A mine car abnormal fault detection method, device, equipment, medium and product
Detecting mine car failures through K-means clustering and artificial intelligence algorithms has solved the problem of untimely detection of mine dump vehicles in the existing technology, realizing timely warning and preventive maintenance of mine car failures, avoiding economic loss.
Patent Information
- Application Number
- CN202410811770.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-21
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-06-21
AI Technical Summary
The existing technology lacks a mature abnormal failure detection scheme for mining dump vehicles, which may result in secondary failures after mechanical failure, resulting in huge losses of economic value.
The K-means clustering algorithm and artificial intelligence algorithm are used to detect abnormal faults of mine trucks through monitoring data collected by multiple sensors, clustering to obtain fault types, train fault type identification models, output fault confidence, and trigger maintenance reminder actions.
Timely detection and early warning of abnormal faults of mine trucks is achieved, which avoids loss of secondary economic value, facilitates maintenance personnel to carry out timely maintenance, and improves the operational reliability of mine trucks.
Smart Images

Figure CN118568638B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of big data analysis, and specifically relates to a method, device, equipment, medium and product for detecting abnormal faults in a mine car. Background Art
[0002] A mine car is a narrow-gauge railway transport vehicle or off-road (the "off-road" here does not mean off-road driving, but is not allowed to be driven on roads because of its extra-wide shape and excessive total mass) mining dump vehicle that transports bulk materials such as coal, ore and waste rock in mines. The former generally requires towing by a locomotive or winch, while the latter is a heavy-duty dump truck used in open-pit mines to complete rock and earth stripping and ore transportation tasks. Its working characteristics are short travel distance and heavy load. It is often loaded with a large electric shovel or hydraulic shovel, and travels back and forth between the mining point and the unloading point.
[0003] Mining dump trucks are expensive to manufacture and have high prices for parts, so if they are repaired after a mechanical failure occurs, the secondary failures may cause huge economic value losses. Therefore, how to detect and analyze abnormal faults of mining trucks based on the monitoring data collected by multiple sensors arranged in the oil circuit and key components such as the engine of the mining truck, and trigger maintenance reminder actions in advance according to the detection and analysis results, so that maintenance personnel can eliminate possible mechanical failures through timely maintenance work to avoid secondary economic value losses, is a topic that technicians in this field urgently need to study. Summary of the invention
[0004] The purpose of the present invention is to provide a method, device, computer equipment, computer-readable storage medium and computer program product for detecting abnormal faults in mining vehicles, so as to solve the problem that there is currently no mature abnormal fault detection solution for mining dump vehicles, so that if maintenance is carried out after a mechanical failure occurs, secondary faults may occur, resulting in huge economic value losses.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions:
[0006] In a first aspect, a method for detecting abnormal faults in a mine car is provided, comprising:
[0007] Obtaining the mine car abnormal fault types of all mine car abnormal fault events and the historical monitoring time series data collected by a plurality of first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the plurality of first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs;
[0008] According to the historical monitoring time series data of all the mine car abnormal failure events, a K-means clustering algorithm is applied to obtain a plurality of cluster centers, wherein the total number of centers of the plurality of cluster centers is equal to the total number of mine car abnormal failure types of all the mine car abnormal failure events;
[0009] For each event in all the abnormal failure events of the mining vehicles, the corresponding Euler distances to the multiple cluster centers are calculated according to the multiple cluster centers and the corresponding historical monitoring time series data, and the corresponding event is used as an event belonging to a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance;
[0010] For each of the plurality of cluster centers, the occurrence frequencies of various types of abnormal fault types of all corresponding mine carts belonging to the events are counted, and a type corresponding to the highest frequency among the abnormal fault types of all corresponding mine carts belonging to the events is used as the corresponding abnormal fault type of the mine cart;
[0011] For various types of mine car abnormal faults in the mine car abnormal fault types of all mine car abnormal fault events, all corresponding positive sample data and all negative sample data are imported into a machine learning model based on an artificial intelligence algorithm for model training to obtain a corresponding fault type recognition model, wherein the positive sample data contains the historical monitoring time series data of the belonging event corresponding to the center used as a model input item and a value "1" used as a model output item, and the negative sample data contains the historical monitoring time series data of the belonging event not corresponding to the center used as a model input item and a value "0" used as a model output item;
[0012] Periodically acquiring current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle in a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors;
[0013] For the various types of abnormal faults of the mine car, the current monitoring time series data is imported into the corresponding fault type identification model, and the confidence of the target mine car having an abnormal fault of the corresponding type after a unit time period is obtained by output;
[0014] When it is found that the confidence level of the target mine car's abnormal fault belonging to a certain mine car abnormal fault type after a unit time period exceeds a preset confidence threshold corresponding to the certain mine car abnormal fault type, a maintenance reminder action corresponding to the certain mine car abnormal fault type is triggered.
[0015] Based on the above invention content, a new solution for detecting abnormal faults of mine cars based on clustering algorithm and artificial intelligence algorithm is provided, that is, firstly, according to the historical monitoring time series data of all abnormal fault events of mine cars, a K-means clustering algorithm is applied to cluster to obtain multiple cluster centers, and then for each center, the corresponding abnormal fault type of the mine car is determined according to the abnormal fault types of the mine car of all corresponding events, and then for various abnormal fault types of mine cars, the positive and negative sample data obtained based on the corresponding center and the non-corresponding center are imported into the machine learning model based on the artificial intelligence algorithm for model training to obtain the corresponding fault type recognition model, and then the corresponding fault type recognition model is obtained. The current monitoring time series data of the target mine car are imported into the corresponding fault type identification model respectively, and the confidence of the target mine car having an abnormal fault of the corresponding type after a unit time period is outputted. Finally, when it is found that the confidence exceeds the standard, the maintenance reminder action is triggered. In this way, the abnormal faults of the mine car can be detected and analyzed based on the monitoring data collected by multiple sensors arranged in the oil circuit and engine and other key components of the mine car, and the maintenance reminder action can be triggered in advance according to the detection and analysis results, so that maintenance personnel can eliminate possible mechanical failures through timely maintenance work to avoid secondary economic value losses, which is convenient for practical application and promotion.
