CN111186741B

CN111186741B - Elevator door system health maintenance method and device

Info

Publication number: CN111186741B
Application number: CN202010014542.XA
Authority: CN
Inventors: 黄冠杰; 史喆; 晋文静; 李�杰
Original assignee: Beijing Cyberinsight Technology Co ltd
Current assignee: Beijing Cyberinsight Technology Co ltd
Priority date: 2020-01-07
Filing date: 2020-01-07
Publication date: 2020-11-24
Anticipated expiration: 2040-01-07
Also published as: CN111186741A

Abstract

The invention discloses a method and a device for maintaining the health of an elevator door system, wherein the method comprises the following steps: collecting current signals in the working process of the elevator door in real time; extracting a predictive feature from the current signal; screening the predicted features to obtain effective features; and inputting the effective characteristics into a pre-established health evaluation model, and obtaining the health value of the elevator door according to the output of the health evaluation model. By using the invention, the accidental fault shutdown of the elevator door can be reduced, and the operation safety of the elevator can be ensured.

Description

Elevator door system health maintenance method and device

Technical Field

The invention relates to the technical field of industrial predictive maintenance, in particular to a method and a device for maintaining the health of an elevator door system.

Background

The elevator is an indispensable transport means in people's daily life, and with the continuous development of social economy, city high-rise building quantity increases rapidly, and the elevator will more widely be used. The elevator door is an important component in an elevator, generally adopts a split door which is divided into a landing door and a car door, is an important safety device for preventing personnel and objects from falling into a hoistway, and mainly comprises a door, a guide rail frame, a pulley, a sliding block and the like. The working principle is that a pulley block is driven by a motor through a belt, so that the opening and closing of the elevator door are realized. However, the elevator door has high use frequency and diversified use scenes, and the stability and the reliability of the performance of the elevator door are related to the safety of the operation of the elevator.

Elevators often fail to operate properly due to door system failures, such as: the elevator door can not be normally opened and closed; the door is not opened after the landing and leveling; the door is not closed after a door closing button is pressed; the elevator receives the floor selection signal and the door is closed but cannot be started; the speed of opening and closing the door is not changed or slowed, and the like. Regular inspection and maintenance cannot guarantee all normal operation, and the maintenance is usually after-accident maintenance after the fault occurs, so that the life of passengers can be influenced, and even great potential safety hazards are generated. Therefore, the maintenance of the elevator door is of great significance. At present, maintenance of elevator doors is mainly faced with the following challenges:

1) the equipment maintenance is basically planned maintenance and highly depends on people, and the redundant maintenance phenomenon is serious;

2) the current monitoring system can only monitor the running state of the elevator door, belongs to passive monitoring and lacks early warning and health state evaluation;

3) post-mortem maintenance, resulting in unplanned shutdowns and high maintenance time and costs.

Disclosure of Invention

The embodiment of the invention provides a health maintenance method and a health maintenance device for an elevator door system, which are used for reducing accidental fault shutdown of the elevator door and ensuring the operation safety of the elevator.

Therefore, the invention provides the following technical scheme:

a method of elevator door system health maintenance, the method comprising:

collecting current signals of the elevator door in a specific working condition process in real time;

extracting a predictive feature from the current signal;

screening the predicted features to obtain effective features;

and inputting the effective characteristics into a pre-established health evaluation model, and obtaining the health value of the elevator door according to the output of the health evaluation model.

Optionally, the method further comprises establishing the health assessment model in the following manner:

collecting a large number of current signals in the working process of the elevator door as training data, and marking the health state of the training data to obtain a label corresponding to each training data;

performing working condition segmentation on the training data to obtain data of a specific working condition;

extracting a prediction characteristic from the data of the specific working condition;

screening the predicted features to obtain effective features;

and training by using the effective features and the labels to obtain a health assessment model.

Optionally, the specific operating conditions include: starting and keeping constant speed; the predictive features include any one or more of: mean, maximum, minimum, variance, and duration of the operating condition.

Optionally, the screening the predicted features to obtain effective features includes:

and screening the prediction features by using a Fisher criterion to obtain effective features.

Optionally, the method further comprises:

determining a failure threshold value by utilizing the training data and the health assessment model in advance;

and if the health value of the elevator door is smaller than the failure threshold value, giving an alarm.

Optionally, the determining a failure threshold using the training data and the health assessment model includes:

inputting each training data into the health assessment model to obtain a health value corresponding to each training data;

determining data distribution of the healthy labels and the invalid labels according to the labels and the healthy values corresponding to the training data;

and determining a failure threshold according to the data distribution.

