an elevator failure prediction apparatus in an embodiment of the present invention.

Detailed Description

The embodiment of the invention provides an elevator fault prediction method, an elevator fault prediction device, elevator fault prediction equipment and a storage medium, wherein risk factors of an elevator are identified through analysis of elevator fault data, a risk evaluation system and a risk evaluation mathematical model are established, elevator operation data are monitored in real time by using the model, the elevator fault prediction performance is improved, units with safety risks are overhauled in time based on big data analysis and intelligent mining technology, event tracking and deep mining of reasons are carried out on the occurring safety faults, and the risks are controlled in the lowest range.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For the sake of understanding, the following describes a specific flow of an embodiment of the present invention, and referring to fig. 1, a first embodiment of an elevator fault prediction method in an embodiment of the present invention includes:

101, acquiring real-time operation data of an elevator;

in this step, the real-time operation data is stored in the elevator system, specifically, in a cache unit of the elevator system in the form of an operation log, and preferably, the real-time operation data may be operation data of a certain time node or operation data of a certain time period.

In the obtaining process, data change in the cache unit is monitored in real time in a specific mode of setting a data crawler tool, wherein before the data crawler tool is started, monitoring time nodes or time periods of the data crawler tool and the total amount of crawled data are configured, the data crawler tool monitors the recording time of the operation log, when the change of the recording time is monitored, the operation log is extracted when the interval between the recording time and the acquisition time of the last data crawler tool meets the configured time period, and then data generated in the operation of the elevator in the operation log is analyzed according to field information, wherein the data comprise circuit parameters in the operation of the elevator, carrying data in each operation of the elevator and change data of the circuit parameters of the elevator in the carrying process;

in this embodiment, the obtaining of the real-time operation data may be specifically realized by configuring a manner of reporting the elevator control system at a fixed time, preferably, a communication unit is set on a control terminal of the elevator, a timer is set based on the communication unit, when the count of the timer reaches, a data reporting mechanism of the communication unit is triggered, the communication unit retrieves corresponding operation data from a cache unit of the elevator, and compresses and packages the operation data based on a communication protocol of the communication unit, and then sends the operation data to an elevator fault prediction module or device, where the elevator fault prediction module or device may be an external big data analysis system or an APP for data analysis, or even an alarm program installed on the elevator control system, and the alarm program monitors the operation data of the elevator and predicts and analyzes faults in real time, even a failure prediction or monitoring function module can be developed on the control system of the elevator to realize monitoring acquisition and analysis of the operation data of the elevator.

In this step, no matter which way to achieve the acquisition of the real-time operation data, the implementation steps specifically include:

after data acquisition is triggered, a data acquisition request is generated according to current time information of the elevator and the operation stage type of the elevator, a data segment corresponding to the time information is read from a cache unit according to the data acquisition request, and screening is carried out according to the operation stage type based on the data segment, so that operation data matched with the current operation stage type is obtained.

Further, when the data needs to be sent out, the method further comprises the steps of cleaning the operation data according to a data sending protocol, uniformly packaging and compressing the operation data into a data packet with a fixed format, and finally sending the data packet to the opposite end for data analysis, wherein the cleaning of the data comprises the steps of removing redundant data in the operation data, or completing the operation data according to a complete data format, or even converting the data format.

102, mining risk factors of the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

in this embodiment, the risk factor mining model is a data analysis model generated by self-learning of the operation data of the known fault through the deep-learning neural network, specifically, an association relationship is established between the fault in the known operation data and the fault factor, even elevator operation stage information corresponding to the operation data is associated, the association relationship is learned by using the data analysis model established based on the neural network, and the fault factor is subjected to extended learning, so as to obtain the risk factor mining model.

The risk factor mining model is used for identifying and extracting the data characteristics in the real-time operation data, so as to obtain a data characteristic set, the risk factor mining model carries out free combination based on the data characteristic set to obtain a plurality of risk factors, then, matching and comparing the plurality of risk factors with the fault factors corresponding to each fault one by one to screen out the corresponding risk factors, wherein when matching and comparing the risk factors, because the risk factors are generated in a free combination mode, the obtained risk factors are not only generated aiming at a certain fault, when the risk factors are matched and compared, the matching results of a plurality of faults exist, and after matching, screening the operation stage where the fault corresponding to the matched risk factor is located according to the operation stage type of the elevator, and screening out the risk factor corresponding to the operation stage corresponding to the real-time operation data.