[0016] In a possible design, the abnormal fault types of all the abnormal fault events of the mine cars include a failure type of being unable to lift, a slow lifting failure type, a failure type of being unable to stop after lifting, a hydraulic system howling failure type, a no-load sound failure type, a failure type of being unable to automatically descend, an engine failure type, a steering failure type, a brake failure type and / or an electrical failure type.
[0017] In one possible design, the plurality of first vehicle-mounted sensors or the plurality of second vehicle-mounted sensors include a noise sensor, a vibration sensor, an oil pressure sensor, a temperature sensor, a displacement sensor, a speed sensor, an inclination sensor and / or an acceleration sensor.
[0018] In a possible design, historical monitoring time series data of all abnormal fault events of mining vehicles collected by a plurality of first vehicle-mounted sensors within a unit time period before the corresponding event occurs are obtained, including:
[0019] Acquire original monitoring time series data of all abnormal fault events of mining vehicles collected by a plurality of first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the plurality of first vehicle-mounted sensors are arranged on the mining vehicles to which the corresponding events belong;
[0020] For each event among all the abnormal failure events of the mine cars and each sensor among the multiple first on-board sensors, the corresponding original monitoring time series data are processed by a data cleaning method, a null value filling method, an abnormal value exclusion method and / or an acquisition time alignment method to obtain the corresponding historical monitoring time series data.
[0021] In a possible design, for each event in all the abnormal fault events of the mining vehicles and each sensor in the plurality of first vehicle-mounted sensors, the corresponding original monitoring time series data is processed by applying the acquisition time alignment method to obtain the corresponding historical monitoring time series data, including:
[0022] For each event in all the abnormal failure events of the mine car and each sensor in the multiple first vehicle-mounted sensors, according to the preset multiple unified collection moments in the unit time period, the collection moment in the corresponding original monitoring time series data is shifted as a whole, so that each shifted collection moment corresponds to each moment in the multiple unified collection moments in the unit time period, and for each shifted collection moment, the corresponding historical monitoring time series data is calculated according to the following formula :
[0023]
[0024] In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The historical monitoring value corresponding to the collection time after the translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, , , and Respectively represent the original monitoring time series data and One-to-one correspondence between adjacent acquisition moments The original monitoring values are solved by the least square method and the coefficients of the local cubic polynomial used to implement cubic spline interpolation. represents a positive integer greater than or equal to 4, The adjacent recording time is the distance from the The closest previous acquisition time after translation A collection moment.
[0025] In a possible design, after obtaining the abnormal fault type of the mining car in each center, the method further includes:
[0026] If it is found that the abnormal fault type of the mine carts of at least two centers is the same, then the at least two centers are sorted in order from high to low according to the occurrence frequency of the same type to obtain a center queue, and the same type continues to be the abnormal fault type of the mine cart of the first center in the center queue;
[0027] For each non-first center in the center queue, the occurrence frequency of various other types of mine car abnormal fault types in the corresponding all belonging events are counted respectively, and a certain other type in the mine car abnormal fault types of all corresponding belonging events and corresponding to the highest frequency is updated as the corresponding mine car abnormal fault type, wherein the other type refers to other mine car abnormal fault types except the same type.
[0028] In a second aspect, a mine car abnormal fault detection device is provided, comprising an event data acquisition unit, a cluster center determination unit, an attribute event classification unit, a fault type classification unit, a recognition model training unit, a current data acquisition unit, a recognition model application unit and a maintenance reminder trigger unit;
[0029] The event data acquisition unit is used to acquire the mine car abnormal fault type of all mine car abnormal fault events and the historical monitoring time series data collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the multiple first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs;
[0030] The cluster center determination unit is communicatively connected to the event data acquisition unit, and is used to apply a K-means clustering algorithm to cluster the historical monitoring time series data of all the mine car abnormal failure events to obtain a plurality of cluster centers, wherein the total number of the plurality of cluster centers is equal to the total number of the mine car abnormal failure types of all the mine car abnormal failure events;
[0031] The belonging event division unit is respectively connected to the event data acquisition unit and the cluster center determination unit in communication, and is used to calculate the corresponding Euler distance to the multiple cluster centers for each event in all the abnormal fault events of the mining vehicles according to the multiple cluster centers and the corresponding historical monitoring time series data, and use the corresponding event as the belonging event of a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance;
[0032] The fault type classification unit is communicatively connected to the attributed event classification unit, and is used to count the occurrence frequencies of various types of abnormal fault types of the mine car of all corresponding attributed events for each center among the multiple cluster centers, and use a certain type of abnormal fault types of the mine car of all corresponding attributed events and corresponding to the highest frequency as the corresponding abnormal fault type of the mine car;
[0033] The identification model training unit is respectively communicatively connected to the event data acquisition unit and the fault type classification unit, and is used for importing all corresponding positive sample data and all corresponding negative sample data into a machine learning model based on an artificial intelligence algorithm for model training for various mine car abnormal fault types in the mine car abnormal fault types of all mine car abnormal fault events, so as to obtain a corresponding fault type identification model, wherein the positive sample data contains the historical monitoring time series data of the belonging event corresponding to the center used as a model input item and a value "1" used as a model output item, and the negative sample data contains the historical monitoring time series data of the belonging event not corresponding to the center used as a model input item and a value "0" used as a model output item;
[0034] The current data acquisition unit is used to periodically acquire current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle in a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors;
[0035] The identification model application unit is respectively connected to the identification model training unit and the current data acquisition unit for communication, and is used to import the current monitoring time series data into the corresponding fault type identification model for the various types of abnormal faults of the mine car, and output the confidence that the target mine car has an abnormal fault of the corresponding type after a unit time period;
[0036] The maintenance reminder trigger unit is communicatively connected to the identification model application unit, and is used to trigger the execution of a maintenance reminder action corresponding to a certain type of mine car abnormal fault when it is found that the confidence level of the mine car abnormal fault belonging to a certain type of mine car abnormal fault in the target mine car after a unit time period exceeds a preset confidence threshold corresponding to the certain type of mine car abnormal fault.
[0037] In a third aspect, the present invention provides a computer device comprising a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the mine car abnormal fault detection method as described in the first aspect or any possible design of the first aspect.