An elevator door system health maintenance device, the device comprising:

the signal acquisition module is used for acquiring a current signal in the specific working condition process of the elevator door in real time;

a feature extraction module for extracting a predicted feature from the current signal;

the characteristic screening module is used for screening the prediction characteristics to obtain effective characteristics;

and the health evaluation module is used for inputting the effective characteristics into a pre-established health evaluation model and obtaining the health value of the elevator door according to the output of the health evaluation model.

Optionally, the apparatus further comprises: the model establishing module is used for establishing a health assessment model; the model building module comprises:

the data collection unit is used for collecting a large number of current signals in the working process of the elevator door as training data, and marking the health state of the training data to obtain a label corresponding to each training data;

the working condition segmentation unit is used for carrying out working condition segmentation on the training data to obtain data of a specific working condition;

the characteristic extraction unit is used for extracting prediction characteristics from the data of the specific working condition;

the characteristic screening unit is used for screening the prediction characteristics to obtain effective characteristics;

and the training unit is used for training by utilizing the effective features and the labels to obtain a health assessment model.

Optionally, the apparatus further comprises:

the threshold value determining module is used for determining a failure threshold value by utilizing the training data and the health assessment model in advance;

and the alarm module is used for giving an alarm when the health value of the elevator door is smaller than the failure threshold value.

Optionally, the threshold determination module includes:

the calculation unit is used for inputting each training data into the health assessment model to obtain a health value corresponding to each training data;

the distribution determining unit is used for determining the data distribution of the healthy labels and the invalid labels according to the labels and the healthy values corresponding to the training data;

and the threshold value determining unit is used for determining a failure threshold value according to the data distribution.

According to the elevator door system health maintenance method and device provided by the embodiment of the invention, the current data of the elevator door in a specific working condition process is collected in real time, and the health state of the operation of the elevator door is evaluated by utilizing the pre-established health evaluation model to obtain the evaluation result, so that the initiative and online monitoring of the health state of the elevator door system is realized.

Furthermore, the elevator door can be early-warned before the elevator door is about to break down according to the evaluation result, and data support and scientific basis for maintenance personnel to make a reasonable maintenance plan are provided, so that accidental fault shutdown of equipment is reduced, and the operation safety of the elevator is guaranteed; moreover, unnecessary shutdown caused by redundant maintenance is avoided, and labor and time cost are saved.

Drawings

In order to more clearly illustrate the embodiments of the present application or technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings can be obtained by those skilled in the art according to the drawings.

FIG. 1 is a flow chart of the establishment of a health assessment model in an embodiment of the present invention;

FIG. 2 is an example of an elevator front door daily health trend graph in an embodiment of the present invention;

fig. 3 is a flow chart of a method of health maintenance of an elevator door system in accordance with an embodiment of the present invention;

FIG. 4 is a flow chart of determining a failure threshold in an embodiment of the present invention;

fig. 5 is a block diagram of one configuration of an elevator door system health maintenance device according to an embodiment of the present invention;

fig. 6 is another block diagram of the health maintenance device of an elevator door system according to an embodiment of the present invention.

Detailed Description

In order to make the technical field of the invention better understand the scheme of the embodiment of the invention, the embodiment of the invention is further described in detail with reference to the drawings and the implementation mode.

The embodiment of the invention provides a method and a device for maintaining the health of an elevator door system, which are used for evaluating the running health state of the elevator door by acquiring current data of the specific working condition process of the elevator door in real time and utilizing a pre-established health evaluation model to obtain an evaluation result, thereby realizing the initiative and online monitoring of the health state of the elevator door system.

Considering that the door opening or closing process at one time comprises a plurality of complex operations, when the health assessment model is established, the collected training data can be subjected to working condition segmentation, and the health assessment model is obtained by utilizing data training under a specific working condition.

As shown in fig. 1, the process of establishing a health assessment model in the embodiment of the present invention includes the following steps:

step 101, collecting a large number of current signals in the working process of the elevator door as training data, and marking the health status of the training data to obtain labels corresponding to the training data.

For example, a hall sensor may be installed at a power supply of a motor driving an elevator door, and the hall sensor collects a current signal generated during operation of the elevator as training data.

Each collected training data can be labeled with a health state, and specifically, the training data can be labeled manually according to the running state of the elevator door. The label can be divided into a health label and a failure label, for example, a period of time immediately after the maintenance is completed can be used as the health label, and a period of time after the long-time operation and without the maintenance can be used as the failure label.

The tag is used to indicate the severity of failure of an elevator door component, which may be, for example, the following: level 0 (healthy), Level 1 (mild), Level2 (moderate), and Level 3 (severe), which correspond to different failure degrees, respectively.