103, taking at least two risk factors as query indexes, and querying a risk assessment system and a risk analysis strategy matched with the at least two risk factors from a risk assessment system relation table constructed in advance by a big data analysis technology;

in this step, specifically, each risk factor is extracted to perform feature extraction, where the feature refers to a data keyword corresponding to the risk factor in the operating data, such as a corresponding code segment, and the code segments are used as query indexes to query from a relationship table.

In this embodiment, the following specific steps may be implemented:

screening out a corresponding relation record which takes the operation stage type as a main node from a pre-constructed risk assessment system relation table, wherein the corresponding relation record is a corresponding relation branch comprising risk factors, a risk assessment system and a risk analysis strategy;

matching the corresponding relationship records item by taking the at least two risk factors as query indexes, and calculating the matching degree of each corresponding relationship record and the at least two risk factors;

and screening out a corresponding relation record with the highest matching degree as a record corresponding to the at least two risk factors, and extracting a risk evaluation system and a risk analysis strategy in the record.

In practical application, the risk assessment system and the risk analysis strategy are obtained by analyzing elevator operation data of known faults, specifically a risk assessment system generated by using risk factors and related factors obtained by a big data analysis technology, and the risk analysis strategy is obtained by solving the faults through adopting technical means.

104, selecting a corresponding risk assessment mathematical model from a preset model library according to a risk assessment system;

in the step, the model base comprises a plurality of models for evaluating various faults, and in practical application, evaluation standards for faults in different operation stages are different, so that after the risk evaluation mathematical model is selected, the selected risk evaluation data model is screened again according to the operation stage type, or configuration parameters in the risk evaluation mathematical model are adjusted according to evaluation parameters corresponding to the operation stage type, and an identification model matched with real-time operation data is obtained.

105, inputting at least two risk factors into a risk assessment mathematical model to calculate a risk coefficient of fault occurrence to obtain an assessment result;

in this step, in the process of calculating the risk coefficient, the risk assessment data model specifically identifies the fault types corresponding to the at least two risk factors according to the risk assessment mathematical model, determines the corresponding assessment rules based on the fault types, calculates the degree of relationship between the at least two risk factors and the fault based on the assessment rules to obtain the risk coefficient, where the risk coefficient may be understood as an average value of the association coefficients between each risk factor and each fault, and selects a closest fault type from the relationship coefficients based on the average value to generate the assessment result.

And 106, analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

In this embodiment, the risk analysis policy may be understood as a preconfigured result check rule, the result check rule is used to check the evaluation result to obtain a final failure prediction result, specifically, the relationship degree corresponding to the evaluation result is compared according to a preset correlation coefficient threshold in the result check rule, and whether the evaluation result meets a preset deviation amplitude is determined based on the comparison result, so as to obtain the failure prediction result.

In practical application, the risk analysis policy may also be an industry inspection standard, where the standard is a probability percentage of various faults occurring in different operation stages, the top N faults are selected based on the percentage, whether the corresponding fault in the evaluation result is in the N faults or not is matched, or a proportion of all the faults included in the evaluation result in the N faults is included, and if the difference between the proportion and the industry standard reaches a certain value, the top several faults in the evaluation result in the N faults are selected as fault prediction results.

By the method, elevator fault prediction can be realized by utilizing the risk assessment data model, and possible elevator fault factors and elevator fault factors which are already caused are analyzed, so that elevator maintenance personnel can timely overhaul the unit with the fault or the unit with the fault, track the fault and carry out time tracking and deep excavation of the reason, the elevator risk is controlled to be minimum, safety of a user is guaranteed, and experience of the user is improved.

Referring to fig. 2, a second embodiment of the elevator fault prediction method according to the embodiment of the present invention includes:

201, acquiring real-time operation data of an elevator;

202, extracting time information and phase marking information in the real-time operation data, and determining the operation phase type of the elevator according to the time information and the phase marking information;

in practical application, when an elevator control system records actual operation data, the recorded data comprises marks of time and operation stages, the information forms fields, the real-time operation data is analyzed through a preset field information and a scheduling byte extraction algorithm after being acquired, the real-time operation data is a data file in a table form, firstly, the field information is used as an identification index to identify the field name of a header in the table, after the field name which is the same as the field information in the header is determined, the data content of a corresponding area under the field name is extracted through a character extraction algorithm, preferably, the data position information of an area corresponding to the field name is calculated under the position where the field name is located, the area range of data reading is determined based on the position information, and the character information in the area range is extracted through the character extraction algorithm, and splicing and combining the extracted character information to obtain time information and phase mark information, determining the operation phase of the elevator according to the time information, the delivery time information and the installation time information of the elevator, and comparing the operation phase with the operation phase type to determine the actual operation phase type of the elevator.

203, determining a corresponding risk factor system from a preset risk factor library according to the operation stage type;

in this step, a plurality of risk factor systems are stored in the risk factor library, and the risk factor systems are divided and stored according to different operation phases, and a corresponding risk factor system is queried from the risk factor library according to the operation phase type, and in practical application, the determined risk factor system includes risk factors (i.e., risk factors) corresponding to a plurality of faults.

204, inputting the risk factor system and the real-time operation data into a risk factor mining model, extracting all risk factors in the real-time operation data by using the risk factor mining model and taking the risk factor system as a reference, and outputting at least two risk factors;

in the step, when the risk factor mining model is used for mining the risk factors, the model is used for analyzing real-time operation data to obtain a plurality of data characteristics, the data characteristics are combined with each other according to a fault factor combination rule generated by a fault to obtain the risk factors, the risk factors are matched with the risk factors in a risk factor system to obtain a matching result, and the consistent risk factors are selected as the risk factors to be output based on the matching result.

In practical application, because the risk factor system actually stores code instances generated when some faults are extracted in advance based on fault cases, the risk factor mining model matches the code data with code segments corresponding to the recorded risk factors in the risk factor system after analyzing the code data in the real-time running data, so as to obtain the risk factors.

205, taking at least two risk factors as query indexes, and querying a risk assessment system and a risk analysis strategy matched with the at least two risk factors from a risk assessment system relation table constructed in advance by a big data analysis technology;

206, selecting a corresponding risk assessment mathematical model from a preset model library according to a risk assessment system;

207, inputting at least two risk factors into the risk assessment mathematical model to calculate the risk coefficient of the fault occurrence, and obtaining an assessment result;

the step can specifically realize risk assessment through information such as weight, risk level and the like in risk factors, and the specific realization comprises the following steps:

determining the weight corresponding to each risk factor in the at least two risk factors according to the operation stage type;

for the calculation of the weight of the risk factors, the risk factors can be scored specifically through a scoring model constructed based on the risk factors, when the scoring model scores the risk factors, firstly, the types of scoring faults in the scoring model are configured according to the types of operation phases, a scoring rule is determined based on the types, after the input risk factors are received, scoring is performed on different fault types to obtain a plurality of score values, Y with higher scores are selected from the plurality of score values to be used as weight output, and further, when the risk factor sequence is generated, the score values of different faults are respectively sequenced to obtain a plurality of risk factor sequences.

Calculating the danger level of the corresponding risk factor according to the weight;

in this embodiment, the danger level refers to a degree division of danger caused by a risk factor to elevator operation, and may also be understood as a degree of loss or injury brought to an elevator, where the greater the loss possibly caused by elevator risk, the higher the risk level is, for example, when the elevator is out of control, falls from high altitude, and a button on a certain floor of the elevator is damaged, and cannot be used on the floor, and at this time, the danger level of the elevator out of control is far higher than the damage of the button of the elevator;

sequencing the at least two risk factors from high to low according to the risk levels to obtain a risk factor sequence;

in this embodiment, the risk level for identifying the risk factors may be specifically determined by querying a level table, for example, dividing the risk levels of different faults according to the faults in the historical operating data in advance, then classifying the risk levels and the fault and fault factors to form a risk level table, querying the fault factors in the risk level table through the identified risk factors in the process of predicting the fault based on the risk level table, then determining the risk levels, after querying the risk levels of all the risk factors, sorting the risk levels according to a sorting manner from high to low, and generating a risk factor sequence.

In practical application, the sequence is obtained by sequencing according to a certain rule, the higher the danger level is, the more possible damage is, therefore, according to the level of the danger level, the risk factors which possibly cause danger are sequenced from high to low to form an elevator risk factor sequence.

Calculating a risk assessment value and a risk prediction result corresponding to each risk factor in the risk factor sequence by using a risk assessment mathematical model;

and comprehensively calculating all risk evaluation values and risk prediction results to obtain evaluation results.

And 208, analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

Through the implementation of the method and the improvement in the prior art, the performance of elevator fault prediction is further improved, based on the application of big data, according to the correction and optimization of risk factor weights in different use scenes and different time periods, a system analyzes a large amount of data to obtain a prediction result and the experience judgment of professional technicians, and then the results obtained by the two are compared and analyzed to obtain the final elevator risk assessment result. Therefore, elevator maintenance personnel can timely overhaul the unit with the fault or the unit with the fault, track the depth excavation of time tracking and reasons after the fault occurs, control the risk of the elevator to the minimum, improve the safety of the elevator, enable a user to use the elevator more securely, bring convenience for the user to go out, and greatly improve experience.

Referring to fig. 3, a third embodiment of the elevator fault prediction method according to the embodiment of the present invention includes:

301. acquiring real-time operation data of an elevator;

302. carrying out risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

303. screening out a corresponding relation record which takes the operation stage type as a main node from a pre-constructed risk assessment system relation table;

in this step, the corresponding relation is recorded as a corresponding relation branch including a risk factor, a risk assessment system and a risk analysis strategy;

304. taking at least two risk factors as query indexes, matching the corresponding relationship records one by one, and calculating the matching degree of each corresponding relationship record and at least two risk factors;

305. screening out a corresponding relation record with the highest matching degree as a record corresponding to at least two risk factors, and extracting a risk evaluation system and a risk analysis strategy in the record;

306. selecting a corresponding risk assessment mathematical model from a preset model library according to a risk assessment system;

in this embodiment, the risk assessment mathematical model is to analyze actual risk factor conditions of collected elevator operation data, learn a pre-formulated calculation formula meeting actual requirements through a deep learning method, wherein the deep learning method is similar to traversing each node thereof from another perspective, deeply mine and analyze risk data, continuously modify risk factors and risk factor weights thereof, substitute the modified risk factors and risk factor weights into a preset risk assessment mathematical model, obtain a risk prediction value through calculation of the risk assessment mathematical model, and predict an assessment result causing elevator risk.

In practical application, the evaluation result is actually a specific fault of an elevator, the specific fault of the elevator comprises a plurality of types, faults occurring in different operation stages can also be different, even the probability of the fault is different, even faults of different types of elevators can also be different, the setting of the risk factors of the elevator comprises 10 categories including basic information of the elevator, a door system, a car and docking system, an electrical and control system and the like, and specific risk factors can be shown in the following table when faults occur in the categories:

the information in the table can be embodied in a risk assessment system relation table, and the purpose of improving the assessment precision of a risk assessment mathematical model and the mining precision of a risk factor mining model on risk factors is achieved.

307. Inputting at least two risk factors into a risk assessment mathematical model to calculate a risk coefficient of a fault to obtain an assessment result;

308. selecting the first N risk factors from the risk factor sequence as calculation factors, and calculating fault risk factors existing in the real-time operation data according to a fault probability calculation formula specified in the fault prediction calculation rule;

309. mining and evaluating faults by using an expert risk evaluation rule according to the fault risk factors to obtain an artificial fault mining result;

310. comparing the evaluation result with the manual fault mining result to obtain a comparison result;

311. and adjusting the evaluation result based on the comparison result to obtain a fault prediction result of the elevator.

In practical application, the risk evaluation value of the elevator needs to be predicted before the risk evaluation result is determined, the risk evaluation model and the evaluation system are further contrastively analyzed on the basis of the existing evaluation result to obtain the evaluation result, the judgment result of professional technicians is obtained according to the working experience of the professional technicians on the elevator and the analysis of elevator risk factors, and the evaluation result calculated by the risk evaluation model and the risk evaluation system and the judgment result of the elevator professional are comprehensively contrastively analyzed to obtain the final prediction result of elevator faults;

in practical application, each professional has its own judgment standard, so that the prediction result obtained by such judgment is not comprehensive enough to the research direction and the living experience of the professional. Therefore, professional technical researchers with different research directions for researching elevator risk prediction can obtain more accurate prediction results and scientific results. For example, when researchers are studying elevator directions as well as elevator systems, they tend to overlook hardware problems when their emphasis is on software.

In this step, when performing risk analysis on the evaluation result, specifically, the evaluation result and the historical fault case are evaluated and matched to obtain a matching degree, if the matching degree is greater than a preset value, it is indicated that the risk evaluation is accurate, and a fault prediction result is calculated based on the evaluation result.

Further, when the evaluation result is a new risk factor or no corresponding case exists in the case, the corresponding risk factor and the corresponding expert analysis data are inquired from the evaluation database of the expert according to the evaluation result, the evaluation result is scored based on the expert analysis data, if the score is greater than a preset value, the evaluation result is accurate, and a fault prediction result is calculated based on the evaluation result.

Further, in order to improve the prediction accuracy, the method further includes: risk data of all aspects are collected regularly, and factors and factor weights of original risk assessment are continuously corrected through mining analysis of the data, so that the factors and weights of an assessment system are gradually improved.

Specific inspection data, accident data, supervision and spot check data and the like are selected, judgment is carried out by professional technicians and evaluation is carried out by using the system, results of the inspection data, the accident data and the supervision and spot check data are compared and analyzed, and effectiveness and practicability of the elevator risk evaluation mathematical model are verified.

A large amount of data generated in each link of elevator supervision is analyzed from the state, systemic risk factors influencing the safe operation of the elevator are determined, indexes are extracted from the systemic risk factors to construct elevator risk factors, and the overall risk of the elevator is comprehensively evaluated.

The elevator risk assessment method based on the large data is improved in the prior art, the elevator fault prediction performance is further improved, based on the application of the large data, according to the correction and optimization of the risk factor weight in different use scenes and different time periods, the system analyzes a large amount of data to obtain prediction results and the experience judgment of professional technicians, and then the results obtained by the two are compared and analyzed to obtain the final elevator risk assessment result. Therefore, elevator maintenance personnel can timely overhaul the unit with the fault or the unit with the fault, track the depth excavation of time tracking and reasons after the fault occurs, control the risk of the elevator to the minimum, improve the safety of the elevator, enable a user to use the elevator more securely, bring convenience for the user to go out, and greatly improve experience.

Referring to fig. 4, a fourth embodiment of the elevator fault prediction method according to the embodiment of the present invention includes:

401. obtaining elevator fault record data generated in different operation stages from a maintenance log table of an elevator control system;

402. extracting fault features and fault labels in elevator fault record data, and classifying the fault features by using a clustering algorithm to obtain a risk factor set generating faults in different operation stages;

403. establishing an incidence relation between a risk factor set fault and a fault label to obtain a risk evaluation system and a risk factor system;

404. analyzing risk factors and fault labels in a risk evaluation system by using a neural network, and performing deep learning on the association relation to obtain a risk evaluation mathematical model prototype;

405. dividing elevator fault record data generated in different operation stages into a training set and a verification set, sequentially inputting the training set and the verification set into a risk assessment mathematical model prototype for training and verifying cyclic operation, and obtaining a risk assessment mathematical model when a verification result meets a preset convergence value;

in practice, it is possible for maintenance logbook entities to be collected from various aspects of manufacturing installation, maintenance, inspection, use, supervised spot inspection, supervision complaints and accident handling by using elevators, for example: collecting inspection data of inspection institutions, fault data of maintenance companies, market supervision spot check data, public opinion/public complaints and elevator accident survey data, analyzing fault reasons in the data based on big data and artificial intelligence technology to obtain risk factors, and establishing a relationship between the risk factors and faults to obtain a risk factor system, wherein the risk factor system is shown in the table. And learning the data based on the table by using a deep learning algorithm to obtain a risk assessment mathematical model.

406. Acquiring real-time operation data of an elevator;

407. carrying out risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

408. at least two risk factors are used as query indexes, and a risk assessment system and a risk analysis strategy which are matched with the at least two risk factors are queried from a risk assessment system relation table which is constructed in advance through a big data analysis technology;

in this embodiment, the risk assessment system may be specifically implemented as follows:

collecting elevator risk factors of all aspects;

in the example, the elevator risk factor is obtained by the identification of the elevator risk factor by the system by using data collected by the elevator from various aspects such as manufacturing installation, maintenance, inspection, use, supervision, accident handling and the like.

Obtaining elevator risk factors through the identification of the risk factors;

in this example, the risk factors of the elevator are identified and analyzed, the risk factors are classified according to the risk which occurs only under different situations, and the risk factors of the elevator, such as sudden turn-off of an elevator lamp, are obtained.

According to the elevator risk factors, determining risk factor evaluation rules through a tag set;

in this embodiment, the label set determines the level of each risk factor by labeling the risk level, and determines the weight according to the level, where a higher risk level corresponds to a higher weight;

further, the risk factor evaluation rule is to calculate the risk coefficient of the elevator according to the risk level of the risk factor and the weight of the risk factor, and can also be summarized as a calculation formula.

And establishing an evaluation system according with elevator risk prediction according to the risk factor evaluation rule.

In this example, the elevator risk coefficient is calculated according to the risk level of the risk factor and the weight of the risk factor, and a risk prediction and evaluation system meeting the actual demand is established according to the calculation mode.

409. Selecting a corresponding risk assessment mathematical model from a preset model library according to a risk assessment system;

410. inputting at least two risk factors into a risk assessment mathematical model to calculate a risk coefficient of a fault to obtain an assessment result;

411. and analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

In the embodiment of the invention, historical data causing elevator risks and risk data possibly causing faults to the elevator are collected, the data are identified and analyzed, the weight of the elevator risk factor is determined by methods such as a fuzzy method and an analytic hierarchy process according to a constructed risk evaluation system and a constructed risk evaluation mathematical model, the predicted value of the elevator faults is calculated, and then the predicted value of the elevator faults calculated by analyzing big data according to artificial intelligence and the final predicted result of the elevator faults is evaluated according to the judgment of professional technicians. Such implementation mode makes elevator maintenance personal can in time give the unit that has the trouble or the unit that breaks down and overhaul to track the degree of depth excavation that has broken down and carry out time tracking and reason, reach minimumly with the risk control of elevator, improve the security of elevator, let the user use more relieved, bring the facility for user's trip, great promotion experience is felt.

With reference to fig. 5, an elevator failure prediction apparatus in an embodiment of the present invention includes:

the acquisition module 501 is used for acquiring real-time operation data of the elevator;

the mining module 502 is used for performing risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

the query module 503 is configured to query, by using the at least two risk factors as a query index, a risk assessment system and a risk analysis policy that are matched with the at least two risk factors from a risk assessment system relation table that is constructed in advance through a big data analysis technology;

a selecting module 504, configured to select a corresponding risk assessment mathematical model from a preset model library according to the risk assessment system;

the evaluation module 505 is configured to input the at least two risk factors into the risk evaluation mathematical model to calculate a risk coefficient of a fault occurrence, so as to obtain an evaluation result;

and the prediction module 506 is used for analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

According to the embodiment of the invention, the risk assessment data model is utilized to realize elevator fault prediction, and also analyze the possible elevator fault factors and the elevator fault factors, so that elevator maintenance personnel can timely overhaul the fault unit or the fault unit, track the fault unit and perform time tracking and deep excavation of reasons, the elevator risk is controlled to be minimum, the safety of a user is guaranteed, and the experience of the user is improved.

Referring to fig. 6, another embodiment of the device for predicting demand of users based on routing according to the embodiment of the present invention includes:

Optionally, the mining module 502 includes:

the first extraction unit 5021 is used for extracting time information and phase mark information in the real-time running data and determining the running phase type of the elevator according to the time information and the phase mark information;

a determining unit 5022, configured to determine a corresponding risk factor system from a preset risk factor library according to the operation phase type;

and the mining unit 5023 is used for inputting the risk factor system and the real-time operation data into the risk factor mining model, extracting all risk factors in the real-time operation data by using the risk factor system as a reference through the risk factor mining model, and outputting at least two risk factors.

Optionally, the query module 503 includes:

a screening unit 5031, configured to screen a correspondence record with the operation stage type as a master node from a pre-constructed risk assessment system relationship table, where the correspondence record is a correspondence branch including a risk factor, a risk assessment system, and a risk analysis policy;

a query unit 5032, configured to perform matching on the corresponding relationship records item by taking the at least two risk factors as query indexes, and calculate matching degrees between each corresponding relationship record and the at least two risk factors;

a second extracting unit 5033, configured to filter out a corresponding relationship record with the highest matching degree as a record corresponding to the at least two risk factors, and extract a risk assessment system and a risk analysis policy in the record.

Optionally, the evaluation module 505 includes:

a weight calculation unit 5051, configured to determine, according to the operation phase type, a weight corresponding to each of the at least two risk factors;

a level calculation unit 5052, configured to calculate a risk level of the corresponding risk factor according to the weight;

a sorting unit 5053, configured to sort the at least two risk factors in order from high to low according to the risk level, so as to obtain a risk factor sequence;

a risk calculation unit 5054, configured to calculate, by using a risk assessment mathematical model, a risk assessment value and a risk prediction result corresponding to each risk factor in the risk factor sequence;

the evaluation unit 5055 is configured to perform comprehensive calculation on all risk evaluation values and risk prediction results to obtain an evaluation result.

Optionally, the risk calculation unit 5054 is specifically configured to: selecting the first N risk factors from the risk factor sequence as calculation factors, and calculating fault risk factors existing in the real-time operation data according to a fault probability calculation formula specified in the fault prediction calculation rule; mining and evaluating the fault by utilizing the expert risk evaluation rule according to the fault risk factor to obtain an artificial fault mining result; comparing the evaluation result with the artificial fault mining result to obtain a comparison result; and adjusting the evaluation result based on the comparison result to obtain a fault prediction result of the elevator.

Optionally, the elevator fault prediction apparatus further includes an analysis module 507, where the analysis module 507 is specifically configured to: obtaining elevator fault record data generated in different operation stages from a maintenance log table of the elevator control system; extracting fault features and fault labels in the elevator fault record data, and classifying the fault features by using a clustering algorithm to obtain risk factor sets generating faults in different operation stages; and establishing an incidence relation between the risk factor set fault and the fault label to obtain a risk evaluation system and a risk factor system.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the elevator failure prediction apparatus further includes a verification module 508, where the verification module 508 is specifically configured to: analyzing risk factors and fault labels in the risk assessment system by using a neural network, and performing deep learning on the association relation to obtain a risk assessment mathematical model prototype; dividing elevator fault record data generated in different operation stages into a training set and a verification set, sequentially inputting the training set and the verification set into the risk assessment mathematical model prototype to perform training and verification cyclic operation, and obtaining the risk assessment mathematical model when a verification result meets a preset convergence value.

In the embodiment of the invention, the risk factors of the elevator are identified through analyzing the elevator fault data, the risk evaluation system and the risk evaluation mathematical model are established, and the model is used for monitoring the elevator operation data in real time so as to realize the prediction and monitoring of the fault, reduce the occurrence probability of the fault and even avoid the fault, greatly improve the use safety of the elevator and reduce the potential safety hazard of the elevator, and meanwhile, the model is used for monitoring so that the obtained result precision is higher.

Referring to fig. 7, an embodiment of an elevator failure prediction apparatus in an embodiment of the present invention will be described in detail below from the viewpoint of hardware processing.

Fig. 7 is a schematic structural diagram of an elevator failure prediction apparatus 700 according to an embodiment of the present invention, where the elevator failure prediction apparatus 700 may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 710 (e.g., one or more processors) and a memory 720, and one or more storage media 730 (e.g., one or more mass storage devices) storing an application 733 or data 732. Memory 720 and storage medium 730 may be, among other things, transient storage or persistent storage. The program stored on the storage medium 730 may include one or more modules (not shown), each of which may include a sequence of instructions operating on the route-based user demand prediction apparatus 700. Further, the processor 710 may be configured to communicate with the storage medium 730 to execute a series of instruction operations in the storage medium 730 on the elevator failure prediction apparatus 700.

The elevator failure prediction apparatus 700 may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input-output interfaces 760, and/or one or more operating systems 731, such as Windows service, Mac OS X, Unix, Linux, FreeBSD, and the like. Those skilled in the art will appreciate that the elevator failure prediction device configuration shown in fig. 7 does not constitute a limitation of the elevator failure prediction device and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A Block chain (Block chain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data Block contains a batch of information for verifying the validity (anti-counterfeiting) of the information and generating a next Block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, which may also be a volatile computer readable storage medium, having stored therein instructions, which, when run on a computer, cause the computer to perform the steps of the elevator failure prediction method.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. An elevator fault prediction method is applied to an elevator control system, and is characterized by comprising the following steps:

acquiring real-time operation data of an elevator;

carrying out risk factor mining processing on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

taking the at least two risk factors as query indexes, and querying a risk assessment system and a risk analysis strategy matched with the at least two risk factors from a risk assessment system relation table constructed in advance by a big data analysis technology;

selecting a corresponding risk assessment mathematical model from a preset model library according to the risk assessment system;

inputting the at least two risk factors into the risk assessment mathematical model to calculate a risk coefficient of the fault occurrence to obtain an assessment result;

and analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

2. The elevator fault prediction method according to claim 1, wherein the mining of the risk factors on the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator comprises:

extracting time information and phase marking information in the real-time operation data, and determining the operation phase type of the elevator according to the time information and the phase marking information;

determining a corresponding risk factor system from a preset risk factor library according to the operation stage type;

and inputting the risk factor system and the real-time operation data into the risk factor mining model, extracting all risk factors in the real-time operation data by using the risk factor system as a reference through the risk factor mining model, and outputting at least two risk factors.

3. The elevator fault prediction method according to claim 2, wherein the querying the risk assessment system and the risk analysis policy matching the at least two risk factors from the risk assessment system relationship table constructed in advance by the big data analysis technology with the at least two risk factors as query indexes comprises:

4. The elevator fault prediction method according to any one of claims 1-3, wherein the inputting the at least two risk factors into the risk assessment mathematical model for calculation of risk coefficients of fault occurrence, and obtaining the assessment result comprises:

5. The elevator fault prediction method according to claim 4, wherein the risk analysis strategy includes a fault prediction calculation rule and an expert risk evaluation rule, and the analyzing the evaluation result according to the risk analysis strategy to obtain the fault prediction result of the elevator includes:

selecting the first N risk factors from the risk factor sequence as calculation factors, and calculating fault risk factors existing in the real-time operation data according to a fault probability calculation formula specified in the fault prediction calculation rule;

mining and evaluating the fault by utilizing the expert risk evaluation rule according to the fault risk factor to obtain an artificial fault mining result;

comparing the evaluation result with the artificial fault mining result to obtain a comparison result;

and adjusting the evaluation result based on the comparison result to obtain a fault prediction result of the elevator.

6. The elevator fault prediction method of any of claims 1-3, further comprising, prior to the obtaining real-time operational data for an elevator:

obtaining elevator fault record data generated in different operation stages from a maintenance log table of the elevator control system;

extracting fault features and fault labels in the elevator fault record data, and classifying the fault features by using a clustering algorithm to obtain risk factor sets generating faults in different operation stages;

and establishing an incidence relation between the risk factor set fault and the fault label to obtain a risk evaluation system and a risk factor system.

7. The elevator failure prediction method of claim 6, further comprising, after the establishing the association between the risk factor set failure and the failure label to obtain a risk assessment system and a risk factor system:

analyzing risk factors and fault labels in the risk assessment system by using a neural network, and performing deep learning on the association relation to obtain a risk assessment mathematical model prototype;

dividing elevator fault record data generated in different operation stages into a training set and a verification set, sequentially inputting the training set and the verification set into the risk assessment mathematical model prototype to perform training and verification cyclic operation, and obtaining the risk assessment mathematical model when a verification result meets a preset convergence value.

8. An elevator failure prediction device, characterized by comprising:

the acquisition module is used for acquiring real-time operation data of the elevator;

the mining module is used for mining risk factors of the real-time operation data by using a preset risk factor mining model to obtain at least two risk factors of the elevator;

the query module is used for taking the at least two risk factors as query indexes and querying a risk assessment system and a risk analysis strategy which are matched with the at least two risk factors from a risk assessment system relation table which is constructed in advance through a big data analysis technology;

the selection module is used for selecting a corresponding risk assessment mathematical model from a preset model library according to the risk assessment system;

the evaluation module is used for inputting the at least two risk factors into the risk evaluation mathematical model to calculate the risk coefficient of the fault occurrence so as to obtain an evaluation result;

and the prediction module is used for analyzing the evaluation result according to the risk analysis strategy to obtain a fault prediction result of the elevator.

9. An elevator failure prediction apparatus, characterized by comprising: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the elevator failure prediction device to perform the elevator failure prediction method of any of claims 1-7.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the elevator fault prediction method according to any one of claims 1-7.