Elevator anomaly detection method and device, computer equipment and storage medium

Abstract

The invention relates to an elevator anomaly detection method and device, computer equipment, a storage medium and a computer program product. The method comprises the steps that a detection parameter set corresponding to a target elevator in a to-be-detected time period is obtained; the detection parameter set comprises detection parameter values corresponding to the detection parameters; on the basis of the training mean value and the training standard deviation corresponding to the detection parameters, the detection parameter values corresponding to the detection parameters are preprocessed, and a target parameter set corresponding to the to-be-detected time period is obtained; the training mean value and the training standard deviation are determined based on a plurality of training parameter sets, and the training parameter sets are used for training the initial anomaly detection model to obtain a target anomaly detection model; and based on the target parameter set corresponding to the to-be-detected time period and the target anomaly detection model, obtaining an anomaly detection result of the target elevator. By adopting the method, the accuracy of elevator anomaly detection can be improved.

Description

Technical Field

[0001] The present application relates to the field of automation technology, and in particular to an elevator abnormality detection method, device, computer equipment and storage medium. Background Art

[0002] With the development of automation technology, elevators are widely used in various scenarios. In order to ensure the stable operation of elevators, it is necessary to regularly perform abnormal detection on elevators to promptly discover and repair elevator faults.

[0003] In traditional technology, maintenance personnel regularly perform abnormality detection on elevators. The abnormality detection results rely on the experience and skills of the maintenance personnel, resulting in low accuracy of the abnormality detection results. Summary of the invention

[0004] Based on this, it is necessary to provide an elevator abnormality detection method, device, computer equipment, computer-readable storage medium and computer program product that can improve the accuracy of elevator abnormality detection in response to the above technical problems.

[0005] In a first aspect, the present application provides an elevator abnormality detection method. The method comprises:

[0006] Acquire a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters;

[0007] Based on the training mean and the training standard deviation corresponding to the detection parameter, the detection parameter value corresponding to the detection parameter is preprocessed to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain a target anomaly detection model;

[0008] Based on the target parameter set corresponding to the time period to be detected and the target anomaly detection model, an anomaly detection result of the target elevator is obtained.

[0009] In one embodiment, obtaining the abnormality detection result of the target elevator based on the target parameter set corresponding to the time period to be detected and the target abnormality detection model includes:

[0010] Based on the sequence of the multiple time periods to be detected, the multiple target parameter sets are combined into an initial matrix;

[0011] For each element value in the initial matrix, convert the element value into a corresponding grayscale value to obtain a grayscale matrix;

[0012] Based on the grayscale matrix, generating a detection feature image;

[0013] The detection feature image is input into the target abnormality detection model to obtain an abnormality detection result of the target elevator.

[0014] In one embodiment, the training of the initial anomaly detection model to obtain the target anomaly detection model includes:

[0015] Acquire a training parameter set corresponding to a plurality of training time periods; the training parameter set includes training parameter values ​​corresponding to the training parameters;

[0016] For each of the training parameters, calculate the mean and standard deviation of the training parameter values ​​corresponding to the training parameters in the plurality of training parameter sets to obtain the training mean and training standard deviation corresponding to the training parameter;

[0017] Preprocessing the training parameter set corresponding to the training time period based on the training mean and the training standard deviation corresponding to the training parameters to obtain a processing parameter set corresponding to the training time period;

[0018] Generate a training feature image based on a set of processing parameters corresponding to a continuous preset number of training time periods;

[0019] The initial anomaly detection model is trained based on the plurality of training feature images to obtain a target anomaly detection model.

[0020] In one embodiment, obtaining a set of detection parameters corresponding to the target elevator in the time period to be detected includes:

[0021] Obtaining a sampled voltage value and a sampled current value corresponding to each sampling time point of the traction machine corresponding to the target elevator during the time period to be detected;

[0022] For each of the sampled voltage values, based on the voltage zero point offset and the voltage proportional factor of the traction machine, determining an actual voltage value corresponding to the sampled voltage value;

[0023] For each of the sampled current values, based on the current zero point offset and the current proportional factor of the traction machine, determining an actual current value corresponding to the sampled current value;

[0024] Based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each of the sampling time points, a detection parameter set corresponding to the target elevator in the time period to be detected is determined.

[0025] In one embodiment, the determining of the detection parameter set corresponding to the target elevator in the time period to be detected based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each sampling time point includes:

[0026] Determining the detection angular velocity value corresponding to the time period to be detected based on the sampling voltage values ​​corresponding to the plurality of sampling time points;

[0027] Determine the detection voltage value corresponding to the time period to be detected based on the actual voltage values ​​corresponding to the plurality of sampling time points;

[0028] Determine the detection current value corresponding to the time period to be detected based on the actual current values ​​corresponding to the plurality of sampling time points;

[0029] Determine the detection power value corresponding to the time period to be detected based on the actual voltage values ​​and the actual current values ​​corresponding to the plurality of sampling time points;

[0030] Based on the detected angular velocity value, the detected voltage value, the detected current value and the detected power value, a detection parameter set corresponding to the target elevator in the time period to be detected is obtained.

[0031] In one embodiment, determining the detection angular velocity value corresponding to the time period to be detected based on the sampled voltage values ​​corresponding to the plurality of sampling time points includes:

[0032] For each phase sampling voltage, based on the directions of the sampling voltage values ​​corresponding to any two adjacent sampling time points, determining the number of zero-crossing points within the time period to be detected;

[0033] Determining the phase voltage frequency based on the length of the time period to be detected and the number of zero crossing points;

[0034] Averaging a plurality of the phase voltage frequencies to obtain a target voltage frequency;

[0035] Based on the target voltage frequency and the number of pole pairs of the traction machine, a detection angular velocity value corresponding to the time period to be detected is determined.

[0036] In a second aspect, the present application also provides an elevator abnormality detection device. The device comprises:

[0037] An acquisition module, used to acquire a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters;

[0038] A preprocessing module, used to preprocess the detection parameter values ​​corresponding to the detection parameters based on the training mean and the training standard deviation corresponding to the detection parameters, so as to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain a target anomaly detection model;

[0039] The detection module is used to obtain an abnormality detection result of the target elevator based on the target parameter set corresponding to the time period to be detected and the target abnormality detection model.

[0040] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0041] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the methods described in the first aspect.

[0042] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of any one of the methods described in the first aspect.