It should be noted that, considering that the current noise in the motor operation process may affect the data analysis processing, the noise reduction processing may be further performed on the acquired current signal. For example, the noise reduction processing may be performed by using a Haar mother wavelet of wavelet decomposition, and since the Haar mother wavelet is similar to a square wave and has a shape similar to a main portion of a current signal, a better processing effect may be obtained.

In addition, it should be noted that, in the process of opening or closing the elevator door, abnormal operation is often encountered, for example, after a passenger in the elevator clicks a door closing button, the elevator door starts to close, in the process, a user outside the elevator door triggers a corresponding button, and the elevator door is opened again, and in the process, the current signal may be abnormal. In the embodiment of the invention, the collected data can be further subjected to anomaly detection, and the anomaly data in the collected data are removed, so that the interference on model training is avoided, and the obtained model parameters are more accurate.

The detection of the abnormal data may adopt the existing related technology, and the embodiment of the present invention is not limited.

And 102, performing working condition segmentation on the training data to obtain data of a specific working condition.

Because the one-time door opening or closing process of the elevator door comprises a plurality of operations, such as door starting, door accelerating, door uniform speed, door decelerating, door locking and the like, wherein the starting and uniform speed section is least influenced by the external influence factor interlocked with the whole structure, and the influence of the elevator door structure problem can be reflected most in the section, therefore, in the embodiment of the invention, the model training can be carried out by mainly utilizing the current signals in the motion process of the starting and uniform speed section. For this reason, four parts in each cycle from door opening to door closing can be separated by adopting a separation method, and 2 starting sections and 2 constant-speed sections are separated. That is, the specific operating conditions may include: the two working conditions of starting and uniform speed specifically comprise starting and uniform speed working conditions in the door opening process and starting and uniform speed working conditions in the door closing process.

The working condition segmentation can be distinguished manually, and the data subjected to wavelet analysis noise reduction can be used for simply identifying the gentle section of the data as the target working condition. The signal collected by the elevator door controller can be used as the basis for judging the running state of the elevator door in the prior art, and the required target working condition can be automatically cut out.

And 103, extracting a prediction characteristic from the data of the specific working condition.

In the embodiment of the present invention, for the current signal, the extracted prediction feature may include, but is not limited to, any one of the following: the average value, the maximum value, the minimum value, the variance, the working condition duration and other 20 characteristics under the two working conditions of starting and uniform speed, and the cycle time from one door opening to door closing are added to the total 21 characteristics.

And 104, screening the prediction characteristics to obtain effective characteristics.

Considering that the characteristics which are not related to the degradation of the elevator door components can bring noise interference to the model establishment, the extracted prediction characteristics are further screened, and effective characteristics are selected from the extracted prediction characteristics to be used as characteristic input of subsequent modeling.

In an embodiment of the present invention, the variable ordering may be performed using fisher (fisher) criterion in Linear Discriminant Analysis (LDA). Specifically, a fisher value is calculated for each feature, the fisher value being equal to the distance between different kinds of features divided by the internal distance of the same kind of features.

After the fisher value of each feature is calculated, the features with the highest fisher value in a set number (for example, 10) are selected as the features in the input model training, that is, the valid features.

Of course, the predicted features may also be screened by using a feature selection method such as a filtering method (Filter), a packing method (Wrapper), an embedding method (Embedded), and the like, so as to obtain effective features, which is not limited in the embodiment of the present invention.

And 105, training by using the effective features and the labels to obtain a health assessment model.

Specifically, a supervised training method may be adopted, in which a health label is used as a reference, a failure label is used as a sample of a decline state for training, a calculation model from an effective feature to a health value, that is, a logistic regression model, is established, the input is the effective feature, and the output is the health value. The process of establishing the health assessment model will be further described in detail below by taking a passenger elevator with front and rear doors both sides being opened and closed simultaneously and front and rear elevator doors having a completely identical and symmetrical structure as an example.

(1) Raw data collection: a Hall sensor is arranged on a power supply of a motor for driving an elevator door of a front door and a rear door, the elevator door is subjected to door opening and closing experiments with the experiment length not shorter than 4 hours every day for 3 months until pulley blocks of the front elevator door fail, and finally 17 groups of sample data of the front elevator door are obtained, wherein 14 groups of samples with health labels and 3 groups of samples with failure labels with different degrees (such as slight abrasion, moderate abrasion and serious failure) are used as training data, and the samples with the health labels of the rear elevator door 17 group are used as test data.

(2) Data preprocessing: the noise of the original signal is denoised using a Haar mother wavelet of the wavelet decomposition.

(3) Data segmentation: after data preprocessing, four parts of door opening starting, door opening, door closing starting and door closing in each door opening to door closing cycle are marked, a starting section and a constant speed section are divided, and finally 4 parts of starting (Segment 1) and constant speed (Segment 2) of a door opening stage and starting (Segment 3) and constant speed (Segment 4) of a door closing stage can be extracted in each cycle.

(4) Feature extraction: and merging the smooth section data of the switch door divided by the data, thus obtaining a collection of 4 stages of data of starting and uniform speed sections of the switch door for a plurality of times, and extracting characteristic values such as an average value, a maximum value, a minimum value, a variance, working condition duration and the like from each stage.

(5) And (3) feature screening: using a Fisher criterion to sort the variables, dividing the distance between different types of features by the internal distance of the same type, and after calculating the fisher value of each variable, selecting 10 variables with the highest fisher value as effective features to finally obtain the following feature value list:

TABLE 1

Numbering	Eigenvalue name
		1	Segment 1-mean value
2	Segment 1-minimum
		3	Segment 3-mean value
4	One complete cycle takes time
		5	Segment 3-minimum
6	Segment 1-duration of operating conditions
		7	Segment 4-duration of operating conditions
8	Segment 1-variance
		9	Segment 3-max
10	Segment 2-duration of operating conditions

(6) Establishing a health evaluation model: the screened effective characteristics in the training data are used for establishing an elevator door health assessment model, the front 14 groups of training data of the front elevator door are selected as health data, health labels are marked, the back 3 groups of training data are selected as failure data, and failure labels are marked. And training by taking the health label as a reference and taking the failure label as a sample of a decline state, and establishing a logistic regression model from the effective characteristics to the health value. As can be seen from the daily health trend chart of the front door shown in fig. 2, the data health values of the first 14 groups of health data and 1 group of slightly worn data are both above 0.8, since slight wear belongs to a sub-health state close to normal, while the health values of the last two test points with moderate wear and severe failure are lower.

According to the elevator door system health maintenance method provided by the embodiment of the invention, the health evaluation model is utilized to realize active and online monitoring of the health state of the elevator door system.

As shown in fig. 3, it is a flowchart of a health maintenance method for an elevator door system according to an embodiment of the present invention, and the method includes the following steps:

and 301, acquiring a current signal in the specific working condition process of the elevator door in real time.

As already explained above, the specific operating conditions include: the two working conditions of starting and uniform speed specifically comprise starting and uniform speed working conditions in the door opening process and starting and uniform speed working conditions in the door closing process.

Correspondingly, in practical application, the current working condition of the elevator door can be determined through a working condition segmentation technology; the current working condition of the elevator door can also be determined according to the time for sending the door opening and closing instruction and the time for detecting the opening or closing of the elevator door to the set position by the corresponding sensor, and the embodiment of the invention is not limited.

Step 302, extracting a predicted feature from the current signal.

Like the predictive features extracted during the previous model training process, the predictive features may include, but are not limited to, any of the following: mean, maximum, minimum, variance, and duration of the operating condition.

And 303, screening the prediction characteristics to obtain effective characteristics.

The screening process of the prediction features is the same as that in the previous model training process, and the Fisher criterion can be used for screening specifically, so that the features irrelevant to the decline of the elevator door components are removed, and the influence of the irrelevant features on the evaluation result is avoided.

And step 304, inputting the effective characteristics into a pre-established health assessment model, and obtaining the health value of the elevator door according to the output of the health assessment model.

According to the health value of the elevator door, whether the elevator door is in a healthy state at present can be determined, and in another embodiment of the method, an alarm can be further given when the health value of the elevator door is monitored to be smaller than a failure threshold value, for example, one or more of displaying alarm prompt information, lighting an alarm indicator lamp, sending an alarm prompt tone and the like can be adopted for giving an alarm.

The failure threshold may be determined using the pre-established health assessment model and the training data described above.

As shown in fig. 4, it is a flowchart of determining the failure threshold in the embodiment of the present invention, and includes the following steps:

step 401, inputting each training data into the health assessment model to obtain a health value corresponding to each training data.

Step 402, determining data distribution of the healthy label and the invalid label according to the label and the healthy value corresponding to each training data.

And 403, determining a failure threshold according to the data distribution.

With continued reference to the front door daily health trend graph shown in fig. 2, since the health values of the test points with the health labels are all above 0.8, the failure threshold value is determined by the division.

Correspondingly, the embodiment of the invention also provides a health maintenance device for an elevator door system, which is a structural block diagram of the health maintenance device as shown in fig. 5.

In this embodiment, the apparatus includes the following modules:

the signal acquisition module 501 is used for acquiring a current signal in the working process of the elevator door in real time;

a feature extraction module 502 for extracting a predicted feature from the current signal;

a feature screening module 503, configured to screen the predicted features to obtain effective features;

and the health evaluation module 504 is used for inputting the effective characteristics into a pre-established health evaluation model 500 and obtaining the health value of the elevator door according to the output of the health evaluation model 500.

It should be noted that, in consideration of the influence of current noise during the operation of the motor on data analysis and processing, a preprocessing module (not shown) may be further provided in the apparatus of the present invention, for performing noise reduction processing on the current signal acquired by the signal acquisition module 501. For example, the noise reduction processing may be performed by using a Haar mother wavelet of wavelet decomposition, and since the Haar mother wavelet is similar to a square wave and has a shape similar to a main portion of a current signal, a better processing effect may be obtained.

The health assessment model may be pre-built by a corresponding model building module (not shown), which may include the following elements:

It should be noted that, in practical applications, the model building module may be a part of the apparatus of the present invention, or may be independent of the apparatus of the present invention, and the embodiment of the present invention is not limited thereto.

Fig. 6 is a block diagram of another embodiment of the health maintenance device for an elevator door system of the present invention.

Unlike the embodiment shown in fig. 5, in this embodiment, the apparatus further includes the following modules:

a threshold determination module 601, configured to determine a failure threshold by using the training data and the health assessment model in advance;

an alarm module 602, configured to alarm when the health value of the elevator door is less than the failure threshold.

The threshold determining module 601 may include the following units:

According to the elevator door system health maintenance method and device provided by the embodiment of the invention, the current data of the elevator door in the working process is collected in real time, the health state of the elevator door operation is evaluated by utilizing the pre-established health evaluation model, and the evaluation result is obtained, so that the initiative and online monitoring of the health state of the elevator door system is realized.

It should be noted that, by using the scheme of the invention, the elevator door system faults caused by various different types of component failures (such as pulley block failure, door lock circuit short circuit, latch hook position deviation and the like) can be monitored, thereby effectively improving the safety of elevator operation.

It should be noted that, for the above embodiments of the apparatus of the present invention, since the function of each module and unit is implemented similarly to that of the corresponding method, the description of each embodiment of the apparatus is relatively simple, and relevant points can be referred to the description of the corresponding parts of the method embodiment.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, 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.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. Furthermore, the above-described system embodiments are merely illustrative, wherein modules and units illustrated as separate components may or may not be physically separate, i.e., may be located on one network element, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Those skilled in the art will appreciate that all or part of the steps in the above method embodiments may be implemented by a program to instruct relevant hardware to perform the steps, and the program may be stored in a computer-readable storage medium, referred to herein as a storage medium, such as: ROM/RAM, magnetic disk, optical disk, etc.

Correspondingly, the embodiment of the invention also provides a device for the health maintenance method of the door system, and the device is an electronic device, such as a mobile terminal, a computer, a tablet device, a personal digital assistant and the like. The electronic device may include one or more processors, memory; wherein the memory is used for storing computer executable instructions and the processor is used for executing the computer executable instructions to realize the method of the previous embodiments.

The present invention has been described in detail with reference to the embodiments, and the description of the embodiments is provided to facilitate the understanding of the method and apparatus of the present invention, and is intended to be a part of the embodiments of the present invention rather than the whole embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative effort shall fall within the protection scope of the present invention, and the content of the present description shall not be construed as limiting the present invention. Therefore, any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method of health maintenance of an elevator door system, the method comprising:

extracting a predicted feature from the current signal, the predicted feature comprising any one or more of: mean, maximum, minimum, variance and duration of operating conditions;

screening the predicted features to obtain effective features;

2. The method of claim 1, further comprising building a health assessment model by:

screening the predicted features to obtain effective features;

3. The method of claim 2, wherein the particular operating condition comprises: starting working conditions and constant speed working conditions.

4. The method of claim 1, wherein the screening the predicted features to obtain valid features comprises:

5. The method according to any one of claims 2 to 3, further comprising:

6. The method of claim 5, wherein determining a failure threshold using the training data and the health assessment model comprises:

and determining a failure threshold according to the data distribution.

7. An elevator door system health maintenance device, the device comprising:

a feature extraction module for extracting a predicted feature from the current signal, the predicted feature comprising any one or more of: mean, maximum, minimum, variance and duration of operating conditions;

8. The apparatus of claim 7, further comprising: the model establishing module is used for establishing a health assessment model; the model building module comprises:

9. The apparatus of claim 8, further comprising:

10. The apparatus of claim 9, wherein the threshold determination module comprises: