CN111960217A - Elevator steel wire rope fault detection method based on Internet of things - Google Patents
Elevator steel wire rope fault detection method based on Internet of things Download PDFInfo
- Publication number
- CN111960217A CN111960217A CN202010688519.9A CN202010688519A CN111960217A CN 111960217 A CN111960217 A CN 111960217A CN 202010688519 A CN202010688519 A CN 202010688519A CN 111960217 A CN111960217 A CN 111960217A
- Authority
- CN
- China
- Prior art keywords
- data
- elevator
- model
- acceleration data
- wire rope
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B7/00—Other common features of elevators
- B66B7/12—Checking, lubricating, or cleaning means for ropes, cables or guides
- B66B7/1207—Checking means
- B66B7/1215—Checking means specially adapted for ropes or cables
- B66B7/1223—Checking means specially adapted for ropes or cables by analysing electric variables
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C5/00—Measuring height; Measuring distances transverse to line of sight; Levelling between separated points; Surveyors' levels
- G01C5/06—Measuring height; Measuring distances transverse to line of sight; Levelling between separated points; Surveyors' levels by using barometric means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y10/00—Economic sectors
- G16Y10/35—Utilities, e.g. electricity, gas or water
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/10—Detection; Monitoring
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Radar, Positioning & Navigation (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Biology (AREA)
- Algebra (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Remote Sensing (AREA)
- Operations Research (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Quality & Reliability (AREA)
- Maintenance And Inspection Apparatuses For Elevators (AREA)
Abstract
The invention relates to an elevator steel wire rope fault detection method based on the Internet of things, which comprises the following steps: a. acquiring acceleration data of the elevator car in the running process, and uploading the acceleration data to a cloud server for grouping storage; b. preprocessing each group of the acceleration data; c. comparing the acceleration data with an abnormal characteristic model in a model database in a cloud server, and judging the fault of the steel wire rope when the acceleration data is matched with the abnormal characteristic model in the model database; d. and when the steel wire rope fails, determining the failure type of the steel wire rope according to the abnormal characteristic model. The detection method of the invention judges whether the steel wire rope fails by using the model in the online database, so the accuracy is higher than that of the method using the threshold value, and the applicability is wider.
Description
Technical Field
The invention relates to the technical field of fault detection, in particular to an elevator steel wire rope fault detection method based on the Internet of things.
Background
With the coming of a series of policies such as urbanization, old building transformation and the like, the number of urban elevators is increasing day by day, and the urban elevators bring convenience to people and are accompanied with some safety accidents. The safety and health degree of an elevator steel wire rope, which is one of main components for ensuring the safety of the elevator, is important for the safe traveling of the elevator.
In the detection mode of the steel wire rope in the prior art, the following defects exist, for example: high construction requirements or high costs for the equipment. Patent CN209536720U discloses a method for detecting elevator steel wire rope, in which the acceleration frequency is detected first, and then the tension deviation of the steel wire rope is obtained through a series of calculation, and finally the deviation is compared with the national standard to determine whether the steel wire rope is in fault. In the method, an acceleration sensor is adsorbed to an elevator steel wire rope through a magnet, and then the steel wire rope is knocked to obtain the acceleration frequency. Finally, the tension deviation of the steel wire rope is obtained by utilizing the frequency, and the tension deviation is compared with the national standard to obtain whether the steel wire rope is abnormal or not. Therefore, the detection mode is complex to install, and the acceleration frequency can be obtained only by knocking the steel wire rope, so that the steel wire rope is easy to be seriously damaged. And because of the influence of different regional environment, the fault occurrence form of each elevator wire rope is different. For example, in cold or hot areas, elevator ropes are prone to fracture due to contraction or expansion, and in wet areas, failures such as corrosion occur. Therefore, data of each application area is acquired in different ways only by using the offline threshold comparison mode, that is, the detection strategy cannot be adjusted according to the environmental influence factors of each area, so that the detection precision is low.
Disclosure of Invention
The invention aims to provide an elevator steel wire rope fault detection method capable of sharing a fault model.
In order to achieve the aim, the invention provides an elevator steel wire rope fault detection method based on the Internet of things, which comprises the following steps of:
a. acquiring acceleration data of the elevator car in the running process, and uploading the acceleration data to a cloud server for grouping storage;
b. preprocessing each group of the acceleration data;
c. comparing the acceleration data with an abnormal characteristic model in a model database in a cloud server, and judging the fault of the steel wire rope when the acceleration data is matched with the abnormal characteristic model in the model database;
d. and when the steel wire rope fails, determining the failure type of the steel wire rope according to the abnormal characteristic model.
According to one aspect of the present invention, in the step (a), the acceleration data is collected in real time by a gyroscope installed inside the elevator car;
the collected acceleration data are transmitted to a cloud server through a PLC (programmable logic controller), and are stored in a distributed storage system under a Hadoop big data architecture in a grouping mode according to the single operation of the elevator.
According to one aspect of the invention, the preprocessing in step (b) includes data washing, padding and filtering;
the data cleaning comprises the steps of removing data with large and small anomalies, a fixed value data set and a data set of a lost sequence in the acquisition process;
the filling is to expand data by using a numerical interpolation method aiming at the data acquired at low frequency, wherein the numerical interpolation method comprises secondary interpolation and mean value interpolation;
the filtering includes high-pass filtering, low-pass filtering, and least-squares filtering.
According to one aspect of the invention, the data of the abnormal large and the abnormal small are respectively data of which the maximum value is more than 1.5 times and data of which the minimum value is less than 0.5 times in the normal value range of the acceleration in the running process of the elevator car;
the data group of the loss sequence is a data group of which the data loss amount in the group of acceleration data exceeds 65 percent of the total number of the group of data;
the fixed value data set is a data set with an acceleration value as a constant during the single operation of the elevator, which is acquired;
and the acquisition frequency in the low-frequency acquisition is below 25 HZ.
According to one aspect of the present invention, in the step (c), the acceleration data is compared with the abnormal feature model in the model database in real time using a flink.
According to an aspect of the present invention, in the comparing in the step (c), if the similarity between the feature of the acceleration data and the feature in the abnormal feature model is higher than 85%, it is determined that the acceleration data and the feature in the abnormal feature model are matched.
According to one aspect of the present invention, there is further provided the step of (e) calculating a numerical characteristic of elevator operation when the wire rope is failed, and comparing the numerical characteristic with a predetermined safety threshold;
and if the numerical value characteristic is greater than a safety threshold value, issuing an alarm signal.
According to one aspect of the invention, the security threshold is at least 1.2 times the numerical feature in the feature database of the cloud server.
According to one aspect of the invention, the numerical characteristics are the cumulative operating mileage of the elevator between adjacent faults and the cumulative number of bends made in the wire rope.
According to one aspect of the invention, the step of calculating the accumulated operating mileage comprises the steps of performing double integration on acceleration data of single operation of the elevator to obtain the single operating mileage of the elevator, and then accumulating the single operating mileage to obtain the accumulated operating mileage, wherein the calculation formula is as follows:
wherein, a (t) is acceleration, t0 is operation starting time, t1 is operation ending time, v is speed, and s is mileage displacement;
the step of calculating the accumulated bending times comprises the steps of firstly calculating the single bending times of the elevator steel wire rope, and then accumulating the single bending times to obtain the accumulated bending times;
the single bending times are at least twice of the times of passing each floor by the elevator, and the times of passing each floor by the elevator are monitored by a barometer arranged in the elevator car;
the calculation formula of the accumulated bending times is as follows:
H＝44300*(1-(P/P0)^(1/5.256))；
wherein H is the altitude and P0 is the atmospheric pressure (0 ℃, 101.325 kPa).
According to one aspect of the invention, the abnormal feature model is generated by training, wherein the training step is that wavelet analysis is firstly carried out on each group of acceleration data to obtain the frequency feature of each group of acceleration data;
and analyzing the fault type corresponding to the group of acceleration data by using the frequency characteristics to obtain an abnormal characteristic model reflecting the mapping relation between the two and storing the abnormal characteristic model into the model database.
According to one aspect of the invention, when the acceleration data is not matched with the abnormal feature model, the acceleration data is compared with the acceleration data when the elevator normally operates, if the similarity of the characteristics of the acceleration data and the abnormal feature model is lower than 60%, the steel wire rope is judged to have a new fault except the fault in the model database at the moment, the abnormal feature model is trained by using the acceleration data, and the trained abnormal feature model is stored in the model database in the cloud server.
According to one aspect of the invention, an abnormal feature model is trained using historical data of elevator operation and then published to an online database. Training of the abnormal feature model can be implemented on line, for example, a model base is established for the first time; the method can also be implemented in a cloud server, for example, when a new steel wire rope fault type is acquired in real time. Therefore, when the online real-time detection is carried out, the acquired data can be compared with the online model, and the fault can be determined if the acquired data is matched with the online model. Therefore, the local computing cost can be reduced, the unlimited storage is realized, the fault type can be updated in real time, the reliability of data is improved, and the universality is stronger. According to the present invention, when the similarity of the feature of the acceleration data and the feature in the abnormal feature model is higher than 85%, it is a match. The models are uploaded to the online, so that the models of all implementation places of the method can be collected, namely the models in the database are shared, compared with a simple threshold judgment mode, the method can be used for detecting in real time, is good in timeliness, more comprehensive in judgment and more accurate and rapid in detection, and can be suitable for various use scenes of the elevator. And if the acceleration data at the current moment is not matched with the abnormal characteristic model, comparing the acceleration data with the acceleration data in the normal running state, and if the similarity of the characteristics of the acceleration data and the abnormal characteristic model is lower than 60%, judging that the fault is a new fault. And training a new model by using the model training device and storing the new model into a model database, thereby continuously perfecting the model database to improve the detection accuracy.