[0043] The above-mentioned elevator anomaly detection method, device, computer equipment, storage medium and computer program product obtain the detection parameter set corresponding to the target elevator in the time period to be detected, and pre-process the detection parameter values ​​corresponding to the detection parameters based on the training mean and training standard deviation corresponding to the detection parameters to obtain the target parameter set corresponding to the time period to be detected. Since the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter set is used to train the initial anomaly detection model to obtain the target anomaly detection model, it can be understood that the detection parameter set is pre-processed using the training mean and training standard deviation corresponding to the training parameters calculated according to the training parameter set. The obtained target parameter set is more suitable for the target anomaly detection model trained using the training parameter set. Based on the target parameter set and the target anomaly detection model corresponding to the time period to be detected, the anomaly detection result of the target elevator is obtained, thereby improving the accuracy of the anomaly detection result. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] FIG1 is a diagram showing an application environment of an elevator abnormality detection method according to an embodiment;

[0045] FIG2 is a schematic diagram of a flow chart of an elevator abnormality detection method in one embodiment;

[0046] FIG3 is a schematic flow chart of a step of determining an abnormality detection result in one embodiment;

[0047] FIG4 is a schematic diagram of detecting a feature image in one embodiment;

[0048] FIG5 is a schematic diagram of a flow chart of a target anomaly detection model determination step in one embodiment;

[0049] FIG6 is a schematic diagram of a flow chart of a step of determining a detection parameter set in one embodiment;

[0050] FIG7 is a schematic flow chart of a step of determining a detection parameter set in another embodiment;

[0051] FIG8 is a schematic diagram of a flow chart of a step of determining a detected angular velocity value in one embodiment;

[0052] FIG9 is a schematic diagram of a flow chart of elevator abnormality detection in one embodiment;

[0053] FIG10 is a block diagram of a structure of an elevator abnormality detection device in one embodiment;

[0054] FIG. 11 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0056] The elevator abnormality detection method provided in the embodiment of the present application can be applied in the application environment shown in FIG1. ​​Among them, the traction machine 102 is the core component of the elevator system, which is mainly used to drive the elevator car and the counterweight to move up and down to realize the rise and fall of the elevator; the data collector 104 is used to collect the voltage value and current value on the traction machine frequency converter. The data collector can be a sensor. The sampling frequency of the data collector can be set according to actual needs. For example, the sampling frequency of the data collector is set to 2 kHz or above 2 kHz; the computer device 106 is used to process the voltage value and current value collected by the data collector 104 to determine the abnormal detection result of the elevator abnormality detection. The data collector 104 and the computer device 106 can be located in the electrical control room of the elevator. The data collector 104 may include a voltage sensor and a current meter. The computer device 106 can be, but is not limited to, various industrial computers, personal computers, laptops, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The computer device 106 can be used alone to execute the elevator anomaly detection method provided in the embodiment of the present application. The computer device 106 and the server can also be used together to execute the elevator anomaly detection method provided in the embodiment of the present application. The server can be implemented as an independent server or a server cluster composed of multiple servers.

[0057] In one embodiment, as shown in FIG. 2 , a method for detecting an abnormality of an elevator is provided. This embodiment is described by taking the method applied to a computer device as an example, and includes steps 202 to 206 .

[0058] Step 202, obtaining a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters.

[0059] The target elevator refers to the elevator for abnormality detection. The time period to be detected refers to the time period for abnormality detection of the target elevator. The detection parameter set refers to a set of detection parameter values, and the detection parameter set includes detection parameter values ​​corresponding to the detection parameters. The detection parameter refers to a parameter characterizing the operation of the target elevator. The detection parameter may be at least one of the detection voltage, the detection current, the detection power and the detection angular velocity. For example, the detection parameter includes the detection current and the detection power, or the detection parameter includes the detection voltage, the detection current and the detection power, or the detection parameter includes the detection current, the detection power and the detection angular velocity, or the detection parameter includes the detection voltage, the detection current, the detection power and the detection angular velocity. The detection parameter value refers to the parameter value that characterizes the operation status of the target elevator. The detection parameter value corresponds to the detection parameter. One detection parameter corresponds to at least one detection parameter value. For example, if the traction machine is powered by a three-phase power supply, the detection voltage corresponds to three detection voltage values, namely, the detection voltage value of phase A, the detection voltage value of phase B, and the detection voltage value of phase C. The detection current corresponds to three detection current values, namely, the detection current value of phase A, the detection current value of phase B, and the detection current value of phase C. The detection power corresponds to three detection power values, namely, the detection power value of phase A, the detection power value of phase B, and the detection power value of phase C. Alternatively, the detection angular velocity corresponds to one detection angular velocity value.

[0060] Exemplarily, the computer device obtains a detection parameter value corresponding to each detection parameter of the target elevator in the time period to be detected, and combines the detection parameter values ​​corresponding to multiple detection parameters into a detection parameter set.

[0061] Step 204, based on the training mean and training standard deviation corresponding to the detection parameters, the detection parameter values ​​corresponding to the detection parameters are preprocessed to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain the target anomaly detection model.

[0062] Among them, the training mean refers to the mean calculated based on the detection parameter values ​​corresponding to the detection parameters in multiple training parameter sets, and the training mean corresponds to the detection parameter one-to-one. The training standard deviation refers to the standard deviation calculated based on the detection parameter values ​​corresponding to the detection parameters in multiple training parameter sets, and the training standard deviation corresponds to the detection parameter one-to-one. The detection parameter corresponds to a set of training means and training standard deviations. Preprocessing refers to the process of standardizing the detection parameter values. The target parameter set refers to the set obtained after preprocessing each detection parameter value in the detection parameter set. The training parameter set refers to the detection parameter set used to train the initial anomaly detection model. The initial anomaly detection model refers to an untrained anomaly detection model, and the target anomaly detection model refers to a trained anomaly detection model. The target anomaly detection model can be a convolutional neural network (CNN). For example, the target anomaly detection model is a ResNet (Residual Network) in a convolutional neural network, such as ResNet 50. Compared with ResNet 18, ResNet50 uses a Bottle neck Block, in which a 1x1 convolution layer is used to reduce the number of channels, the amount of calculation and the amount of parameters, and then a 3x3 convolution layer is used for convolution operations, so that ResNet 50 can reduce the computing cost and improve the computing efficiency while ensuring the network depth. In addition, due to the increase in the depth of the network, ResNet 50 can learn more texture details of the feature image, better distinguish normal feature images from abnormal feature images, and improve the accuracy of anomaly detection.

[0063] Exemplarily, the computer device obtains the training mean and training standard deviation corresponding to each detection parameter, and for the detection parameter value corresponding to each detection parameter in the detection parameter set, performs standardization on each detection parameter value corresponding to the detection parameter based on the training mean and training standard deviation corresponding to the detection parameter, to obtain the target parameter value corresponding to each detection parameter value, and obtains the target parameter set based on the target parameter value corresponding to each detection parameter.

[0064] In one embodiment, before step 204, it also includes: obtaining a training parameter set corresponding to multiple training time periods; the training parameter set includes training parameter values ​​corresponding to the detection parameters; for each detection parameter, calculating the mean and standard deviation of the training parameter values ​​corresponding to the detection parameter in multiple training parameter sets to obtain the training mean and training standard deviation corresponding to the detection parameter.

[0065] In one embodiment, before step 204, it also includes: obtaining a training parameter set corresponding to a plurality of training time periods; the training parameter set includes a training parameter value corresponding to the training parameter; for each training parameter, calculating the mean and standard deviation of the training parameter values ​​corresponding to the training parameter in the plurality of training parameter sets, to obtain a training mean and a training standard deviation corresponding to the training parameter; based on the training mean and the training standard deviation corresponding to the training parameter, preprocessing the training parameter set corresponding to the training time period to obtain a processing parameter set corresponding to the training time period; based on the processing parameter sets corresponding to a preset number of consecutive training time periods, generating a training feature image; and training an initial anomaly detection model based on the plurality of training feature images to obtain a target anomaly detection model.

[0066] In one embodiment, before step 204, it also includes: based on the detection parameter sets corresponding to a continuous preset number of time periods to be detected, based on the sequence of multiple time periods to be detected, multiple detection parameter sets are formed into a detection matrix; step 204 includes: based on the training mean and training standard deviation corresponding to the detection parameters, the detection parameter values ​​corresponding to the detection parameters in the initial matrix are preprocessed to obtain the initial matrix; for each element value in the initial matrix, the element value is converted into a corresponding grayscale value to obtain a grayscale matrix; based on the grayscale matrix, a detection feature image is generated.

[0067] Step 206, based on the target parameter set corresponding to the time period to be detected and the target anomaly detection model, obtain the anomaly detection result of the target elevator.

[0068] Among them, the abnormal detection result refers to the result of abnormal detection on the target elevator, and the abnormal detection result includes but is not limited to at least one of the detection result and the abnormal probability, etc. For example, the abnormal detection result can be one of the results of normal and fault, or the abnormal detection result can be the abnormal probability.

[0069] Exemplarily, the computer device generates a detection feature image based on a target parameter set corresponding to multiple time periods to be detected, inputs the detection feature image into a target anomaly detection model, and obtains an anomaly detection result of the target elevator.

[0070] In one embodiment, based on the sequence of multiple time periods to be detected, multiple target parameter sets are formed into an initial matrix; for each element value in the initial matrix, the element value is converted into a corresponding grayscale value to obtain a grayscale matrix; the grayscale matrix is ​​input into a target anomaly detection model to obtain an anomaly detection result of the target elevator.

[0071] In one embodiment, after step 206, the method further includes: generating fault prompt information corresponding to the target elevator when the abnormality detection result is a fault. The fault prompt information may be a prompt tone or a prompt message.

[0072] In this embodiment, a detection parameter set corresponding to the target elevator in the time period to be detected is obtained, and the detection parameter values ​​corresponding to the detection parameters are preprocessed based on the training mean and the training standard deviation corresponding to the detection parameters to obtain a target parameter set corresponding to the time period to be detected. Since the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter set is used to train the initial anomaly detection model to obtain the target anomaly detection model, it can be understood that the detection parameter set is preprocessed using the training mean and the training standard deviation corresponding to the training parameters calculated according to the training parameter set, and the obtained target parameter set is more suitable for the target anomaly detection model trained using the training parameter set. Based on the target parameter set and the target anomaly detection model corresponding to the time period to be detected, the anomaly detection result of the target elevator is obtained, thereby improving the accuracy of the anomaly detection result.

[0073] In one embodiment, as shown in FIG3 , based on the target parameter set corresponding to the time period to be detected and the target anomaly detection model, an anomaly detection result of the target elevator is obtained, including:

[0074] Step 302: Based on the sequence of multiple time periods to be detected, multiple target parameter sets are formed into an initial matrix.

[0075] Among them, the initial matrix refers to a matrix composed of target parameter sets corresponding to multiple time periods to be detected in the order of multiple time periods to be detected. For example, there are 20 target parameter sets corresponding to time periods to be detected, and each target parameter set includes 10 target parameter values, namely, the target voltage value of phase A, the target voltage value of phase B, the target voltage value of phase C, the target current value of phase A, the target current value of phase B, the target current value of phase C, the target power value of phase A, the target power value of phase B, the target power value of phase C and the target angular velocity value, thus forming a 20×10 initial matrix, and the 10 target parameter values ​​in each row are a target parameter set corresponding to a time period to be detected.

[0076] Exemplarily, the computer device forms an initial matrix with target parameter sets corresponding to a preset number of time periods to be detected according to the sequence of the preset number of time periods to be detected.

[0077] Step 304: for each element value in the initial matrix, convert the element value into a corresponding grayscale value to obtain a grayscale matrix.

[0078] The gray value refers to a value representing the degree of brightness and darkness, and the gray value can range from 0 to 255. The gray matrix refers to a matrix composed of gray values, and the size of the gray matrix is ​​the same as the size of the initial matrix.

[0079] Exemplarily, for each element value in the initial matrix, the computer device converts the element value into a corresponding grayscale value based on a preset mapping relationship to obtain a grayscale matrix corresponding to the initial matrix.

[0080] Step 306: Generate a detection feature image based on the grayscale matrix.

[0081] Among them, the detection feature image refers to the feature grayscale image generated according to the grayscale matrix. It can be understood that each element in the grayscale matrix represents the grayscale value of a pixel in the detection feature image. For example, the detection feature image is shown in Figure 4. The initial matrix of 20×10 generates a 20×10 detection feature image. The left figure is a detection feature image with a normal anomaly detection result, and the right figure is a detection feature image with an abnormal anomaly detection result of abnormal.

[0082] Exemplarily, the computer device generates a detection feature image corresponding to the grayscale matrix.

[0083] Step 308: input the detection feature image into the target anomaly detection model to obtain an anomaly detection result of the target elevator.

[0084] Exemplarily, the computer device inputs the detection feature image into the target anomaly detection model, the target anomaly detection model outputs the anomaly detection result, and the computer device obtains the anomaly detection result of the target elevator.

[0085] In this embodiment, a detection feature image is generated from a grayscale matrix, that is, the correlation between the grayscale values ​​in the grayscale matrix is ​​converted into an adjacent relationship on the Y-axis, and the continuous relationship in time is converted into an adjacent relationship on the X-axis. The detection feature image is input into the target anomaly detection model. When the convolution kernel slides in the Y-axis direction, it will analyze the characteristic performance of different grayscale values ​​in the same time window. When the convolution kernel slides in the X-axis direction, it will analyze the changes of the same grayscale value in adjacent time windows. Through convolution processing, the target anomaly detection model can simultaneously analyze the corresponding relationship between the grayscale values ​​of various physical quantities in the same time period and the changing trend of the grayscale values ​​in different time periods, thereby improving the accuracy of the anomaly detection results.

[0086] In one embodiment, as shown in FIG. 5 , the initial anomaly detection model is trained to obtain a target anomaly detection model, including:

[0087] Step 502, obtaining a training parameter set corresponding to a plurality of training time periods; the training parameter set includes training parameter values ​​corresponding to the detection parameters.

[0088] The training parameter set refers to a set consisting of training parameter values, and the training parameter set includes the training parameter values ​​corresponding to the training parameters in the training time period. The training time period refers to the time period for obtaining the training parameter values, and the training time period is before the time period to be detected. The number of training parameter values ​​in the training parameter set is the same as the number of detection parameter values ​​in the detection parameter set.

[0089] Exemplarily, the computer device obtains a set of training parameters corresponding to multiple training time periods.

[0090] Step 504 , for each detection parameter, calculate the mean and standard deviation of the training parameter values ​​corresponding to the detection parameter in multiple training parameter sets to obtain the training mean and training standard deviation corresponding to the detection parameter.

[0091] The mean refers to the average value of a data set, and the standard deviation refers to the value that characterizes the degree of dispersion between each value in the data set and the mean.

[0092] Exemplarily, for each detection parameter, the computer device obtains training parameter values ​​corresponding to the detection parameter from multiple training parameter sets, calculates the mean and standard deviation of the obtained multiple training parameter values, and obtains the training mean and training standard deviation corresponding to the detection parameter.

[0093] Step 506 , preprocessing the training parameter set corresponding to the training time period based on the training mean and the training standard deviation corresponding to the detection parameters, to obtain a processing parameter set corresponding to the training time period.

[0094] The processing parameter set refers to a set consisting of processing parameter values. The processing parameter value refers to a value obtained by preprocessing the training parameter value.

[0095] Exemplarily, for each training parameter value in the training parameter set, the training parameter value is standardized based on the training mean and training standard deviation of the detection parameter corresponding to the training parameter value to obtain the processing parameter value corresponding to the training parameter value, and based on the processing parameter value corresponding to each training parameter value in the training parameter set, a processing parameter set corresponding to the training time period is obtained.

[0096] Step 508: Generate a training feature image based on a set of processing parameters corresponding to a continuous preset number of training time periods.

[0097] The preset number refers to a preset number, which can be set according to actual needs and is not limited here. For example, the preset number is 20 or 30. The training feature image refers to a feature grayscale image generated based on a processing parameter set corresponding to a continuous preset number of training time periods. The size of the training feature image is the same as the size of the detection feature image.

[0098] Exemplarily, the computer device organizes the target parameter sets corresponding to the preset number of training time periods into a training matrix based on the chronological order of a consecutive preset number of training time periods, and for each element value in the training matrix, converts the element value into a corresponding grayscale value based on a preset mapping relationship to obtain a sample matrix corresponding to the training matrix, and generates a training feature image corresponding to the sample matrix.

[0099] Step 510: Train the initial anomaly detection model based on multiple training feature images to obtain a target anomaly detection model.

[0100] Exemplarily, the computer device trains the initial anomaly detection model based on multiple training feature images and anomaly detection labels corresponding to the training feature images to obtain a target anomaly detection model. The anomaly detection label refers to an identifier that characterizes anomaly detection results corresponding to the training feature images.

[0101] In this embodiment, the training parameter set corresponding to the training time period is preprocessed by the training mean and training standard deviation corresponding to the detection parameters to obtain the processing parameter set corresponding to the training time period. The preprocessing eliminates the influence of different detection parameters due to different dimensions. The training feature image is generated based on the processing parameter set corresponding to a continuous preset number of training time periods. The initial anomaly detection model is trained based on multiple training feature images to obtain the target anomaly detection model. Not only the performance and stability of the target anomaly detection model are improved, but also the accuracy of the target anomaly detection model is improved by using multiple training feature images to train the initial anomaly detection model.

[0102] In one embodiment, as shown in FIG. 6 , obtaining a detection parameter set corresponding to the target elevator in the time period to be detected includes:

[0103] Step 602, obtaining the sampled voltage value and the sampled current value corresponding to each sampling time point of the traction machine corresponding to the target elevator during the detection time period.

[0104] Among them, the sampled voltage value refers to the voltage value corresponding to the traction machine corresponding to the target elevator at the sampling time point. It can be understood that the analog voltage signal provided by the traction machine corresponding to the target elevator is converted by AD (Analog to Digital) to obtain the sampled voltage values ​​corresponding to multiple sampling time points. The sampled voltage value is not the real voltage value. The sampled voltage value represents the voltage amplitude corresponding to the sampling time point. The sampled current value refers to the current value corresponding to the traction machine corresponding to the target elevator at the sampling time point. It can be understood that the analog current signal provided by the traction machine corresponding to the target elevator is converted by AD to obtain the sampled current values ​​corresponding to multiple sampling time points. The sampled current value is not the real current value. The sampled current value represents the current amplitude corresponding to the sampling time point.

[0105] Exemplarily, the computer device obtains the analog voltage signal and analog current signal of the traction motor corresponding to the target elevator during the detection time period through a data acquisition device. The computer samples the analog voltage signal based on a preset frequency to obtain the sampled voltage value corresponding to each sampling time point, and samples the analog current signal based on the preset frequency to obtain the sampled current value corresponding to each sampling time point.

[0106] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, and the computer device obtains the analog voltage signal corresponding to A, the analog voltage signal corresponding to B, the analog voltage signal corresponding to C, the analog current signal corresponding to A, the analog current signal corresponding to B and the analog current signal corresponding to C of the traction machine during the detection time period through a data collector, and the computer device samples the analog voltage signal corresponding to A based on a preset frequency to obtain the sampled voltage value corresponding to A at each sampling time point, samples the analog voltage signal corresponding to B based on a preset frequency to obtain the sampled voltage value corresponding to B at each sampling time point, samples the analog voltage signal corresponding to C based on a preset frequency to obtain the sampled voltage value corresponding to C at each sampling time point, samples the analog current signal corresponding to A based on a preset frequency to obtain the sampled current value corresponding to A at each sampling time point, samples the analog current signal corresponding to B based on a preset frequency to obtain the sampled current value corresponding to B at each sampling time point, and samples the analog current signal corresponding to C based on a preset frequency to obtain the sampled current value corresponding to C at each sampling time point.

[0107] Step 604: for each sampled voltage value, based on the voltage zero point offset and the voltage proportional factor of the traction machine, determine the actual voltage value corresponding to the sampled voltage value.

[0108] The voltage zero offset is used to correct the reference offset caused by hardware or noise in voltage measurement. The voltage scale factor is the proportional relationship between the sampled voltage value and the actual voltage value.

[0109] Exemplarily, for each sampled voltage value, the computer multiplies the difference between the sampled voltage value and the voltage zero point offset corresponding to the sampled voltage value by the voltage proportional factor to obtain the actual voltage value corresponding to the sampled voltage value.

[0110] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, then the actual voltage value corresponding to A is , the actual voltage value corresponding to B The actual voltage value corresponding to C The calculation formula is as follows:

[0111] Formula (1)

[0112] Where, t is the sampling time point; is the sampling voltage value of phase A corresponding to the sampling time point; is the voltage zero point offset corresponding to A; is the voltage proportional factor corresponding to A;

[0113] is the sampling voltage value of phase B corresponding to the sampling time point; is the voltage zero point offset corresponding to B; is the voltage proportional factor corresponding to B; is the sampling voltage value of phase C corresponding to the sampling time point; is the voltage zero point offset corresponding to C; is the voltage proportional factor corresponding to C.

[0114] Step 606: for each sampled current value, based on the current zero offset and the current proportional factor of the traction machine, determine the actual current value corresponding to the sampled current value.

[0115] The current zero offset is used to correct the reference offset caused by hardware or noise in current measurement. The current proportional factor is the proportional relationship between the sampled current value and the actual current value.

[0116] Exemplarily, for each sampled current value, the computer device multiplies the difference between the sampled current value and the current zero offset corresponding to the sampled current value by the current proportional factor to obtain the actual current value corresponding to the sampled current value.

[0117] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, then the actual current value corresponding to A is , the actual current value corresponding to B The actual current value corresponding to C The calculation formula is as follows:

[0118] Formula (2)

[0119] Where, t is the sampling time point; is the sampling current value of phase A corresponding to the sampling time point; is the current zero point offset corresponding to A; is the current proportional factor corresponding to A; is the sampling current value of phase B corresponding to the sampling time point; is the current zero point offset corresponding to B;

[0120] is the current proportional factor corresponding to B; is the sampling current value of phase C corresponding to the sampling time point; is the current zero point offset corresponding to C; is the current proportional factor corresponding to C.

[0121] Step 608: Determine a detection parameter set corresponding to the target elevator in the detection time period based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each sampling time point.

[0122] Exemplarily, the computer device determines the detection angular velocity value corresponding to the time period to be detected based on the sampled voltage values ​​corresponding to multiple sampling time points, determines the detection voltage value, detection current value and detection power value corresponding to the time period to be detected based on the actual voltage values ​​and actual current values ​​corresponding to the multiple sampling time points, and obtains the detection parameter set corresponding to the target elevator in the time period to be detected based on the detection angular velocity value, detection voltage value, detection current value and detection power value corresponding to the time period to be detected.

[0123] In this embodiment, for each sampled voltage value, the actual voltage value corresponding to the sampled voltage value is determined based on the voltage zero point offset and the voltage proportional factor of the traction machine. For each sampled current value, the actual current value corresponding to the sampled current value is determined based on the current zero point offset and the current proportional factor of the traction machine. Based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each sampling time point, the detection parameter set corresponding to the target elevator in the time period to be detected is determined, so as to provide accurate basic data for the subsequent determination of the abnormal detection result of the target elevator.

[0124] In one embodiment, as shown in FIG. 7 , based on the sampled voltage value, the actual voltage value, and the actual current value corresponding to each sampling time point, a detection parameter set corresponding to the target elevator in the time period to be detected is determined, including:

[0125] Step 702: Determine the detection angular velocity value corresponding to the time period to be detected based on the sampled voltage values ​​corresponding to the multiple sampling time points.

[0126] The detected angular velocity value refers to a value representing the angular velocity of the traction machine during the time period to be detected.

[0127] Exemplarily, the computer device determines the number of zero-crossing points in the time period to be detected using a zero-crossing method based on the sampled voltage values ​​corresponding to multiple sampling time points, and determines the detection angular velocity value corresponding to the time period to be detected based on the number of zero-crossing points.

[0128] Step 704: Determine a detection voltage value corresponding to the time period to be detected based on the actual voltage values ​​corresponding to the multiple sampling time points.

[0129] Exemplarily, the computer device calculates the average of the actual voltage values ​​corresponding to the multiple sampling times to obtain the detection voltage value corresponding to the time period to be detected.

[0130] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, and the actual voltage value of the sampling point corresponding to A is , the actual voltage value of the sampling point corresponding to B The actual voltage value at the sampling point corresponding to C , the calculation formula of the detection voltage value is as follows:

[0131] Formula (3)

[0132] Formula (4)

[0133] Formula (5)

[0134] in, is the detection voltage value of phase A corresponding to the time period to be detected;

[0135] is the detection voltage value of phase B corresponding to the time period to be detected; is the detection voltage value of phase C corresponding to the time period to be detected; t is the sampling time point; and n is the number of sampling time points in the time period to be detected.

[0136] Step 706 , based on the actual current values ​​corresponding to the multiple sampling time points, determine the detection current value corresponding to the time period to be detected.

[0137] Exemplarily, the computer device calculates the average of the actual current values ​​corresponding to multiple sampling time points to obtain the detection current value corresponding to the time period to be detected.

[0138] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, then the actual current value corresponding to A is , the actual current value corresponding to B The actual current value corresponding to C , then the calculation formula of the detection current value is as follows:

[0139] Formula (6)

[0140] Formula (7)

[0141] Formula (8)

[0142] in, is the detection current value of phase A corresponding to the time period to be detected; is the detection current value of phase B corresponding to the time period to be detected; is the detection current value of phase C corresponding to the time period to be detected; t is the sampling time point; and n is the number of sampling time points in the time period to be detected.

[0143] Step 708: Determine a detection power value corresponding to the time period to be detected based on the actual voltage values ​​and the actual current values ​​corresponding to the multiple sampling time points.

[0144] Exemplarily, for each sampling time point, the computer device multiplies the actual voltage value corresponding to the sampling time point by the actual current value to obtain the actual power value corresponding to the sampling time point, calculates the average of the actual power values ​​corresponding to multiple sampling time points, and obtains the detection power value corresponding to the time period to be detected.

[0145] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, and the calculation formula of the detection power value corresponding to the detection time period is as follows:

[0146] Formula (9)

[0147] Formula (10)

[0148] Formula (11)

[0149] in, The time period to be tested; is the detection power value of phase A corresponding to the time period to be detected; is the detection power value of phase B corresponding to the time period to be detected; is the detection power value of phase C corresponding to the time period to be detected; t is the sampling time point; n is the number of sampling time points in the time period to be detected;

[0150] Step 710, based on the detected angular velocity value, the detected voltage value, the detected current value and the detected power value, obtain a detection parameter set corresponding to the target elevator in the time period to be detected.

[0151] Exemplarily, the computer device combines the detection angular velocity value, the detection voltage value, the detection current value and the detection power value corresponding to the time period to be detected into a detection parameter set corresponding to the target elevator in the time period to be detected.

[0152] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, and the computer device combines the detection angular velocity value corresponding to the time period to be detected, the detection voltage value corresponding to phase A in the time period to be detected, the detection voltage value corresponding to phase B in the time period to be detected, the detection voltage value corresponding to phase C in the time period to be detected, the detection current value corresponding to phase A in the time period to be detected, the detection current value corresponding to phase B in the time period to be detected, the detection current value corresponding to phase C in the time period to be detected, the detection power value corresponding to phase A in the time period to be detected, the detection power value corresponding to phase B in the time period to be detected and the detection power value corresponding to phase C in the time period to be detected to form a detection parameter set corresponding to the target elevator in the time period to be detected.

[0153] In this embodiment, the detection parameter set corresponding to the target elevator in the detection time period is determined by the sampled voltage value, actual voltage value and actual current value corresponding to each sampling time point, providing accurate basic data for the subsequent determination of the abnormal detection result of the target elevator.

[0154] In one embodiment, as shown in FIG8 , based on the sampled voltage values ​​corresponding to the plurality of sampling time points, determining the detection angular velocity value corresponding to the time period to be detected includes:

[0155] Step 802: for each phase sampled voltage, based on the directions of the sampled voltage values ​​corresponding to any two adjacent sampling time points, determine the number of zero-crossing points within the time period to be detected.

[0156] The direction of the sampled voltage value refers to the polarity of the sampled voltage value, and the direction of the sampled voltage value can be represented by "+" and "-". The number of zero crossings refers to the total number of zero crossings of the sampled voltage value in the time period to be detected. It can be understood that if the directions of the sampled voltage values ​​corresponding to two adjacent sampling times are opposite, it is a zero crossing. The total number of zero crossings in the time period to be detected is the number of zero crossings.

[0157] Exemplarily, the computer device determines the number of zero-crossing points within the time period to be detected based on the directions of the sampled voltage values ​​corresponding to any two adjacent sampling time points for each phase sampled voltage.

[0158] Step 804: determining the phase voltage frequency based on the duration of the time period to be detected and the number of zero crossings.

[0159] The duration refers to the length of the time period to be detected, for example, the duration is 0.1 seconds. The phase voltage frequency refers to the voltage frequency of one phase, for example, the phase voltage frequency of phase A, the phase voltage frequency of phase B, and the phase voltage frequency of phase C.

[0160] Exemplarily, the computer device divides the number of zero-crossing points of each phase by twice the duration to obtain the phase voltage frequency of the item.

[0161] In one embodiment, the calculation formula of the phase voltage frequency is as follows:

[0162] Formula (12)

[0163] Wherein, P is the phase identifier, and P can be A, B or C; is the phase voltage frequency of phase P; N is the number of zero crossing points; T S The length of the time period to be detected.

[0164] Step 806: average multiple phase voltage frequencies to obtain a target voltage frequency.

[0165] Exemplarily, the computer device averages a plurality of phase voltage frequencies to obtain a target voltage frequency.

[0166] In one embodiment, the traction machine corresponding to the target elevator is powered by a three-phase power supply, and the target voltage frequency The calculation formula is as follows:

[0167] Formula (13)

[0168] in, is the phase voltage frequency of phase A; is the phase voltage frequency of phase B;

[0169] is the phase voltage frequency of phase C.

[0170] Step 808: Determine the detection angular velocity value corresponding to the time period to be detected based on the target voltage frequency and the pole pair number of the traction machine.

[0171] The number of pole pairs refers to the number of magnetic pole pairs inside the motor of the traction machine.

[0172] Exemplarily, the computer device determines the detection angular velocity value corresponding to the time period to be detected based on the ratio between the target voltage frequency and the number of pole pairs of the traction machine.

[0173] In one embodiment, the detected angular velocity value corresponding to the time period to be detected is The calculation formula is as follows:

[0174] Formula (14)

[0175] in, is the target voltage frequency; P is the number of pole pairs of the traction machine.

[0176] In this embodiment, based on the sampled voltage values ​​corresponding to multiple sampling time points, the detection angular velocity value corresponding to the time period to be detected is determined, providing accurate basic data for the subsequent determination of the detection parameter set corresponding to the target elevator in the time period to be detected.

[0177] In an exemplary embodiment, a flow chart of elevator abnormality detection is shown in FIG9 , including:

[0178] Anomaly detection model training process. The industrial computer obtains a set of training parameters corresponding to multiple training time periods. For each detection parameter, the industrial computer obtains the training parameter values ​​corresponding to the detection parameter from the multiple training parameter sets, calculates the mean and standard deviation of the obtained multiple training parameter values, and obtains the training mean and training standard deviation corresponding to the detection parameter; for each training parameter value in the training parameter set, based on the training mean and training standard deviation of the detection parameter corresponding to the training parameter value, the above training parameter value is standardized to obtain the processing parameter value corresponding to the above training parameter value, and based on the processing parameter value corresponding to each training parameter value in the training parameter set, the processing parameter set corresponding to the training time period is obtained; based on the sequence of a continuous preset number of training time periods, the target parameter sets corresponding to the preset number of training time periods are formed into a training matrix, for each element value in the training matrix, based on a preset mapping relationship, the element value is converted into a corresponding grayscale value, to obtain a sample matrix corresponding to the training matrix, and a training feature image corresponding to the sample matrix is ​​generated; based on multiple training feature images and the anomaly detection labels corresponding to the training feature images, the initial anomaly detection model is trained to obtain a target anomaly detection model.

[0179] Elevator abnormality detection process. The traction machine corresponding to the target elevator is powered by a three-phase power supply. The industrial computer obtains the analog voltage signal corresponding to A, the analog voltage signal corresponding to B, and the analog voltage signal corresponding to C of the traction machine during the detection time period through the voltage sensor, and obtains the analog current signal corresponding to A, the analog current signal corresponding to B, and the analog current signal corresponding to C of the traction machine during the detection time period through the current sensor. The industrial computer samples the analog voltage signal corresponding to A based on the preset frequency to obtain the sampled voltage value corresponding to A at each sampling time point, samples the analog voltage signal corresponding to B based on the preset frequency to obtain the sampled voltage value corresponding to B at each sampling time point, and samples the analog voltage signal corresponding to C based on the preset frequency to obtain the sampled voltage value corresponding to C at each sampling time point; samples the analog current signal corresponding to A based on the preset frequency to obtain the sampled current value corresponding to A at each sampling time point, samples the analog current signal corresponding to B based on the preset frequency to obtain the sampled current value corresponding to B at each sampling time point, and samples the analog current signal corresponding to C based on the preset frequency to obtain the sampled current value corresponding to C at each sampling time point.

[0180] For each sampling time point, use formula (1) to determine the actual voltage value corresponding to A , the actual voltage value corresponding to B The actual voltage value corresponding to C ; Use formula (2) to determine the actual current value corresponding to A , the actual current value corresponding to B The actual current value corresponding to C .

[0181] The industrial computer uses formulas (3), (4) and (5) to calculate the detection voltage value corresponding to phase A in the time period to be detected, the detection voltage value corresponding to phase B in the time period to be detected and the detection voltage value corresponding to phase C in the time period to be detected; uses formulas (6), (7) and (8) to calculate the detection current value corresponding to phase A in the time period to be detected, the detection current value corresponding to phase B in the time period to be detected and the detection current value corresponding to phase C in the time period to be detected; uses formulas (9), (10) and (11) to calculate the detection power value corresponding to phase A in the time period to be detected, the detection power value corresponding to phase B in the time period to be detected and the detection power value corresponding to phase C in the time period to be detected.

[0182] For phase A, the industrial computer uses the zero-crossing method to determine the number of zero-crossing points of phase A in the time period to be detected, and substitutes the number of zero-crossing points of phase A in the time period to be detected into formula (12) to obtain the phase voltage frequency of phase A; uses the zero-crossing method to determine the number of zero-crossing points of phase B in the time period to be detected, and substitutes the number of zero-crossing points of phase B in the time period to be detected into formula (12) to obtain the phase voltage frequency of phase B; uses the zero-crossing method to determine the number of zero-crossing points of phase C in the time period to be detected, and substitutes the number of zero-crossing points of phase C in the time period to be detected into formula (12) to obtain the phase voltage frequency of phase C. Substitute the phase voltage frequency of phase A, the phase voltage frequency of phase B, and the phase voltage frequency of phase C into formula (13) to obtain the target voltage frequency; substitute the target voltage frequency into formula (14) to obtain the detection angular velocity value corresponding to the time period to be detected.

[0183] The industrial computer combines the detection angular velocity value corresponding to the time period to be detected, the detection voltage value corresponding to phase A in the time period to be detected, the detection voltage value corresponding to phase B in the time period to be detected, the detection voltage value corresponding to phase C in the time period to be detected, the detection current value corresponding to phase A in the time period to be detected, the detection current value corresponding to phase B in the time period to be detected, the detection current value corresponding to phase C in the time period to be detected, the detection power value corresponding to phase A in the time period to be detected, the detection power value corresponding to phase B in the time period to be detected and the detection power value corresponding to phase C in the time period to be detected to form a detection parameter set corresponding to the target elevator in the time period to be detected.

[0184] Based on the detection parameter sets corresponding to the preset number of consecutive time periods to be detected, and based on the order of multiple time periods to be detected, multiple detection parameter sets are combined into a detection matrix; based on the training mean and training standard deviation corresponding to the detection parameters, the detection parameter values ​​corresponding to the detection parameters in the initial matrix are preprocessed to obtain the initial matrix; for each element value in the initial matrix, the element value is converted into the corresponding grayscale value to obtain the grayscale matrix; based on the grayscale matrix, a detection feature image is generated. The detection feature image is input into the target anomaly detection model to obtain the anomaly detection result of the target elevator.

[0185] When the abnormal detection result is a fault, fault prompt information corresponding to the target elevator is generated.

[0186] In this embodiment, a detection parameter set corresponding to the target elevator in the time period to be detected is obtained, and the detection parameter values ​​corresponding to the detection parameters are preprocessed based on the training mean and the training standard deviation corresponding to the detection parameters to obtain a target parameter set corresponding to the time period to be detected. Since the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter set is used to train the initial anomaly detection model to obtain the target anomaly detection model, it can be understood that the detection parameter set is preprocessed using the training mean and the training standard deviation corresponding to the training parameters calculated according to the training parameter set, and the obtained target parameter set is more suitable for the target anomaly detection model trained using the training parameter set. Based on the target parameter set and the target anomaly detection model corresponding to the time period to be detected, the anomaly detection result of the target elevator is obtained, thereby improving the accuracy of the anomaly detection result.

[0187] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0188] Based on the same inventive concept, the embodiment of the present application also provides an elevator anomaly detection device for implementing the elevator anomaly detection method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more elevator anomaly detection device embodiments provided below can refer to the limitations of the elevator anomaly detection method above, and will not be repeated here.

[0189] In one embodiment, as shown in FIG10 , an elevator abnormality detection device is provided, comprising: an acquisition module, a preprocessing module and a detection module, wherein:

[0190] An acquisition module 1002 is used to acquire a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters;

[0191] A preprocessing module 1004 is used to preprocess the detection parameter values ​​corresponding to the detection parameters based on the training mean and training standard deviation corresponding to the detection parameters to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain a target anomaly detection model;

[0192] The detection module 1006 is used to obtain an abnormality detection result of the target elevator based on the target parameter set corresponding to the time period to be detected and the target abnormality detection model.

[0193] In one embodiment, the detection module 1006 is also used to: based on the sequence of multiple time periods to be detected, form an initial matrix with multiple target parameter sets; for each element value in the initial matrix, convert the element value into a corresponding grayscale value to obtain a grayscale matrix; based on the grayscale matrix, generate a detection feature image; input the detection feature image into a target anomaly detection model to obtain an anomaly detection result of the target elevator.

[0194] In one embodiment, the elevator abnormality detection device also includes a training module, which is used to: obtain a training parameter set corresponding to multiple training time periods; the training parameter set includes training parameter values ​​corresponding to the detection parameters; for each detection parameter, calculate the mean and standard deviation of the training parameter values ​​corresponding to the detection parameter in multiple training parameter sets to obtain the training mean and training standard deviation corresponding to the detection parameter; based on the training mean and training standard deviation corresponding to the detection parameter, pre-process the training parameter set corresponding to the training time period to obtain a processing parameter set corresponding to the training time period; based on the processing parameter sets corresponding to a continuous preset number of training time periods, generate a training feature image; based on multiple training feature images, train the initial abnormality detection model to obtain a target abnormality detection model.

[0195] In one embodiment, the acquisition module 1002 is also used to: obtain the sampled voltage value and the sampled current value corresponding to each sampling time point of the traction machine corresponding to the target elevator within the time period to be detected; for each sampled voltage value, based on the voltage zero point offset and the voltage proportional factor of the traction machine, determine the actual voltage value corresponding to the sampled voltage value; for each sampled current value, based on the current zero point offset and the current proportional factor of the traction machine, determine the actual current value corresponding to the sampled current value; based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each sampling time point, determine the detection parameter set corresponding to the target elevator in the time period to be detected.

[0196] In one embodiment, the acquisition module 1002 is also used to: determine the detection angular velocity value corresponding to the time period to be detected based on the sampled voltage values ​​corresponding to multiple sampling time points; determine the detection voltage value corresponding to the time period to be detected based on the actual voltage values ​​corresponding to multiple sampling time points; determine the detection current value corresponding to the time period to be detected based on the actual current values ​​corresponding to multiple sampling time points; determine the detection power value corresponding to the time period to be detected based on the actual voltage values ​​and actual current values ​​corresponding to multiple sampling time points; and obtain the detection parameter set corresponding to the target elevator in the time period to be detected based on the detection angular velocity value, the detection voltage value, the detection current value and the detection power value.

[0197] In one embodiment, the acquisition module 1002 is also used to: for each phase sampling voltage, determine the number of zero-crossing points in the time period to be detected based on the direction of the sampling voltage values ​​corresponding to any two adjacent sampling time points; determine the phase voltage frequency based on the length of the time period to be detected and the number of zero-crossing points; average multiple phase voltage frequencies to obtain a target voltage frequency; determine the detection angular velocity value corresponding to the time period to be detected based on the target voltage frequency and the number of pole pairs of the traction machine.

[0198] Each module in the above elevator abnormality detection device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0199] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG11. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, an elevator abnormality detection method is implemented. The display unit of the computer device is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0200] Those skilled in the art will understand that the structure shown in FIG. 11 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0201] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0202] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0203] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0204] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0205] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0206] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0207] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. An elevator abnormality detection method, characterized in that: The method comprises: Acquire a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters; Based on the training mean and the training standard deviation corresponding to the detection parameter, the detection parameter value corresponding to the detection parameter is preprocessed to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain a target anomaly detection model; Based on the target parameter set corresponding to the time period to be detected and the target anomaly detection model, an anomaly detection result of the target elevator is obtained.

2. The method according to claim 1, characterized in that: The step of obtaining the abnormality detection result of the target elevator based on the target parameter set corresponding to the time period to be detected and the target abnormality detection model includes: Based on the sequence of the multiple time periods to be detected, the multiple target parameter sets are combined into an initial matrix; For each element value in the initial matrix, convert the element value into a corresponding grayscale value to obtain a grayscale matrix; Based on the grayscale matrix, generating a detection feature image; The detection feature image is input into the target abnormality detection model to obtain an abnormality detection result of the target elevator.

3. The method according to claim 1, characterized in that The training of the initial anomaly detection model to obtain the target anomaly detection model includes: Acquire a training parameter set corresponding to a plurality of training time periods; the training parameter set includes training parameter values ​​corresponding to the detection parameters; For each of the detection parameters, calculate the mean and standard deviation of the training parameter values ​​corresponding to the detection parameter in the plurality of training parameter sets to obtain the training mean and training standard deviation corresponding to the detection parameter; Preprocessing the training parameter set corresponding to the training time period based on the training mean and the training standard deviation corresponding to the detection parameter to obtain the processing parameter set corresponding to the training time period; Generate a training feature image based on a set of processing parameters corresponding to a continuous preset number of training time periods; The initial anomaly detection model is trained based on the plurality of training feature images to obtain a target anomaly detection model.

4. The method according to claim 1, characterized in that: The step of obtaining a detection parameter set corresponding to the target elevator in the time period to be detected includes: Obtaining a sampled voltage value and a sampled current value corresponding to each sampling time point of the traction machine corresponding to the target elevator during the time period to be detected; For each of the sampled voltage values, based on the voltage zero point offset and the voltage proportional factor of the traction machine, determining an actual voltage value corresponding to the sampled voltage value; For each of the sampled current values, based on the current zero point offset and the current proportional factor of the traction machine, determining an actual current value corresponding to the sampled current value; Based on the sampled voltage value, the actual voltage value and the actual current value corresponding to each of the sampling time points, a detection parameter set corresponding to the target elevator in the time period to be detected is determined.

5. The method according to claim 4, characterized in that The step of determining a detection parameter set corresponding to the target elevator in the time period to be detected based on the sampled voltage value, the actual voltage value, and the actual current value corresponding to each of the sampling time points includes: Determining the detection angular velocity value corresponding to the time period to be detected based on the sampling voltage values ​​corresponding to the plurality of sampling time points; Determine the detection voltage value corresponding to the time period to be detected based on the actual voltage values ​​corresponding to the plurality of sampling time points; Determine the detection current value corresponding to the time period to be detected based on the actual current values ​​corresponding to the plurality of sampling time points; Determine the detection power value corresponding to the time period to be detected based on the actual voltage values ​​and the actual current values ​​corresponding to the plurality of sampling time points; Based on the detected angular velocity value, the detected voltage value, the detected current value and the detected power value, a detection parameter set corresponding to the target elevator in the time period to be detected is obtained.

6. The method according to claim 5, characterized in that The step of determining the detection angular velocity value corresponding to the time period to be detected based on the sampled voltage values ​​corresponding to the plurality of sampling time points comprises: For each phase sampling voltage, based on the directions of the sampling voltage values ​​corresponding to any two adjacent sampling time points, determining the number of zero-crossing points within the time period to be detected; Determining the phase voltage frequency based on the length of the time period to be detected and the number of zero crossing points; Averaging a plurality of the phase voltage frequencies to obtain a target voltage frequency; Based on the target voltage frequency and the number of pole pairs of the traction machine, a detection angular velocity value corresponding to the time period to be detected is determined.

7. An elevator abnormality detection device, characterized in that: The device comprises: An acquisition module, used to acquire a detection parameter set corresponding to the target elevator in the time period to be detected; the detection parameter set includes detection parameter values ​​corresponding to the detection parameters; A preprocessing module, used to preprocess the detection parameter values ​​corresponding to the detection parameters based on the training mean and the training standard deviation corresponding to the detection parameters, so as to obtain a target parameter set corresponding to the time period to be detected; the training mean and the training standard deviation are determined based on multiple training parameter sets, and the training parameter sets are used to train the initial anomaly detection model to obtain a target anomaly detection model; The detection module is used to obtain an abnormality detection result of the target elevator based on the target parameter set corresponding to the time period to be detected and the target abnormality detection model.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images