[0038] In a fourth aspect, the present invention provides a computer-readable storage medium having instructions stored thereon. When the instructions are executed on a computer, the method for detecting abnormal faults of a mine car as described in the first aspect or any possible design of the first aspect is executed.
[0039] In a fifth aspect, the present invention provides a computer program product, comprising a computer program or instructions, which, when executed by a computer, implements the method for detecting abnormal faults of a mine car as described in the first aspect or any possible design of the first aspect.
[0040] Beneficial effects of the above scheme:
[0041] (1) The present invention creatively provides a new solution for detecting abnormal faults of mine cars based on clustering algorithms and artificial intelligence algorithms, that is, firstly, according to the historical monitoring time series data of all abnormal fault events of mine cars, a K-means clustering algorithm is applied to cluster to obtain multiple cluster centers, and then for each center, the corresponding abnormal fault type of the mine car is determined according to the abnormal fault types of all corresponding events, and then for various abnormal fault types of mine cars, the positive and negative sample data obtained based on the corresponding center and the non-corresponding center are imported into a machine learning model based on an artificial intelligence algorithm for model training to obtain the corresponding fault type recognition model, and then the corresponding fault type recognition model is obtained. The current monitoring time series data of the target mine car are imported into the corresponding fault type identification model respectively, and the confidence of the target mine car having an abnormal fault of the corresponding type after a unit time period is outputted. Finally, when it is found that the confidence exceeds the standard, the maintenance reminder action is triggered. In this way, the abnormal faults of the mine car can be detected and analyzed based on the monitoring data collected by multiple sensors arranged in the oil circuit and engine and other key components of the mine car, and the maintenance reminder action can be triggered in advance according to the detection and analysis results, so that maintenance personnel can eliminate possible mechanical failures through timely maintenance work to avoid secondary economic value losses, which is convenient for practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be 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.
[0043] Figure 1 A schematic flow chart of a method for detecting abnormal faults in a mine car provided in an embodiment of the present application.
[0044] Figure 2 A schematic diagram of the structure of a mine car abnormal fault detection device provided in an embodiment of the present application.
[0045] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.
[0047] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.
[0048] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.
[0049] Example
[0050] like Figure 1 As shown, the mine car abnormal fault detection method provided in the first aspect of this embodiment can be, but is not limited to, executed by a computer device having certain computing resources and communicating with a plurality of second vehicle-mounted sensors arranged on the target mine car, for example, by an on-board ECU (Electronic Control Unit, i.e., computer control module) device of the target mine car. Figure 1 As shown, the mine car abnormal fault detection method may include but is not limited to the following steps S1 to S8.
[0051] S1. Obtain the mine car abnormal fault types of all mine car abnormal fault events and the historical monitoring time series data collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the multiple first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs.
[0052] In step S1, the mine car abnormal failure event is a historical event in which a certain mine car abnormal failure has occurred, so a mine car abnormal failure type representing the certain mine car abnormal failure will be recorded. The unit time period can be specifically the time required for the mine car to travel from the mining point to the unloading point / from the unloading point to the mining point, or the time required for a round trip (for example, 15 minutes), so that the maintenance reminder action can be triggered in time before going to the unloading point, the mining point, or a round trip. The next closest unit time period refers to a unit time period before the most recent unit time period. For example, if the occurrence time of a mine car abnormal failure event is 12:00, then the most recent unit time period before the occurrence of the mine car abnormal failure event is from 11:45 to 12:00, and the second most recent unit time period before the occurrence of the mine car abnormal failure event is from 11:30 to 11:45. Specifically, the abnormal fault types of all the abnormal fault events of the mine car include but are not limited to the type of failure to lift, the type of slow lifting, the type of failure to stop after lifting, the type of hydraulic system whistling, the type of no-load sounding, the type of failure to automatically descend, the type of engine failure, the type of steering failure, the type of brake failure and / or the type of electrical failure, etc.; the aforementioned fault types are all common fault types of existing mine cars. Specifically, the multiple first on-board sensors include but are not limited to noise sensors, vibration sensors, oil pressure sensors, temperature sensors, displacement sensors, speed sensors, inclination sensors and / or acceleration sensors, etc.; the aforementioned sensors are used to collect multiple corresponding sensor values arranged in sequence in time series in real time as corresponding monitoring time series data, for example, the temperature sensor can collect the temperature value within 15 minutes and at every minute; the arrangement positions of the aforementioned sensors can be conventionally designed according to the collection needs, for example, the oil pressure sensor can be arranged in the oil circuit of the mine car, the noise sensor or the vibration sensor can be arranged on or near the engine of the mine car, etc.; the aforementioned sensors can be implemented using corresponding existing sensor products. In addition, the historical monitoring time series data can be obtained based on the regular records of the real-time collection results of the first vehicle-mounted sensor.
[0053] In the step S1, considering that the historical monitoring time series data contains some dirty data, some empty value data, some abnormal value data and / or the collection time of different sensors are not aligned, in order to avoid these situations from affecting the accuracy of subsequent data analysis results, preferably, the historical monitoring time series data of all abnormal failure events of mining cars and collected by multiple first vehicle-mounted sensors within a nearly unit time period before the corresponding event occurs are obtained, including but not limited to the following steps S11 to S12.
[0054] S11. Acquire original monitoring time series data of all abnormal failure events of mining vehicles collected by a plurality of first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the plurality of first vehicle-mounted sensors are arranged on the mining vehicles to which the corresponding events belong.
[0055] S12. For each event among all the abnormal failure events of the mine cars and each sensor among the multiple first on-board sensors, apply a data cleaning method, a null value filling method, an abnormal value exclusion method and / or an acquisition time alignment method to process the corresponding original monitoring time series data to obtain the corresponding historical monitoring time series data.
[0056] In the step S12, for example, the data cleaning method, the null value filling method, the abnormal value exclusion method and / or the acquisition time alignment method are applied to process the original monitoring time series data collected by a first on-board sensor in a unit time period before an abnormal failure event of a mine car occurs, and obtain the corresponding historical monitoring time series data. The data cleaning method, the null value filling method and the abnormal value exclusion method are used to ensure that there is no dirty data, null value data and abnormal value data in the historical monitoring time series data obtained after processing, and the specific processing process can be conventionally derived using existing corresponding technical means. The acquisition time alignment method is used to ensure that the acquisition times of different sensors in the historical monitoring time series data obtained after processing are aligned, for example, all sensors collect corresponding sensor values at the end of the 1st minute, the end of the 2nd minute, ..., the end of the 14th minute and the end of the 15th minute in the unit time period. Considering the characteristic that adjacent collection moments of different sensors have equal time intervals (because the sensor values are collected periodically, for example, they are collected every 1 minute), in order to better maintain the consistency of the historical monitoring time series data of all sensors in the collection timing, preferably, for each event in the abnormal failure events of all mine cars and each sensor in the multiple first vehicle-mounted sensors, the corresponding original monitoring time series data is processed by the collection time alignment method to obtain the corresponding historical monitoring time series data, including but not limited to the following steps: for each event in the abnormal failure events of all mine cars and each sensor in the multiple first vehicle-mounted sensors, according to a plurality of unified collection moments in a preset unit time period, the collection moments in the corresponding original monitoring time series data are overall shifted, so that each shifted collection moment corresponds to each moment in the multiple unified collection moments in the unit time period, and for each shifted collection moment, the corresponding historical monitoring time series data is calculated according to the following formula :
[0057]
[0058] In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The historical monitoring value corresponding to the collection time after the translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, , , and Respectively represent the original monitoring time series data and One-to-one correspondence between adjacent acquisition moments The original monitoring values are solved by the least square method and the coefficients of the local cubic polynomial used to implement cubic spline interpolation. represents a positive integer greater than or equal to 4, The adjacent recording time is the distance from the The closest previous acquisition time after translation A collection moment.
[0059] For example, if the multiple unified collection times within the preset unit time period are the end time of the 1st minute, the end time of the 2nd minute, ..., the end time of the 14th minute and the end time of the 15th minute within the unit time period, and the collection times in the original monitoring time series data collected by a first on-board sensor within the nearly unit time period before an abnormal failure event of a mine car occurs are 10:00:21, 10:01:21, ..., 10:13:21 and 10:14:21, then after the collection times in the original monitoring time series data are shifted backward by 39 seconds as a whole, the collection times after the shift can be 10:01, 10:02, ..., 10:14 and 10:15, and correspond one by one to and align with the end time of the 1st minute, the end time of the 2nd minute, ..., the end time of the 14th minute and the end time of the 15th minute within the unit time period.
[0060] S2. Based on the historical monitoring time series data of all the mine car abnormal failure events, a K-means clustering algorithm is applied to obtain a plurality of cluster centers, wherein the total number of centers of the plurality of cluster centers is equal to the total number of mine car abnormal failure types of all the mine car abnormal failure events.
[0061] In step S2, the K-means clustering algorithm is an iterative clustering analysis algorithm, and its steps are generally: pre-dividing the data into K groups, and randomly selecting K objects as the initial cluster centers, and then calculating the distance between each object and each seed cluster center, assigning each object to the cluster center closest to it, and continuously updating the iterative cluster center until the error sum (i.e., the criterion function) converges to obtain the clustering result. Considering that the existing K-means clustering algorithm will change the compactness and discreteness of the clusters due to the influence of extreme values during the clustering process, reducing the accuracy of the entire clustering result, preferably, according to the historical monitoring time series data of all abnormal fault events of the mine car, the K-means clustering algorithm is applied to cluster to obtain multiple cluster centers, including but not limited to the following steps S21 to S25.
[0062] S21. For each event among all the mine car abnormal failure events, the corresponding Euler distance with each other event is calculated according to the corresponding historical monitoring time series data and the historical monitoring time series data of each other event among all the mine car abnormal failure events, and the number of distances whose corresponding Euler distance is less than a preset distance threshold is statistically obtained.
[0063] In step S21, the specific calculation formula of the Euler distance is an existing formula. For example, for a certain event, if there are 999 other events, 999 Euler distances will be calculated, and if 81 of the 999 Euler distances are less than the distance threshold, then the number of distances of the certain event whose Euler distance is less than the preset distance threshold is 81. The aforementioned number of distances positively reflects the number of other events located around the certain event in the multidimensional space, that is, it positively reflects the size of the neighbor density corresponding to the certain event.
[0064] S22. According to the number of distances of each event and the distribution position of each event in the multidimensional space, determine a plurality of events with maximum number of distances among all the abnormal failure events of the mining vehicles, wherein each dimension in the multidimensional space corresponds one-to-one to each sensor in the plurality of first vehicle-mounted sensors.
[0065] In step S22, the number of distances of each event among the multiple events is greater than the number of distances of adjacent events in each dimension, which can be similar to the method of finding extreme values through partial derivatives of multivariate functions. According to the number of distances of each event and the distribution position of each event in the multidimensional space, the multiple events with the maximum number of distances are determined.
[0066] S23. Arrange the multiple events in descending order according to the maximum values of the number of distances to obtain an event sequence.
[0067] S24. Select the previous events, and the previous The historical monitoring time series data of each event is used as Initial cluster centers, where Represents the total number of minecart abnormal failure types for all minecart abnormal failure events.
[0068] S25. Based on the Initial cluster centers are selected, and multiple cluster centers are obtained by clustering using other steps of the K-means clustering algorithm that are located after the initial cluster center selection step.
[0069] Therefore, based on the above steps S21 to S25, the sample distribution information can be used to select multiple points with the highest density as the initial clustering centers, so as to effectively solve the problem of clustering falling into local optimality due to human interference or extreme value influence, so that the final clustering result can meet the conditions that the similarity of the same cluster is the highest and the similarity of different clusters is the lowest, thereby ensuring the stability of the clustering result.
[0070] S3. For each event among all the abnormal failure events of the mining vehicles, the corresponding Euler distances to the multiple cluster centers are calculated according to the multiple cluster centers and the corresponding historical monitoring time series data, and the corresponding event is taken as an event belonging to a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance.
[0071] In the step S3, for example, if the Euler distances of the event to the multiple cluster centers are calculated based on the historical monitoring time series data of the multiple cluster centers, the Euler distances to the event are: the Euler distance to cluster center A is 100 unit distances, the Euler distance to cluster center B is 87 unit distances; the Euler distance to cluster center C is 34 unit distances, the Euler distance to cluster center D is 73 unit distances, and the Euler distance to cluster center E is 66 unit distances, then the event can be regarded as the belonging event of cluster center C.
[0072] S4. For each center among the multiple cluster centers, the occurrence frequency of various types of mine car abnormal failure types in all corresponding belonging events is counted, and a certain type among the mine car abnormal failure types in all corresponding belonging events and corresponding to the highest frequency is used as the corresponding mine car abnormal failure type.
[0073] In the step S4, for example, if among the mine car abnormal fault types of all the belonging events of the cluster center C, the mine car abnormal fault type A appears 88 times (that is, among all the belonging events of the cluster center C, there are 88 events with the mine car abnormal fault type A), the mine car abnormal fault type B appears 45 times (that is, among all the belonging events of the cluster center C, there are 45 events with the mine car abnormal fault type B), the mine car abnormal fault type C appears 12 times (that is, among all the belonging events of the cluster center C, there are 12 events with the mine car abnormal fault type C) and the mine car abnormal fault type D appears 3 times (that is, among all the belonging events of the cluster center C, there are 3 events with the mine car abnormal fault type D), then the mine car abnormal fault type A can be used as the mine car abnormal fault type of the cluster center C.
[0074] After step S4, consider the situation where there are at least two centers with the same type of mine car abnormal failure type, for example, the mine car abnormal failure type of cluster center C and cluster center E are both mine car abnormal failure type A. In order to avoid this situation, ensure that the multiple cluster centers have different mine car abnormal failure types. Preferably, after obtaining the mine car abnormal failure type of each center, the method also includes but is not limited to the following steps S41 to S42.
[0075] S41. If it is found that there are at least two centers with the same type of abnormal fault type of the mine cart, then the at least two centers are sorted in order from high to low according to the frequency of occurrence of the same type to obtain a center queue, and the same type continues to be the abnormal fault type of the mine cart of the first center in the center queue.
[0076] In the step S41, for example, if the mine cart abnormal failure types of cluster center A, cluster center B and cluster center D are all type A, and in the mine cart abnormal failure types of all events belonging to cluster center A, type A appears 82 times, in the mine cart abnormal failure types of all events belonging to cluster center B, type A appears 51 times, and in the mine cart abnormal failure types of all events belonging to cluster center D, type A appears 62 times, then the obtained center queue is: cluster center A, cluster center D and cluster center B, and type A can be used as the mine cart abnormal failure type of cluster center A.
[0077] S42. For each non-first center in the center queue, the occurrence frequencies of various other types of mine car abnormal fault types in the corresponding all belonging events are counted respectively, and a certain other type in the mine car abnormal fault types of the corresponding all belonging events and corresponding to the highest frequency is updated as the corresponding mine car abnormal fault type, wherein the other type refers to other mine car abnormal fault types except the same type.
[0078] In step S42, for example, if among the mine cart abnormal fault types of all the belonging events of cluster center B, type A appears 51 times, type B appears 12 times, type C appears 39 times, and type D appears 48 times, then type D (i.e., relative to type A of the same type, type B, type C, and type D are the other types respectively) can be updated as the mine cart abnormal fault type of cluster center B. In addition, if after the aforementioned steps S41 to S42, a new situation occurs in which the mine cart abnormal fault type of at least two centers is the same type, the aforementioned steps S41 to S42 will continue to be avoided until each cluster center has a different mine cart abnormal fault type.
[0079] S5. For various mine car abnormal failure types among the mine car abnormal failure types of all the mine car abnormal failure events, all corresponding positive sample data and all negative sample data are imported into a machine learning model based on an artificial intelligence algorithm for model training to obtain a corresponding fault type identification model, wherein the positive sample data contains the historical monitoring time series data of the belonging events corresponding to the center used as model input items and the numerical value "1" used as model output items, and the negative sample data contains the historical monitoring time series data of the belonging events not corresponding to the center used as model input items and the numerical value "0" used as model output items.
[0080] In step S5, the artificial intelligence algorithm is a core artificial intelligence algorithm that specifically studies how computers simulate or implement human learning behavior to acquire new knowledge or skills, reorganize existing knowledge structures to continuously improve their own performance, and is the fundamental way to make computers intelligent; specifically, the artificial intelligence algorithm preferably uses a linear regression algorithm based on the Python sklearn library to quickly and accurately find the rules in the data. In this way, based on a certain amount of the positive sample data and the negative sample data, through a conventional calibration verification modeling process (specifically including a calibration process and a verification process of the model, that is, first by comparing the model simulation results with the measured data, and then adjusting the model parameters according to the comparison results, so that the simulation results are consistent with the actual process), the fault type identification model can be trained. Preferably, in the calibration verification modeling process of the fault type identification model, a Bayesian optimization algorithm based on a tree structure is used to tune the model parameters. In addition, the artificial intelligence algorithm can also be, but not limited to, an artificial intelligence algorithm based on a support vector machine, a stochastic gradient descent method, a multivariate linear regression, a multilayer perceptron, a decision tree, a back propagation neural network or a radial basis function network.
[0081] S6. Periodically obtain current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle within a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors.
[0082] In step S6, the plurality of second vehicle-mounted sensors also include but are not limited to noise sensors, vibration sensors, oil pressure sensors, temperature sensors, displacement sensors, speed sensors, inclination sensors and / or acceleration sensors, etc. In addition, the current monitoring time series data also needs to be pre-processed by applying the data cleaning method, the null value filling method, the abnormal value exclusion method and / or the acquisition time alignment method, etc., with reference to the aforementioned steps S11 to S12, so as to ensure the correctness of the subsequent prediction results.
[0083] S7. For the various types of abnormal faults in mining cars, the current monitoring time series data is imported into the corresponding fault type identification model, and the confidence level of the target mining car having an abnormal fault of the corresponding type after a unit time period is obtained by output.
[0084] S8. When it is found that the confidence level of the target mine car having an abnormal fault belonging to a certain mine car abnormal fault type after a unit time period exceeds the preset confidence threshold corresponding to the certain mine car abnormal fault type, a maintenance reminder action corresponding to the certain mine car abnormal fault type is triggered.
[0085] In the step S8, the preset confidence threshold is exemplified but not limited to 62%. The maintenance reminder action is specifically but not limited to an audible and visual alarm. In order to correct and update the confidence of the mine car abnormal fault based on the clustering result, so as to further improve the accuracy of the maintenance reminder action, preferably, when it is found that the target mine car has a mine car abnormal fault belonging to a certain mine car abnormal fault type after a unit time period, the confidence exceeds the preset confidence threshold corresponding to the certain mine car abnormal fault type, the maintenance reminder action corresponding to the certain mine car abnormal fault type is triggered, including but not limited to the following steps S81 to S82.
[0086] S81. Calculate the new confidence level of the target mine car having a mine car abnormal fault belonging to a certain mine car abnormal fault type after a unit period of time according to the following formula :
[0087]
[0088] In the formula, Indicates less than or equal to A positive integer, The total number of minecart abnormal failure types representing all minecart abnormal failure events. Indicates that according to the first The cluster centers and the current monitoring time series data are calculated and the The Euler distance of the cluster centers, represents the confidence that the target mine car will have a mine car abnormal fault belonging to the certain mine car abnormal fault type after a unit period of time, the first The abnormal fault type of the mine car in the cluster center is the abnormal fault type of the mine car. represents the maximum value function, represents the minimum value function, Represents a negative correlation function.
[0089] In step S81, the negative correlation function Examples include ,in, Represents a function variable.
[0090] S82. When it is found that the new confidence of the target mine car having an abnormal fault belonging to the certain mine car abnormal fault type after a unit time period exceeds the preset confidence threshold corresponding to the certain mine car abnormal fault type, the maintenance reminder action corresponding to the certain mine car abnormal fault type is triggered.
[0091] Therefore, based on the mine car abnormal fault detection method described in the aforementioned steps S1 to S8, a new scheme for detecting mine car abnormal faults based on clustering algorithm and artificial intelligence algorithm is provided, that is, firstly, according to the historical monitoring time series data of all mine car abnormal fault events, the K-means clustering algorithm is applied to cluster to obtain multiple cluster centers, and then for each center, the corresponding mine car abnormal fault type is determined according to the mine car abnormal fault type of all corresponding belonging events, and then for various mine car abnormal fault types, the positive and negative sample data obtained based on the corresponding center and the non-corresponding center are imported into the machine learning model based on the artificial intelligence algorithm for model training to obtain the corresponding fault type. A fault type identification model is established, and the current monitoring time series data of the target mine car is imported into the corresponding fault type identification model respectively, and the confidence level of the target mine car having an abnormal fault of the corresponding type after a unit period of time is outputted, and finally, a maintenance reminder action is triggered when it is found that the confidence level exceeds the standard. In this way, the abnormal faults of the mine car can be detected and analyzed based on the monitoring data collected by multiple sensors arranged in the oil circuit and engine and other key components of the mine car, and the maintenance reminder action can be triggered in advance according to the detection and analysis results, so that maintenance personnel can eliminate possible mechanical failures through timely maintenance work to avoid secondary economic value losses, which is convenient for practical application and promotion.
[0092] like Figure 2 As shown, the second aspect of this embodiment provides a virtual device for implementing the abnormal fault detection method of a mine car described in the first aspect, including an event data acquisition unit, a cluster center determination unit, an attribute event classification unit, a fault type classification unit, a recognition model training unit, a current data acquisition unit, a recognition model application unit and a maintenance reminder trigger unit;
[0093] The event data acquisition unit is used to acquire the mine car abnormal fault type of all mine car abnormal fault events and the historical monitoring time series data collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the multiple first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs;
[0094] The cluster center determination unit is communicatively connected to the event data acquisition unit, and is used to apply a K-means clustering algorithm to cluster the historical monitoring time series data of all the mine car abnormal failure events to obtain a plurality of cluster centers, wherein the total number of the plurality of cluster centers is equal to the total number of the mine car abnormal failure types of all the mine car abnormal failure events;
[0095] The belonging event division unit is respectively connected to the event data acquisition unit and the cluster center determination unit in communication, and is used to calculate the corresponding Euler distance to the multiple cluster centers for each event in all the abnormal fault events of the mining vehicles according to the multiple cluster centers and the corresponding historical monitoring time series data, and use the corresponding event as the belonging event of a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance;
[0096] The fault type classification unit is communicatively connected to the attributed event classification unit, and is used to count the occurrence frequencies of various types of abnormal fault types of the mine car of all corresponding attributed events for each center among the multiple cluster centers, and use a certain type of abnormal fault types of the mine car of all corresponding attributed events and corresponding to the highest frequency as the corresponding abnormal fault type of the mine car;
[0097] The identification model training unit is respectively communicatively connected to the event data acquisition unit and the fault type classification unit, and is used for importing all corresponding positive sample data and all corresponding negative sample data into a machine learning model based on an artificial intelligence algorithm for model training for various mine car abnormal fault types in the mine car abnormal fault types of all mine car abnormal fault events, so as to obtain a corresponding fault type identification model, wherein the positive sample data contains the historical monitoring time series data of the belonging event corresponding to the center used as a model input item and a value "1" used as a model output item, and the negative sample data contains the historical monitoring time series data of the belonging event not corresponding to the center used as a model input item and a value "0" used as a model output item;
[0098] The current data acquisition unit is used to periodically acquire current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle in a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors;
[0099] The identification model application unit is respectively connected to the identification model training unit and the current data acquisition unit for communication, and is used to import the current monitoring time series data into the corresponding fault type identification model for the various types of abnormal faults of the mine car, and output the confidence that the target mine car has an abnormal fault of the corresponding type after a unit time period;
[0100] The maintenance reminder trigger unit is communicatively connected to the identification model application unit, and is used to trigger the execution of a maintenance reminder action corresponding to a certain type of mine car abnormal fault when it is found that the confidence level of the mine car abnormal fault belonging to a certain type of mine car abnormal fault in the target mine car after a unit time period exceeds a preset confidence threshold corresponding to the certain type of mine car abnormal fault.
[0101] The working process, working details and technical effects of the aforementioned device provided in the second aspect of this embodiment can be referred to the mine car abnormal fault detection method described in the first aspect, and will not be repeated here.
[0102] like Figure 3 As shown, the third aspect of this embodiment provides a computer device for executing the abnormal fault detection method of the mine car as described in the first aspect, including a memory, a processor and a transceiver that are sequentially connected in communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the abnormal fault detection method of the mine car as described in the first aspect. For example, the memory may include, but is not limited to, a random access memory (Random-Access Memory, RAM), a read-only memory (Read-Only Memory, ROM), a flash memory (Flash Memory), a first-in-first-out memory (First Input First Output, FIFO) and / or a first-in-last-out memory (First Input Last Output, FILO), etc.; the processor may include, but is not limited to, a microprocessor of the STM32F105 series. In addition, the computer device may also include, but is not limited to, a power module, a display screen and other necessary components.
[0103] The working process, working details and technical effects of the aforementioned computer device provided in the third aspect of this embodiment can be referred to the mine car abnormal fault detection method described in the first aspect, and will not be repeated here.
[0104] The fourth aspect of this embodiment provides a computer-readable storage medium storing instructions including the mine car abnormal fault detection method as described in the first aspect, that is, the computer-readable storage medium stores instructions, and when the instructions are run on a computer, the mine car abnormal fault detection method as described in the first aspect is executed. The computer-readable storage medium refers to a carrier for storing data, which may include but is not limited to computer-readable storage media such as floppy disks, optical disks, hard disks, flash memories, USB flash drives, and / or memory sticks, and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
[0105] The working process, working details and technical effects of the aforementioned computer-readable storage medium provided in the fourth aspect of this embodiment can be referred to the mine car abnormal fault detection method described in the first aspect, and will not be repeated here.
[0106] A fifth aspect of this embodiment provides a computer program product, including a computer program or instructions, wherein the computer program or instructions, when executed by a computer, implements the mine car abnormal fault detection method as described in the first aspect. The computer may be a general-purpose computer, a special-purpose computer, a computer network or other programmable device.
[0107] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A method for detecting abnormal faults in a mine car, characterized in that: include: Obtaining the mine car abnormal fault types of all mine car abnormal fault events and the historical monitoring time series data collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the multiple first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs, and obtaining the historical monitoring time series data of all mine car abnormal fault events and collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, including: obtaining the original monitoring time series data of all mine car abnormal fault events and collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs; for each event in the said all mine car abnormal fault events and each sensor in the said multiple first vehicle-mounted sensors, applying a data cleaning method, a null value filling method, an abnormal value exclusion method and / or a collection time alignment method to process the corresponding original monitoring time series data, and obtaining the corresponding historical monitoring time series data; For each event in all the mine car abnormal failure events and each sensor in the multiple first vehicle-mounted sensors, the corresponding original monitoring time series data is processed by applying the collection time alignment method to obtain the corresponding historical monitoring time series data, including: for each event in all the mine car abnormal failure events and each sensor in the multiple first vehicle-mounted sensors, according to a plurality of unified collection times in a preset unit time period, the collection time in the corresponding original monitoring time series data is overall shifted, so that each shifted collection time corresponds to each time in the multiple unified collection times in the unit time period, and for each shifted collection time, the corresponding historical monitoring time series data is calculated according to the following formula : In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The historical monitoring value corresponding to the collection time after the translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, , , and Respectively represent the original monitoring time series data and One-to-one correspondence between adjacent acquisition moments The original monitoring values are solved by the least square method and the coefficients of the local cubic polynomial used to implement cubic spline interpolation. represents a positive integer greater than or equal to 4, The adjacent collection time refers to the distance from the The closest previous acquisition time after translation A collection moment; According to the historical monitoring time series data of all the mine car abnormal failure events, a K-means clustering algorithm is applied to obtain a plurality of cluster centers, wherein the total number of centers of the plurality of cluster centers is equal to the total number of mine car abnormal failure types of all the mine car abnormal failure events; For each event in all the abnormal failure events of the mining vehicles, the corresponding Euler distances to the multiple cluster centers are calculated according to the multiple cluster centers and the corresponding historical monitoring time series data, and the corresponding event is used as an event belonging to a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance; For each of the plurality of cluster centers, the occurrence frequencies of various types of abnormal fault types of all corresponding mine carts belonging to the events are counted, and a type corresponding to the highest frequency among the abnormal fault types of all corresponding mine carts belonging to the events is used as the corresponding abnormal fault type of the mine cart; After obtaining the mine car abnormal fault type of each center, the method further includes: if it is found that the mine car abnormal fault type of at least two centers is the same type, then according to the occurrence frequency of the same type, the at least two centers are sorted in order from high to low to obtain a center queue, and the same type continues to be used as the mine car abnormal fault type of the first center in the center queue; for each non-first center in the center queue, the occurrence frequency of various other types in the mine car abnormal fault type of all corresponding belonging events is counted respectively, and a certain other type in the mine car abnormal fault type of all corresponding belonging events and corresponding to the highest frequency is updated as the corresponding mine car abnormal fault type, wherein the other type refers to other mine car abnormal fault types other than the same type; For various types of mine car abnormal faults in the mine car abnormal fault types of all mine car abnormal fault events, all corresponding positive sample data and all negative sample data are imported into a machine learning model based on an artificial intelligence algorithm for model training to obtain a corresponding fault type recognition model, wherein the positive sample data contains the historical monitoring time series data of the belonging event corresponding to the center used as a model input item and a value "1" used as a model output item, and the negative sample data contains the historical monitoring time series data of the belonging event not corresponding to the center used as a model input item and a value "0" used as a model output item; Periodically acquiring current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle in a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors; For the various types of abnormal faults of the mine car, the current monitoring time series data is imported into the corresponding fault type identification model, and the confidence of the target mine car having an abnormal fault of the corresponding type after a unit time period is obtained by output; When it is found that the confidence level of the target mine car's abnormal fault belonging to a certain mine car abnormal fault type after a unit time period exceeds a preset confidence threshold corresponding to the certain mine car abnormal fault type, a maintenance reminder action corresponding to the certain mine car abnormal fault type is triggered.
2. The mine car abnormal fault detection method according to claim 1, characterized in that: The abnormal failure types of all the abnormal failure events of the mine cars include a failure type of being unable to lift, a failure type of lifting slowly, a failure type of being unable to stop after lifting, a failure type of hydraulic system howling, a failure type of making noise when no-load, a failure type of being unable to automatically descend, an engine failure type, a steering failure type, a brake failure type and / or an electrical failure type.
3. The mine car abnormal fault detection method according to claim 1, characterized in that: The plurality of first vehicle-mounted sensors or the plurality of second vehicle-mounted sensors include a noise sensor, a vibration sensor, an oil pressure sensor, a temperature sensor, a displacement sensor, a speed sensor, an inclination sensor and / or an acceleration sensor.
4. A mine car abnormal fault detection device, characterized in that: It includes an event data acquisition unit, a cluster center determination unit, an attribute event classification unit, a fault type classification unit, an identification model training unit, a current data acquisition unit, an identification model application unit and a maintenance reminder trigger unit; The event data acquisition unit is used to acquire the mine car abnormal fault type of all mine car abnormal fault events and the historical monitoring time series data collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, wherein the multiple first vehicle-mounted sensors are arranged on the mine car to which the corresponding event belongs, and acquire the historical monitoring time series data of all mine car abnormal fault events and collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs, including: acquiring the original monitoring time series data of all mine car abnormal fault events and collected by multiple first vehicle-mounted sensors within a unit time period before the corresponding event occurs; for each event in the all mine car abnormal fault events and each sensor in the multiple first vehicle-mounted sensors, applying a data cleaning method, a null value filling method, an abnormal value exclusion method and / or an acquisition time alignment method to process the corresponding original monitoring time series data to obtain the corresponding historical monitoring time series data; For each event in all the mine car abnormal failure events and each sensor in the multiple first vehicle-mounted sensors, the corresponding original monitoring time series data is processed by applying the collection time alignment method to obtain the corresponding historical monitoring time series data, including: for each event in all the mine car abnormal failure events and each sensor in the multiple first vehicle-mounted sensors, according to a plurality of unified collection times in a preset unit time period, the collection time in the corresponding original monitoring time series data is overall shifted, so that each shifted collection time corresponds to each time in the multiple unified collection times in the unit time period, and for each shifted collection time, the corresponding historical monitoring time series data is calculated according to the following formula : In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The historical monitoring value corresponding to the collection time after the translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, , , and Respectively represent the original monitoring time series data and One-to-one correspondence between adjacent acquisition moments The original monitoring values are solved by the least square method and the coefficients of the local cubic polynomial used to implement cubic spline interpolation. represents a positive integer greater than or equal to 4, The adjacent collection time refers to the distance from the The closest previous acquisition time after translation A collection moment; The cluster center determination unit is communicatively connected to the event data acquisition unit, and is used to apply a K-means clustering algorithm to cluster the historical monitoring time series data of all the mine car abnormal failure events to obtain a plurality of cluster centers, wherein the total number of the plurality of cluster centers is equal to the total number of the mine car abnormal failure types of all the mine car abnormal failure events; The belonging event division unit is respectively connected to the event data acquisition unit and the cluster center determination unit in communication, and is used to calculate the corresponding Euler distance to the multiple cluster centers for each event in all the abnormal fault events of the mining vehicles according to the multiple cluster centers and the corresponding historical monitoring time series data, and use the corresponding event as the belonging event of a cluster center among the multiple cluster centers and corresponding to the shortest Euler distance; The fault type classification unit is communicatively connected to the belonging event classification unit, and is used for counting the occurrence frequencies of various types of mine car abnormal fault types of all corresponding belonging events for each center in the multiple clustering centers, and taking a certain type of mine car abnormal fault types of all corresponding belonging events and corresponding to the highest frequency as the corresponding mine car abnormal fault type, and after obtaining the mine car abnormal fault types of each center, if it is found that the mine car abnormal fault types of at least two centers are the same type, then according to the occurrence frequencies of the same type, the at least two centers are sorted in order from high to low to obtain a center queue, and the same type continues to be used as the mine car abnormal fault type of the first center in the center queue, and then for each non-first center in the center queue, the occurrence frequencies of various other types of mine car abnormal fault types of all corresponding belonging events are counted respectively, and a certain other type of mine car abnormal fault types of all corresponding belonging events and corresponding to the highest frequency is updated as the corresponding mine car abnormal fault type, wherein the other type refers to other mine car abnormal fault types except the same type; The identification model training unit is respectively communicatively connected to the event data acquisition unit and the fault type classification unit, and is used for importing all corresponding positive sample data and all corresponding negative sample data into a machine learning model based on an artificial intelligence algorithm for model training for various mine car abnormal fault types in the mine car abnormal fault types of all mine car abnormal fault events, so as to obtain a corresponding fault type identification model, wherein the positive sample data contains the historical monitoring time series data of the belonging event corresponding to the center used as a model input item and a value "1" used as a model output item, and the negative sample data contains the historical monitoring time series data of the belonging event not corresponding to the center used as a model input item and a value "0" used as a model output item; The current data acquisition unit is used to periodically acquire current monitoring time series data collected by a plurality of second vehicle-mounted sensors arranged on the target mining vehicle in a recent unit time period, wherein the plurality of second vehicle-mounted sensors correspond one-to-one to the plurality of first vehicle-mounted sensors; The identification model application unit is respectively connected to the identification model training unit and the current data acquisition unit for communication, and is used to import the current monitoring time series data into the corresponding fault type identification model for the various types of abnormal faults of the mine car, and output the confidence that the target mine car has an abnormal fault of the corresponding type after a unit time period; The maintenance reminder trigger unit is communicatively connected to the identification model application unit, and is used to trigger the execution of a maintenance reminder action corresponding to a certain type of mine car abnormal fault when it is found that the confidence level of the mine car abnormal fault belonging to a certain type of mine car abnormal fault in the target mine car after a unit time period exceeds a preset confidence threshold corresponding to the certain type of mine car abnormal fault.
5. A computer device, characterized in that: It comprises a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the mine car abnormal fault detection method as described in any one of claims 1 to 3.
6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on the computer, the mine car abnormal fault detection method as described in any one of claims 1 to 3 is executed.
7. A computer program product comprising a computer program or instructions, characterized in that When the computer program or the instruction is executed by a computer, the method for detecting abnormal faults of a mining vehicle according to any one of claims 1 to 3 is implemented.
