Elevator operation fault detection and alarm system based on artificial intelligence

Abstract

The present invention relates to the field of data processing technology, and in particular to an elevator operation fault detection alarm system based on artificial intelligence; obtaining a suspected abnormal moment according to a vibration speed curve; obtaining a mutation coefficient according to the difference characteristics of the change speed of the vibration speed at the suspected abnormal moment and other moments; obtaining a discrete coefficient according to the difference characteristics of the discrete degree of the vibration speed in three axes at the suspected abnormal moment and other moments; obtaining a stability coefficient according to the distribution characteristics of the vibration speed in three axes at the suspected abnormal moment and adjacent moments; obtaining an abnormal index at the suspected abnormal moment according to the mutation coefficient, the discrete coefficient and the stability coefficient and determining the final abnormal moment. The present invention constructs an abnormality detection model based on the vibration speed training of the three axes at the final abnormal moment, detects and warns the elevator abnormality according to the abnormality detection model, and improves the accuracy of elevator abnormality detection.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an elevator operation fault detection and alarm system based on artificial intelligence. Background Art

[0002] Long-term use of elevators can cause mechanical structures to wear and vibrate, or even cause accidents. This is especially true for elevators in hospitals, which need to transport patients, have large spaces, high load capacity, and shorter mechanical structure lifespans. Therefore, elevator inspection and maintenance are required to ensure stable operation of the elevator. When an elevator is operating abnormally, it is easy to vibrate. Therefore, in order to quickly analyze whether the elevator is abnormal, the abnormality of the elevator can be determined based on the degree of vibration. However, hospital elevators often quickly transport beds and patients for emergency surgery, and family members of doctors and patients shake in the elevator, which causes the elevator to vibrate due to human factors, resulting in large vibration data values, which in turn triggers an abnormal elevator warning. Therefore, the accuracy of elevator abnormality detection based directly on the vibration size of the elevator is low. Summary of the invention

[0003] In order to solve the above-mentioned technical problem of low accuracy of elevator abnormality detection directly based on the vibration magnitude of the elevator, the purpose of the present invention is to provide an elevator operation fault detection and alarm system based on artificial intelligence, and the technical solution adopted is as follows:

[0004] A data acquisition module, used to respectively acquire vibration velocity curves of the three axes of the elevator;

[0005] A vibration analysis module is used to obtain a suspected abnormal moment according to the vibration velocity curves in three axes; obtain a mutation coefficient according to the difference characteristics of the vibration velocity change speed between the suspected abnormal moment and other moments; obtain a discrete coefficient according to the difference characteristics of the discrete degree of the vibration velocity in three axes between the suspected abnormal moment and other moments; obtain a stability coefficient according to the distribution characteristics of the vibration velocity in three axes between the suspected abnormal moment and adjacent moments;

[0006] An abnormality judgment module; used to obtain the abnormality index of the suspected abnormal moment according to the mutation coefficient, the dispersion coefficient and the stability coefficient; and obtain the final abnormal moment according to the abnormality index;

[0007] The anomaly detection module is used to train and build an anomaly detection model based on the vibration speed of the three axes at the final abnormal moment; and detect and warn of elevator anomalies based on the anomaly detection model.

[0008] Furthermore, the step of obtaining the suspected abnormal moment according to the vibration velocity curves in three axes includes:

[0009] The maximum value of the vibration speed in the three axes at any moment is taken as the effective value of the vibration at the any moment; when the effective value of the vibration exceeds the preset speed threshold, the any moment is the suspected abnormal moment.

[0010] Furthermore, the step of obtaining the mutation coefficient according to the difference characteristics of the vibration speed change at the suspected abnormal moment and other moments includes:

[0011] Calculate the rate of change of the vibration effective value at any moment and the adjacent previous moment to obtain the change characteristic value; calculate the absolute value of the difference between the change characteristic value at any moment and the average value of the change characteristic values ​​at all moments to obtain the change difference characteristic value at any moment; calculate the difference between the change difference characteristic value at the suspected abnormal moment and the average value of the change difference characteristic values ​​at all moments and normalize them to obtain the mutation coefficient at the suspected abnormal moment.

[0012] Furthermore, the step of obtaining the discrete coefficient according to the discrete degree difference characteristics of the vibration speed in three axes at the suspected abnormal moment and other moments includes:

[0013] Calculate the sum of the average value of the three axial vibration velocities at any moment and a preset minimum positive number to obtain the average characterization value of the axial velocity; calculate the ratio of the standard deviation of the three axial vibration velocities at any moment to the average characterization value of the axial velocity to obtain the discrete degree characteristic value; calculate the difference between the discrete degree characteristic value at the suspected abnormal moment and the average value of the discrete degree characteristic values ​​at all moments and normalize them to obtain the discrete coefficient at the suspected abnormal moment.

[0014] Furthermore, the step of obtaining the stability coefficient according to the distribution characteristics of the vibration speeds in three axes at the suspected abnormal moment and adjacent moments includes:

[0015] In the vibration velocity curve, the Euclidean distance between the suspected abnormal moment and the vibration velocity at the adjacent previous moment is calculated to obtain a first distance characteristic value; the Euclidean distance between the suspected abnormal moment and the vibration velocity at the adjacent subsequent moment is calculated to obtain a second distance characteristic value; the product of a preset first weight and the first distance characteristic value is calculated to obtain a first distance coefficient; the product of a preset second weight and the second distance characteristic value is calculated to obtain a second distance coefficient; the sum of the first distance coefficient and the second distance coefficient is calculated and negatively correlated to obtain an axial stability factor at the suspected abnormal moment; the sum of the three axial stability factors at the suspected abnormal moment is calculated to obtain the stability coefficient.

[0016] Furthermore, the step of obtaining the abnormality index at the suspected abnormal moment according to the mutation coefficient, the dispersion coefficient and the stability coefficient comprises:

[0017] The difference between constant 1 and the mutation coefficient is calculated to obtain a first abnormal factor; the difference between constant 1 and the discrete coefficient is calculated to obtain a second abnormal factor; the average value of the first abnormal factor, the second abnormal factor and the stability coefficient is calculated to obtain an abnormal index at the suspected abnormal moment.

[0018] Furthermore, the step of obtaining the final abnormal time according to the abnormal index includes:

[0019] When the abnormality index of the suspected abnormal moment exceeds a preset abnormality threshold, the suspected abnormal moment is the final abnormal moment.

[0020] The present invention has the following beneficial effects:

[0021] In the present invention, obtaining the suspected abnormal moment can preliminarily determine the moment of elevator shaking, and then the cause of elevator shaking can be analyzed according to the suspected abnormal moment; obtaining the mutation coefficient can be analyzed according to the change speed difference characteristics of the vibration speed caused by the abnormal elevator shaking and the human factor shaking, so as to determine the cause of the elevator shaking. Obtaining the discrete coefficient can be analyzed according to the discrete degree of the vibration speed of different axes caused by the abnormal elevator shaking and the human factor shaking, so as to determine the cause of the elevator shaking. Obtaining the stability coefficient can be analyzed according to the continuous stability characteristics of the vibration speed caused by the abnormal elevator shaking and the human factor shaking, so as to determine the cause of the electric shaking. Obtaining the abnormal index can accurately analyze the cause of the elevator shaking according to the mutation coefficient, the discrete coefficient and the abnormal index, and determine the moment of shaking caused by the elevator abnormality, so as to make the vibration speed of each axis learned in the abnormal detection model when the elevator is abnormally shaking more accurate, so that the elevator abnormal detection model at all times has higher detection accuracy for elevator abnormalities. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 A block diagram of an elevator operation fault detection and alarm system based on artificial intelligence provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of an elevator operation fault detection and alarm system based on artificial intelligence proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0025] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0026] The specific scheme of an elevator operation fault detection and alarm system based on artificial intelligence provided by the present invention is described in detail below with reference to the accompanying drawings.

[0027] See also Figure 1 , which shows a block diagram of an elevator operation fault detection and alarm system based on artificial intelligence provided by an embodiment of the present invention, the system includes the following modules:

[0028] The data acquisition module S1 is used to respectively acquire the vibration velocity curves of the three axes of the elevator.

[0029] In the embodiment of the present invention, the implementation scenario is to judge the working state of the elevator and improve the accuracy of elevator abnormality detection; the most common situation when the elevator is abnormal is jitter, and the vibration speed can quantify the degree of jitter of the elevator, so the vibration speed curves of the three axes of the elevator are obtained respectively; the three axes are the three directions of the three-dimensional rectangular coordinate system, and the vibration measuring instrument is installed at the center position of the top of the elevator. The vibration speeds of the three axes are collected simultaneously according to the preset collection frequency, and the vibration speeds of the same axis at adjacent moments are connected to obtain three different vibration speed curves of the three axes; in the embodiment of the present invention, the preset collection frequency is 1 time per second, and the implementer can determine it according to the implementation scenario.

[0030] The vibration analysis module S2 is used to obtain the suspected abnormal moment according to the vibration velocity curve of the three axes; obtain the mutation coefficient according to the difference characteristics of the change speed of the vibration velocity between the suspected abnormal moment and other moments; obtain the dispersion coefficient according to the difference characteristics of the dispersion degree of the vibration velocity in the three axes between the suspected abnormal moment and other moments; obtain the stability coefficient according to the distribution characteristics of the vibration velocity in the three axes between the suspected abnormal moment and adjacent moments.

[0031] The elevator is relatively smooth during normal travel, and its vibration speed is low. When the elevator shakes due to any situation, its vibration speed will increase, so the suspected abnormal moment can be obtained according to the vibration speed curve of the three axes; preferably, in the embodiment of the present invention, the step of obtaining the suspected abnormal moment includes: taking the maximum value of the vibration speed of the three axes at any time as the effective value of the vibration at that time, and the effective value of the vibration represents the maximum situation of the vibration speed at that time. The larger the effective value of the vibration, the more likely it is that abnormal shaking has occurred. When the effective value of the vibration exceeds the preset speed threshold, the arbitrary moment is a suspected abnormal moment; the suspected abnormal moment represents that the elevator has shaken at that moment, but it cannot accurately determine that the elevator has an abnormal condition, and further analysis is required based on the vibration speed of different axes.

[0032] Due to human factors such as doctors pushing the bed into the elevator quickly and people in the elevator shaking, the vibration speed will increase instantaneously, and the abnormal shaking of the elevator will keep the vibration speed at a high value for a long time; therefore, the change of the vibration speed can be analyzed to determine the reason for the increase in the vibration speed, so the mutation coefficient is obtained according to the difference characteristics of the change speed of the vibration speed at the suspected abnormal moment and other moments; preferably, in an embodiment of the present invention, the step of obtaining the mutation coefficient includes: calculating the change rate of the vibration effective value at any moment and the adjacent previous moment, obtaining the change characteristic value, and the first moment does not participate in the calculation; the change rate is the ratio of the change amount of the vibration speed to the corresponding interval time, which belongs to the prior art and will not be repeated; the larger the change characteristic value, the more likely it is that the elevator shaking is caused by human factors. Calculate the absolute value of the difference between the change characteristic value at any moment and the average value of the change characteristic value at all moments to obtain the change difference characteristic value at any moment; the change difference characteristic value reflects the degree of difference between the vibration change speed at any moment and the average vibration change speed, and the larger the change difference characteristic value, the more the vibration change speed at any moment deviates from the average level. Calculate the difference between the change difference characteristic value at the suspected abnormal moment and the average value of the change difference characteristic value at all moments and normalize them to obtain the mutation coefficient at the suspected abnormal moment. When the mutation coefficient is larger, it means that the change speed of the effective value of the vibration at the suspected abnormal moment is greater than the average level, and the effective value of the vibration presents more mutation characteristics. The suspected abnormal moment is more likely to be caused by human factors. When the mutation coefficient is smaller, it means that the change speed of the effective value of the vibration at the suspected abnormal moment is less obvious, and it is more likely to represent the real abnormal jitter of the elevator. The formula for obtaining the mutation coefficient includes:

[0033]

[0034] In the formula, represents the mutation coefficient of the i-th suspected abnormal moment, represents the normalization function, represents the change characteristic value at the i-th suspected abnormal moment, represents the average value of the changing characteristic value at all times, represents the change difference characteristic value of the suspected abnormal moment, N represents the number of all moments, represents the change difference characteristic value at the nth moment, Represents the average value of the change difference characteristic value at all times.

[0035] Furthermore, when the elevator shakes abnormally, the vibration speeds of the three axes collected usually change, and the overall stability of the elevator decreases; when the elevator shakes due to human factors, the vibration speeds of only one or two axes often change. Therefore, the cause of the shaking can be analyzed based on the difference characteristics of the vibration speeds of the three axes at the suspected abnormal moment, so the dispersion coefficient is obtained based on the difference characteristics of the dispersion degree of the vibration speeds of the three axes at the suspected abnormal moment and other moments.

[0036] Preferably, in an embodiment of the present invention, the step of obtaining the discrete coefficient includes: calculating the sum of the average value of the vibration speed of the three axes at any time and a preset minimum positive number to obtain the average characterization value of the axial speed; calculating the ratio of the standard deviation of the vibration speed of the three axes at any time to the average characterization value of the axial speed to obtain the discrete degree characteristic value; when the discrete degree characteristic value is larger, it means that the difference of the vibration speed of the three axes at any time is larger and more discrete. Calculate the difference between the discrete degree characteristic value at the suspected abnormal moment and the average value of the discrete degree characteristic value at all times and normalize it to obtain the discrete coefficient at the suspected abnormal moment; when the discrete degree characteristic value at the suspected abnormal moment is higher than the average value of the discrete degree characteristic value at all times, the larger the discrete coefficient, the larger the difference of the vibration speed of the three axes at the suspected abnormal moment, and the vibration speed of only one or two axes may increase significantly. The suspected abnormal moment is more likely to be caused by human factors. When the discrete coefficient is smaller, the difference of the vibration speed of the three axes at the suspected abnormal moment is smaller, and the vibration speed of the three axes increases significantly at the same time, which is more likely to be caused by elevator abnormality.

[0037] Furthermore, since the jitter caused by elevator abnormalities will exist for a long time, while the jitter caused by human factors is only instantaneous, the elevator abnormalities will cause the vibration speed values ​​of the three axes to be larger within a period of time, while the vibration speed caused by human factors will only be larger at a certain moment, and the vibration speed will drop at subsequent moments; therefore, in the vibration speed curve, if the vibration speeds at adjacent moments are close, the straight-line distance is shorter, and if the vibration speed difference is large, the straight-line distance is longer, and the stability coefficient can be obtained according to the distribution characteristics of the vibration speeds of the three axes at the suspected abnormal moment and the adjacent moment.

[0038] Preferably, in an embodiment of the present invention, the step of obtaining the stability coefficient includes: calculating the Euclidean distance between the suspected abnormal moment and the vibration velocity of the adjacent previous moment in the vibration velocity curve to obtain a first distance characteristic value; calculating the Euclidean distance between the suspected abnormal moment and the vibration velocity of the adjacent next moment to obtain a second distance characteristic value; it should be noted that the moments at both ends of the vibration velocity curve do not participate in the calculation; when the first distance characteristic value and the second distance characteristic value are smaller, it means that the vibration velocity of the suspected abnormal moment and the adjacent moment in the axis is similar, and it is more likely that the vibration is caused by the elevator abnormality. Calculate the product of the preset first weight and the first distance characteristic value to obtain the first distance coefficient; calculate the product of the preset second weight and the second distance characteristic value to obtain the second distance coefficient; in the embodiment of the present invention, the preset first weight is 0.4, and the preset second weight is 0.6, which can be determined by the implementer according to the implementation scenario. Calculate the sum of the first distance coefficient and the second distance coefficient and negatively correlated map to obtain a stability factor on an axis of the suspected abnormal moment; when the first distance coefficient and the second distance coefficient are smaller, the stability factor is larger, which means that the vibration velocity of the suspected abnormal moment and the adjacent moment is more similar, and the vibration characteristics are more stable. Calculate the sum of the stability factors of the three axes at the suspected abnormal moment to obtain the stability coefficient; when the stability coefficient is smaller, it means that the stability factors of the three axes at the suspected abnormal moment are smaller, the difference in vibration speed of the same axis at the suspected abnormal moment and the adjacent moment is greater, and the suspected abnormal moment is more likely to be caused by human factors. Conversely, when the stability coefficient is larger, it means that the vibration is more likely to be caused by elevator abnormality. The formula for obtaining the stability coefficient includes:

[0039]

[0040] In the formula, R represents the stability coefficient at the suspected abnormal moment, M represents the number of axes, Indicates the preset first weight, Indicates the preset second weight, represents the first distance eigenvalue corresponding to the mth axis, represents the second distance eigenvalue corresponding to the mth axis, represents the first distance coefficient, represents the second distance coefficient, represents an exponential function with a natural constant as base, Represents the stability factor.

[0041] Abnormal judgment module S3; used to obtain the abnormal index of the suspected abnormal moment according to the mutation coefficient, dispersion coefficient and stability coefficient; and obtain the final abnormal moment according to the abnormal index.

[0042] The mutation coefficient, the dispersion coefficient and the stability coefficient can all reflect the cause of the abnormal vibration speed at the suspected abnormal moment, so the abnormal index at the suspected abnormal moment can be obtained according to the mutation coefficient, the dispersion coefficient and the stability coefficient; preferably, in the embodiment of the present invention, the step of obtaining the abnormal index includes: calculating the difference between the constant 1 and the mutation coefficient to obtain the first abnormal factor; calculating the difference between the constant 1 and the dispersion coefficient to obtain the second abnormal factor; calculating the average value of the first abnormal factor, the second abnormal factor and the stability coefficient to obtain the abnormal index at the suspected abnormal moment. When the mutation coefficient is smaller, the dispersion coefficient is smaller, and the stability coefficient is larger, the abnormal index is larger, which means that the moment is more likely to be caused by the elevator abnormality; when the abnormal index is smaller, it is more likely to be caused by human factors.

[0043] Furthermore, the final abnormal moment can be obtained according to the abnormality index, specifically including: when the abnormality index of the suspected abnormal moment exceeds the preset abnormality threshold, the suspected abnormal moment is the final abnormal moment; in an embodiment of the present invention, the preset abnormality threshold is 0.5, and the implementer can determine it according to the implementation scenario. The final abnormal moment is the jitter moment caused by the elevator abnormality.

[0044] The anomaly detection module S4 is used to build an anomaly detection model based on the vibration speed training of the three axes at the final abnormal moment; and detect and warn of elevator anomalies based on the anomaly detection model.

[0045] According to the vibration velocity curve, the final abnormal moment is determined, and then the abnormal detection model can be constructed according to the vibration velocity training of the three axes at the final abnormal moment. It should be noted that the construction of the abnormal detection model through neural network training belongs to the prior art, and the specific steps are not repeated. According to the vibration velocity curves of multiple time periods of multiple different elevators and the vibration velocity of the three axes corresponding to the final abnormal moment, the neural network model is put into the training to identify the vibration velocity range of each axis when the elevator is abnormal, so that the constructed abnormal detection model can be detected during the operation of the elevator. Then, the elevator abnormality can be detected and warned according to the abnormal detection model. When the vibration velocity of the three axes of the elevator meets the abnormal vibration velocity range in the abnormal detection model for a continuous period of time, the detection and warning are carried out, thereby improving the accuracy of the abnormal fault detection of the elevator. It should be noted that the length of the continuous period of time is determined by the implementer according to the implementation scenario. So far, by obtaining the suspected abnormal moment in the vibration velocity curve of different axes of the elevator, the vibration velocity of the three axes at the suspected abnormal moment is analyzed according to the characteristic differences between the abnormal jitter of the elevator and the artificial jitter, and the real abnormal moment is determined; and the abnormal detection model is constructed according to the vibration velocity of each axis corresponding to the real abnormal moment, and the accuracy of the abnormal fault detection of the elevator is improved by the model.

[0046] In summary, the embodiment of the present invention provides an elevator operation fault detection and alarm system based on artificial intelligence; obtains the suspected abnormal moment according to the vibration speed curve; obtains the mutation coefficient according to the difference characteristics of the change speed of the vibration speed at the suspected abnormal moment and other moments; obtains the discrete coefficient according to the difference characteristics of the discrete degree of the vibration speed in three axes at the suspected abnormal moment and other moments; obtains the stability coefficient according to the distribution characteristics of the vibration speed in three axes at the suspected abnormal moment and adjacent moments; obtains the abnormal index of the suspected abnormal moment according to the mutation coefficient, the discrete coefficient and the stability coefficient and determines the final abnormal moment. The present invention constructs an abnormality detection model based on the vibration speed training of the three axes at the final abnormal moment, detects and warns the elevator abnormality according to the abnormality detection model, and improves the accuracy of elevator abnormality detection.

[0047] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0048] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An elevator operation fault detection and alarm system based on artificial intelligence, characterized in that: The system includes the following modules: A data acquisition module, used to respectively acquire vibration velocity curves of the three axes of the elevator; A vibration analysis module is used to obtain a suspected abnormal moment according to the vibration velocity curves in three axes; obtain a mutation coefficient according to the difference characteristics of the vibration velocity change speed between the suspected abnormal moment and other moments; obtain a discrete coefficient according to the difference characteristics of the discrete degree of the vibration velocity in three axes between the suspected abnormal moment and other moments; obtain a stability coefficient according to the distribution characteristics of the vibration velocity in three axes between the suspected abnormal moment and adjacent moments; An abnormality judgment module; used to obtain the abnormality index of the suspected abnormal moment according to the mutation coefficient, the dispersion coefficient and the stability coefficient; and obtain the final abnormal moment according to the abnormality index; An anomaly detection module is used to train and build an anomaly detection model based on the vibration speed of the three axes at the final abnormal moment; and detect and warn of elevator anomalies based on the anomaly detection model; The step of obtaining the stability coefficient according to the distribution characteristics of the vibration speeds in three axes at the suspected abnormal moment and the adjacent moments comprises: In the vibration velocity curve, the Euclidean distance between the suspected abnormal moment and the vibration velocity at the adjacent previous moment is calculated to obtain a first distance characteristic value; the Euclidean distance between the suspected abnormal moment and the vibration velocity at the adjacent subsequent moment is calculated to obtain a second distance characteristic value; the product of a preset first weight and the first distance characteristic value is calculated to obtain a first distance coefficient; the product of a preset second weight and the second distance characteristic value is calculated to obtain a second distance coefficient; the sum of the first distance coefficient and the second distance coefficient is calculated and negatively correlated to obtain an axial stability factor at the suspected abnormal moment; the sum of the three axial stability factors at the suspected abnormal moment is calculated to obtain the stability coefficient.

2. The elevator operation fault detection and alarm system based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the suspected abnormal moment according to the vibration velocity curves of the three axes comprises: The maximum value of the vibration speed in the three axes at any moment is taken as the effective value of the vibration at the any moment; when the effective value of the vibration exceeds the preset speed threshold, the any moment is the suspected abnormal moment.

3. The elevator operation fault detection and alarm system based on artificial intelligence according to claim 2 is characterized in that: The step of obtaining the mutation coefficient according to the difference characteristics of the change speed of the vibration speed at the suspected abnormal moment and other moments comprises: Calculate the rate of change of the vibration effective value at any moment and the adjacent previous moment to obtain the change characteristic value; calculate the absolute value of the difference between the change characteristic value at any moment and the average value of the change characteristic values ​​at all moments to obtain the change difference characteristic value at any moment; calculate the difference between the change difference characteristic value at the suspected abnormal moment and the average value of the change difference characteristic values ​​at all moments and normalize them to obtain the mutation coefficient at the suspected abnormal moment.

4. The elevator operation fault detection and alarm system based on artificial intelligence according to claim 2 is characterized in that: The step of obtaining the dispersion coefficient according to the dispersion degree difference characteristics of the vibration speed in three axes at the suspected abnormal moment and other moments comprises: Calculate the sum of the average value of the three axial vibration velocities at any moment and a preset minimum positive number to obtain the average characterization value of the axial velocity; calculate the ratio of the standard deviation of the three axial vibration velocities at any moment to the average characterization value of the axial velocity to obtain the discrete degree characteristic value; calculate the difference between the discrete degree characteristic value at the suspected abnormal moment and the average value of the discrete degree characteristic values ​​at all moments and normalize them to obtain the discrete coefficient at the suspected abnormal moment.

5. The elevator operation fault detection and alarm system based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the abnormality index at the suspected abnormal moment according to the mutation coefficient, the dispersion coefficient and the stability coefficient comprises: The difference between constant 1 and the mutation coefficient is calculated to obtain a first abnormal factor; the difference between constant 1 and the discrete coefficient is calculated to obtain a second abnormal factor; the average value of the first abnormal factor, the second abnormal factor and the stability coefficient is calculated to obtain an abnormal index at the suspected abnormal moment.

6. The elevator operation fault detection and alarm system based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the final abnormal time according to the abnormal index comprises: When the abnormality index of the suspected abnormal moment exceeds a preset abnormality threshold, the suspected abnormal moment is the final abnormal moment.

Images