According to one aspect of the invention, the numerical characteristics of the fault rooms adjacent to each other in every two times in the history of the elevator are calculated by using the operation history data of the elevator and are uploaded to the characteristic database on the line. And after detecting and finding the fault on line in real time, further calculating the numerical characteristic between the fault and the last fault, if the numerical characteristic is at least 1.2 times larger than the numerical characteristic in the on-line database, sending an alarm signal, and if not, not sending the alarm signal. Thus, frequent maintenance can be avoided, and the cost on manpower and time can be saved. The numerical characteristics comprise the accumulated running mileage of the elevator and the accumulated bending times of the steel wire rope, and the accumulated running mileage and the accumulated bending times of the steel wire rope are closely related to the service life of the steel wire rope, so that whether the steel wire rope needs to be maintained at the moment can be effectively judged.
According to one scheme of the invention, the acceleration data of the elevator car is detected by using the gyroscope installed in the car, so that parts needing to be installed on the steel wire rope can be avoided, the damage to the steel wire rope is reduced to the greatest extent, the construction difficulty is reduced, and the safety coefficient is improved. And the measured acceleration data can be used as the basis for subsequent off-line training and real-time detection. The altitude height of the elevator implementation is detected through the barometer, so that the times of passing each floor of the elevator car can be counted, and the accumulated bending times of the steel wire rope can be obtained. The altitude is an objective factor reflecting the floor where the elevator car is located, and the elevator can be used no matter where the elevator car is located, and when the altitude and air pressure difference of different application areas is large, the barometer can be correspondingly adjusted or replaced. Therefore, the mode of using the air pressure as the judgment floor is not only accurate, but also has wide adaptability, low cost and easy popularization.
According to one aspect of the invention, the collected acceleration data is preprocessed before anomaly detection. The preprocessing process comprises data cleaning, data filling of low-frequency acquisition and filtering. The subsequent abnormal detection can be carried out, so that the data is not interfered by the dirty data, the data quantity is sufficient, and the data effectiveness is high. The data of the large abnormality and the small abnormality are respectively the data of the maximum value of the acceleration more than 1.5 times and the data of the minimum value less than 0.5 times when the elevator car works normally. The acquisition frequency in low-frequency acquisition is below 25HZ, and the limitation of the frequency range can ensure that the filled data can still maintain higher detection precision in subsequent detection. The numerical range can effectively eliminate data which influence the subsequent detection accuracy. The data group of the loss sequence is a data group in which the data loss amount in the set of acceleration data exceeds 65% of the total number of the data groups. Therefore, the data groups do not need to be analyzed in the subsequent big data analysis process, so that the calculation amount is reduced, and meanwhile, the influence of invalid data on the detection accuracy is further avoided.
Drawings
Fig. 1 is a flow chart schematically representing a detection method according to an embodiment of the present invention.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
The present invention is described in detail below with reference to the drawings and the specific embodiments, which are not repeated herein, but the embodiments of the present invention are not limited to the following embodiments.
The equipment used by the elevator steel wire rope fault detection method based on the Internet of things comprises a data acquisition module, a data preprocessing module and an abnormality detection module. The method utilizes the main elevator operation parameter as the acceleration of the elevator car when detecting, so that the data acquisition module comprises a gyroscope arranged in the elevator car and is used for acquiring the acceleration data of the elevator car in real time. The acceleration detection mode does not need to install components such as sensors on the steel wire rope, so that the construction difficulty is reduced. Referring to fig. 1, in the method of the present invention, first, the gyroscope is used to collect acceleration data in real time. The acceleration data of a single point acquired by the front-end gyroscope is transmitted to a cloud server (on the same line) in the Internet of things through a PLC technology, and is stored in a distributed storage system under a Hadoop big data framework in a group mode according to single operation of the elevator. The purpose of the grouped storage is that the data collected by the gyroscope are single-point data, and in order to analyze the frequency characteristics of the acceleration signals, the data need to be stored according to data groups so as to meet the subsequent big data analysis condition. According to the operation principle of the elevator, the acceleration value can be changed within a certain range when the elevator car is started or stopped every time, and the change rule of the acceleration value during the speed increasing or reducing is unchanged under the condition that a steel wire rope is not defective. That is, during the single operation of the elevator, the acceleration values corresponding to different time points are different and the change of the acceleration values needs to be within a certain range. Therefore, if the data acquired by the gyroscope are fixed values or abnormally large and small jump values, the data represent that the group of data is abnormal during acquisition, which has no practical significance for detecting the condition of the steel wire rope and may influence the accuracy of subsequent large data analysis, that is, the data can be summarized into dirty data. After the data preprocessing module extracts the acceleration data from the distributed storage system, the acceleration data is cleaned, so that the dirty data is removed. The step of data cleaning includes removing noise data with large and small anomalies, removing fixed value data groups which are constant, and also removing data groups of missing sequences in the signal acquisition process. When removing the abnormal large and small data, the removed data is the data with the numerical value higher than a certain multiple of the maximum value and the minimum value in the normal acceleration value range when the elevator car normally operates. Specifically, the abnormal data is data with the maximum value within the acceleration normal value range being more than 1.5 times; and the data with small abnormality is the data with the minimum value within the normal value range of the acceleration being less than 0.5 times. By setting the multiple, most abnormal data can be removed in the step, and the normal data can be prevented from being removed by mistake, so that the detection accuracy is improved. Missing sequences refer to data that was not acquired during the acquisition process. For example, in a set of collected data, when the data that is not collected or lost exceeds 65% of the total collection number, the set of data is considered invalid and should be discarded. And if the data loss amount in the group of data is smaller than the interval, the data is reserved, so that the data amount is prevented from being insufficient during the subsequent big data analysis. After the cleaning steps, the interference of the dirty data on the subsequent analysis can be reduced. In addition, the data preprocessing module of the present invention also fills and filters the data in the preprocessing step. The filling work may also be referred to as interpolation, i.e. the signal is augmented with numerical interpolation for the low frequency acquired data, wherein numerical interpolation includes quadratic interpolation and mean interpolation. The interpolation work is mainly to avoid the result of under-analysis caused by insufficient data volume. The acquisition frequency during low-frequency acquisition is below 25HZ (i.e. acquiring acceleration data at intervals of 40ms or longer), and this frequency range is limited because interpolation is not needed when the frequency is too high, and the data set after interpolation is too different from the true value when the frequency is too low, thereby affecting the subsequent detection. Therefore, the frequency range can ensure that the filled data can still maintain higher detection precision during subsequent detection. The filtering process is also called filtering smoothing, high-pass filtering, low-pass filtering, least square filtering and the like, and the filtering process can improve the effectiveness of subsequent big data analysis.
After the preprocessing process, the data collected by the front-end gyroscope already has the condition of big data analysis, and then the abnormality detection can be carried out. As can be seen from fig. 1, the operation of the anomaly detection module is divided into model training and on-line detection. Firstly, in the embodiment, the abnormal characteristic model is obtained by offline training and training by using historical operation data of the elevator, and the model is sent to an online model database after the training is finished for comparison in online detection. From this step it is known that after the initial installation of the apparatus for implementing the method on the elevator, it is possible to perform a training for a certain period of time before starting normal use in order to obtain a sufficient number of models for subsequent use. Of course, the initial training process may be idle, which is the most safe. But it is also possible to train in normal operation of the elevator for some elevators already having a maintenance plan with sufficient safety. The above is the training process when the method is initially used, but since the technology of the steel wire rope is continuously updated, new faults may occur in the future when the steel wire rope is normally used. The abnormal characteristic models in the model database need to be continuously updated to adapt to the long-term use of the elevator. According to this concept, the acceleration data during these faults should be data that does not match all of the abnormal feature models in the model database. Therefore, in order to avoid the confusion with the data in normal operation, the data which is not matched with the model is compared with the data in normal operation, and if the similarity between the data and the data in normal operation is not enough, namely the current time is neither normal operation without fault nor original fault in the database, the steel wire rope at the time corresponding to the acceleration data is determined to have a new fault. These data can then be used as the basis for training the model to refine the anomaly feature model in the model database. Therefore, the mode of updating the model in the model database can effectively improve the detection accuracy of the method. Therefore, model training is closely related to on-line detection, the model trained by the model is applied when the model is detected on line, and data detected in real time during subsequent normal use can be used for training the model. The mode can continuously update the model database to adapt to the elevator in each service environment, and can continuously improve the detection accuracy while using new fault types appearing in the future. As can be seen from the above, the step of model training can be performed either on-line or on-line.
Because the model database is an online database, model training only needs to train the model and upload the model to the database. Because the preprocessed acceleration data need to be subjected to wavelet transformation to obtain the frequency of the acceleration data, the wavelet analysis is firstly carried out on the preprocessed acceleration data in the model training process, so that the frequency characteristic of the acceleration data can be obtained. The frequency characteristics of the elevator in normal operation should be substantially the same, and a failure of the wire rope causes a change in the acceleration frequency characteristics. Therefore, the fault type corresponding to each frequency feature needs to be analyzed, and further the fault type corresponding to the acceleration data is obtained. In fact, wavelet analysis can analyze multidimensional characteristics of acceleration data, and frequency characteristics are only one of them. Therefore, the abnormal feature model can also represent the mapping relation between the multi-dimensional features and the fault types. In fact, this step can also be understood as a training process of the above model, that is, a model capable of reflecting the mapping relationship between the acceleration data and the fault type is obtained through training, and the model is the abnormal feature model of the present invention. Therefore, the mode of the abnormal acceleration data, namely the abnormal characteristic model, stored in the model database is actually the mode of the abnormal acceleration data, and the characteristic of the real-time acceleration data is utilized to be compared with the characteristic in the model in the real-time comparison.
Through the model training process, an abnormal characteristic model is obtained, and the on-line detection process is realized by closely depending on the model. In the invention, the acceleration data after pretreatment is analyzed in real time by on-line detection by using flink, the analysis step is to compare the acceleration data with the abnormal characteristic model in real time, and if the acceleration data is matched with the abnormal characteristic model in the model database, the fault is determined. The abnormal characteristic model represents the mapping relation between the characteristics of the acceleration data and the fault type, so that the flink real-time analysis is actually to query the mapping relation in the model database, and if the similarity between the characteristics of the data at the moment and the characteristics in the abnormal characteristic model in the model database is higher than 85%, the data are judged to be matched, namely the steel wire rope is suspected to have a fault at the moment. And if the characteristics of the real-time acceleration data and the characteristics of the abnormal characteristic models in the model database are higher than 85%, judging that the steel wire rope at the moment has various faults. If the acceleration data is not matched with the abnormal characteristic model, the characteristics of the acceleration data are compared with the characteristics of the acceleration data in normal operation, and the comparison mode still adopts a similarity mode. That is, when the similarity of the characteristics of the two is lower than 60%, it is determined that a new failure has occurred in the wire rope at that time. The acceleration data at this time should be used to train a new model, thereby continuously perfecting the model database. The subsequent processing steps when it is determined that a new failure has occurred are the same as the processing steps when it is determined that a failure has occurred.
To this end, online detection discovers the occurrence of a fault and determines the type of fault by comparing the real-time data to the models in the model database. But also to determine whether an alarm signal needs to be issued. The elevator has maintenance plan, and the alarm signal is used for informing maintenance personnel when the elevator needs maintenance. Therefore, the alarm signal should be sent according to a set maintenance schedule, otherwise, if the alarm signal is sent when a fault is just found, the maintenance process is more frequent, and waste is caused.
In the invention, whether an alarm is needed when a fault occurs is judged according to the numerical characteristics of the running of the elevator. Since the method mainly monitors the fault of the steel wire rope, the numerical characteristics of the invention comprise the accumulated running mileage of the elevator and the accumulated bending times of the steel wire rope. At this time, similar to the comparison method of the abnormal operation model, the numerical characteristics are obtained by using the historical data through offline (or online) calculation, and then transmitted to the online characteristic database for online detection. In the calculation process, the numerical characteristic required to be calculated is the numerical characteristic between every two adjacent faults in the history of the elevator, and the interval between the two adjacent faults is the time interval from the end of the previous fault to the start of the next fault (the end of the fault means that the fault is maintained). The method for calculating the accumulated running mileage adopts a mode of firstly calculating the single running mileage of the elevator and then accumulating according to the total running times. The single-run mileage refers to the stroke or displacement of each run of the elevator. Because the parameters acquired by the data acquisition module are the car acceleration, double integration can be performed by using the acceleration data so as to obtain the single-time running mileage of the elevator. The method does not need to use parts such as a mileage wheel and the like which need to be installed on the steel wire rope, not only has low construction difficulty, but also can reduce the damage to the steel wire rope. The specific calculation formula of the running mileage of the cumulative operation is as follows:
where a (t) is acceleration, t0 is operation start time, t1 is operation end time, v is velocity, and s is mileage displacement.
The calculation of the accumulated bending times in the invention also adopts a mode of firstly calculating the single bending times of the steel wire rope and then accumulating. In this embodiment, the number of single bends is defined as twice the number of times that the elevator passes through each floor, i.e., arrival at and departure from the floor is considered to pass through the floor once, and is also referred to as bending the wire rope twice. This step can also be understood as counting how many floors the elevator passes through in total between two faults, and then multiplying the number by 2 to obtain the accumulated number of times of bending. Of course, according to this concept, the number of times of bending of the wire rope is related to the number of pulleys, and if the number of elevator pulleys to be used is large, the calculation magnification of the number of times of bending of each time can be correspondingly increased. In the invention, the floor where the elevator car is located is monitored by utilizing the altitude and air pressure mode, and the work of monitoring the air pressure can be completed by utilizing the barometer. The air pressure collected by the barometer is also a parameter of the elevator, so the barometer is also a component in the data collection module, and meanwhile, the barometer is also installed in the elevator car. Because the altitude barometric pressure corresponding to each floor is fixed, the number of times of passing through each floor can be counted by counting the altitude barometric pressure collected by the barometer. The barometer can be a component integrating a statistical function, and also can only have a collecting function, and other equipment with the statistical function completes statistical work. The transmission mode of the air pressure data acquired by the front-end barometer is the same as that of the acceleration data, and the air pressure data and the acceleration data are finally stored in the distributed storage system. The specific calculation formula of the accumulated bending times is as follows:
H＝44300*(1-(P/P0)^(1/5.256))；
wherein H is the altitude and P0 is the atmospheric pressure (0 ℃, 101.325 kPa).
Therefore, the calculated accumulated running mileage and the accumulated bending times are both released to an online characteristic database. At this time, after the result of the wire rope fault is obtained by utilizing flink real-time analysis, the numerical characteristics between the fault distance of this time and the fault of the last time need to be further calculated. The alarm signal can only be issued if the value characteristic is greater than a safety threshold, otherwise, the detection is continued until the value characteristic is greater than the safety threshold. In the invention, the safety threshold is at least 1.2 times of the numerical characteristic in the characteristic database, the multiple corresponds to the same accumulated running mileage and the same accumulated bending times, but different fault types need to be in one-to-one correspondence during comparison. If the feature database contains a plurality of numerical features, the multiple corresponds to the numerical value in which the accumulated operating mileage or the accumulated bending times is respectively the maximum. Therefore, the invention utilizes the Internet of things equipment to connect the front-end acquisition module and the mobile phone client of the maintenance personnel, thereby providing effective information for the maintenance and other related personnel.
The above description is only one embodiment of the present invention, and is not intended to limit the present invention, and it is apparent to those skilled in the art that various modifications and variations can be made in the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (12)
1. An elevator steel wire rope fault detection method based on the Internet of things comprises the following steps:
a. acquiring acceleration data of the elevator car in the running process, and uploading the acceleration data to a cloud server for grouping storage;
b. preprocessing each group of the acceleration data;
c. comparing the acceleration data with an abnormal characteristic model in a model database in a cloud server, and judging the fault of the steel wire rope when the acceleration data is matched with the abnormal characteristic model in the model database;
d. and when the steel wire rope fails, determining the failure type of the steel wire rope according to the abnormal characteristic model.
2. The detecting method according to claim 1, wherein in the step (a), the acceleration data is collected in real time by a gyroscope installed inside an elevator car;
the collected acceleration data are transmitted to a cloud server through a PLC (programmable logic controller), and are stored in a distributed storage system under a Hadoop big data architecture in a grouping mode according to the single operation of the elevator.
3. The detection method according to claim 1, wherein the preprocessing in the step (b) comprises data washing, padding and filtering;
the data cleaning comprises the steps of removing data with large and small anomalies, a fixed value data set and a data set of a lost sequence in the acquisition process;
the filling is to expand data by using a numerical interpolation method aiming at the data acquired at low frequency, wherein the numerical interpolation method comprises secondary interpolation and mean value interpolation;
the filtering includes high-pass filtering, low-pass filtering, and least-squares filtering.
4. The detection method according to claim 3, wherein the data of the abnormal large and the abnormal small are respectively data of 1.5 times or more of the maximum value and 0.5 times or less of the minimum value in the normal value range of the acceleration during the operation of the elevator car;
the data group of the loss sequence is a data group of which the data loss amount in the group of acceleration data exceeds 65 percent of the total number of the group of data;
the fixed value data set is a data set with an acceleration value as a constant during the single operation of the elevator, which is acquired;
and the acquisition frequency in the low-frequency acquisition is below 25 HZ.
5. The detection method according to claim 1, wherein in the step (c), the acceleration data is compared with an abnormal feature model in the model database in real time by using flink.
6. The detection method according to claim 5, wherein in the comparison in step (c), if the similarity between the feature of the acceleration data and the feature in the abnormal feature model is higher than 85%, it is determined that the acceleration data and the abnormal feature model are matched.
7. The detection method according to claim 5 or 6, further comprising the steps of (e) calculating a numerical characteristic of elevator operation when the wire rope is failed, and comparing the numerical characteristic with a predetermined safety threshold;
and if the numerical value characteristic is greater than a safety threshold value, issuing an alarm signal.
8. The detection method of claim 7, wherein the security threshold is at least 1.2 times a numerical feature in a feature database of the cloud server.
9. The detection method according to claim 7, wherein the numerical characteristics are the accumulated operating mileage of the elevator between the adjacent faults and the accumulated bending times of the wire rope.
10. The detection method according to claim 9, wherein the step of calculating the accumulated operating mileage comprises the steps of performing double integration on acceleration data of single operation of the elevator to obtain the single operating mileage of the elevator, and then accumulating the single operating mileage to obtain the accumulated operating mileage, wherein the calculation formula is as follows:
wherein, a (t) is acceleration, t0 is operation starting time, t1 is operation ending time, v is speed, and s is mileage displacement;
the step of calculating the accumulated bending times comprises the steps of firstly calculating the single bending times of the elevator steel wire rope, and then accumulating the single bending times to obtain the accumulated bending times;
the single bending times are at least twice of the times of passing each floor by the elevator, and the times of passing each floor by the elevator are monitored by a barometer arranged in the elevator car;
the calculation formula of the accumulated bending times is as follows:
H＝44300*(1-(P/P0)^(1/5.256))；
wherein H is the altitude and P0 is the atmospheric pressure (0 ℃, 101.325 kPa).
11. The detection method according to claim 1, wherein the abnormal feature model is generated by training, and the training step is that wavelet analysis is firstly carried out on each group of acceleration data to obtain the frequency feature of each group of acceleration data;
and analyzing the fault type corresponding to the group of acceleration data by using the frequency characteristics to obtain an abnormal characteristic model reflecting the mapping relation between the two and storing the abnormal characteristic model into the model database.
12. The detection method according to claim 6, wherein when the acceleration data is not matched with the abnormal feature model, the acceleration data is compared with acceleration data of an elevator in normal operation, if the similarity of the characteristics of the acceleration data and the abnormal feature model is lower than 60%, it is determined that a new fault except the fault in the model database occurs in the steel wire rope at the time of the comparison, the abnormal feature model is trained by using the acceleration data corresponding to the new fault, and the trained abnormal feature model is stored in the model database in the cloud server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010688519.9A CN111960217B (en) | 2020-07-16 | 2020-07-16 | Elevator steel wire rope fault detection method based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010688519.9A CN111960217B (en) | 2020-07-16 | 2020-07-16 | Elevator steel wire rope fault detection method based on Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111960217A true CN111960217A (en) | 2020-11-20 |
| CN111960217B CN111960217B (en) | 2022-04-19 |
Family
ID=73362058
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010688519.9A Active CN111960217B (en) | 2020-07-16 | 2020-07-16 | Elevator steel wire rope fault detection method based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111960217B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0849208A1 (en) * | 1996-12-20 | 1998-06-24 | Otis Elevator Company | Method and Apparatus to Inspect Hoisting Ropes |
| NO20005336D0 (en) * | 1999-10-21 | 2000-10-23 | Inventio Ag | Rope device and for this suitable artificial fiber rope and its use |
| CN1397797A (en) * | 2001-07-12 | 2003-02-19 | 因温特奥股份公司 | Detector for detecting wear of bearing rope |
| CN103364289A (en) * | 2013-07-22 | 2013-10-23 | 宁波市鄞州隆茂冲压件厂 | Testing device of repeated bending stress of metal spring clip |
| CN104679834A (en) * | 2015-02-06 | 2015-06-03 | 北京工商大学 | Time series data cleaning method and system |
| CN106629330A (en) * | 2017-01-21 | 2017-05-10 | 上海峰景移动科技有限公司 | On-line detection system for tension of steel wire rope for elevator |
| CN107892223A (en) * | 2017-11-15 | 2018-04-10 | 中南大学 | A kind of hoisting container monitoring system and tension force fault monitoring method |
| CN108313846A (en) * | 2017-12-13 | 2018-07-24 | 浙江新再灵科技股份有限公司 | A kind of elevator platform velocity-measuring system and method based on gyroscope |
| CN108726299A (en) * | 2018-06-22 | 2018-11-02 | 广西烽火信息技术有限公司 | A method of based on air pressure change monitoring elevator running position |
| CN110386530A (en) * | 2019-07-16 | 2019-10-29 | 浙江大学 | A kind of elevator monitoring systems and method towards fault diagnosis and safe early warning |
| CN111302173A (en) * | 2019-11-25 | 2020-06-19 | 广东马上到网络科技有限公司 | Elevator steel wire rope abnormity detection method, device and equipment and readable storage medium |
-
2020
- 2020-07-16 CN CN202010688519.9A patent/CN111960217B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0849208A1 (en) * | 1996-12-20 | 1998-06-24 | Otis Elevator Company | Method and Apparatus to Inspect Hoisting Ropes |
| NO20005336D0 (en) * | 1999-10-21 | 2000-10-23 | Inventio Ag | Rope device and for this suitable artificial fiber rope and its use |
| CN1397797A (en) * | 2001-07-12 | 2003-02-19 | 因温特奥股份公司 | Detector for detecting wear of bearing rope |
| CN103364289A (en) * | 2013-07-22 | 2013-10-23 | 宁波市鄞州隆茂冲压件厂 | Testing device of repeated bending stress of metal spring clip |
| CN104679834A (en) * | 2015-02-06 | 2015-06-03 | 北京工商大学 | Time series data cleaning method and system |
| CN106629330A (en) * | 2017-01-21 | 2017-05-10 | 上海峰景移动科技有限公司 | On-line detection system for tension of steel wire rope for elevator |
| CN107892223A (en) * | 2017-11-15 | 2018-04-10 | 中南大学 | A kind of hoisting container monitoring system and tension force fault monitoring method |
| CN108313846A (en) * | 2017-12-13 | 2018-07-24 | 浙江新再灵科技股份有限公司 | A kind of elevator platform velocity-measuring system and method based on gyroscope |
| CN108726299A (en) * | 2018-06-22 | 2018-11-02 | 广西烽火信息技术有限公司 | A method of based on air pressure change monitoring elevator running position |
| CN110386530A (en) * | 2019-07-16 | 2019-10-29 | 浙江大学 | A kind of elevator monitoring systems and method towards fault diagnosis and safe early warning |
| CN111302173A (en) * | 2019-11-25 | 2020-06-19 | 广东马上到网络科技有限公司 | Elevator steel wire rope abnormity detection method, device and equipment and readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 韩春文: "160吨提梁机应用施工技术", 《科技创新导报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111960217B (en) | 2022-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110208019B (en) | Dynamic threshold early warning method for monitoring state of mobile equipment | |
| CN103852271B (en) | High-speed train running gear fault diagnosis and remote monitoring system based on Internet of Things | |
| CN108069308B (en) | A kind of electric staircase failure diagnosis method based on sequential probability | |
| CN204689281U (en) | For the lift management system that elevator faults is diagnosed in advance | |
| CN110647133A (en) | Rail transit equipment state detection maintenance method and system | |
| CN107797537A (en) | A kind of prognostic and health management method applied to automatic production line | |
| CN110092255B (en) | Elevator monitoring and maintenance method and system based on Internet of things cloud platform | |
| CN105045983B (en) | A kind of bullet train axletree aging analysis method based on axle temperature data | |
| CN109099975A (en) | A kind of building structure health monitoring systems | |
| CN103529822A (en) | Method and device for detecting vehicle failure | |
| CN116105802B (en) | Underground facility safety monitoring and early warning method based on Internet of things | |
| CN116739384A (en) | Mining equipment operation management system based on 5G wireless communication | |
| CN109507992B (en) | Method, device and equipment for predicting faults of locomotive brake system components | |
| CN116227752A (en) | Park facility management system based on Internet of things | |
| CN108844612A (en) | A kind of identification method of transformer internal faults based on mathematical statistics probabilistic model | |
| CN114487361A (en) | Oil on-line monitoring system | |
| CN110471395A (en) | A kind of fault detection method, device, equipment and storage medium | |
| CN110533115B (en) | Quantitative evaluation method for transmission characteristics of track circuit based on variational modal decomposition | |
| CN116823220A (en) | Cable running state monitoring platform and equipment | |
| CN111960217B (en) | Elevator steel wire rope fault detection method based on Internet of things | |
| CN113032907A (en) | Method and system for correcting vehicle shaking disease data deviation based on waveform correlation | |
| CN109894476B (en) | Fault diagnosis method and device for hydraulic equipment of cold-rolled silicon steel production line | |
| KR102545672B1 (en) | Method and apparatus for machine fault diagnosis | |
| CN115650006B (en) | Elevator safety monitoring and early warning method and system based on big data | |
| CN114722866B (en) | Management and control system is tracked to equipment based on industry internet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |