Industrial equipment abnormal data identification method and system based on artificial intelligence

Abstract

The invention provides an industrial equipment abnormal data identification method and system based on artificial intelligence. The method comprises the steps of obtaining operation parameters and historical fault record data of different industrial equipment; according to the operation parameters and the historical fault record data, sampling frequencies of different industrial devices are calculated; according to the determined sampling frequency, operating state data, working condition data and environment data of the corresponding industrial equipment are collected; preprocessing the operation state data of the industrial equipment; according to the working condition data, the operation state data, the working condition data and the environment data of the industrial equipment are sliced and labeled, and normal fragment data and abnormal fragment data are obtained; and inputting the normal fragment data and the abnormal fragment data into a pre-constructed basic network model for training to obtain a multi-working-condition industrial equipment abnormal data identification model. According to the method, the abnormal data of the industrial equipment can be accurately and efficiently identified, and the evaluation accuracy of the abnormal condition of the operation state data of the industrial equipment is improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and system for identifying abnormal data of industrial equipment based on artificial intelligence. Background Technology

[0002] With the development of Industry 4.0, the data generated by industrial equipment is exploding. This data often contains crucial information about equipment malfunctions, but traditional methods are insufficient to effectively identify them.

[0003] Existing data-driven methods rely on a large number of labeled anomaly samples, while the probability of anomalies in high-end industrial equipment is extremely low (<0.1%), leading to severe imbalance in model training. Currently, different industrial equipment has different probabilities of failure and varying durations of failure. Real-time collection of industrial equipment operating status data generates a large amount of data for equipment that rarely experiences anomalies. This large amount of data often fails to identify anomalies, making it impossible to accurately and efficiently identify abnormal data in industrial equipment. If the operating status data of industrial equipment is collected at intervals, it will result in missed anomaly data for equipment with a high frequency of anomalies. Therefore, how to set an appropriate sampling frequency to match the probability and duration of equipment failures and achieve accurate and efficient identification of abnormal data in industrial equipment is a pressing technical problem that needs to be solved. In addition, existing technologies do not include the evaluation of anomalies in the operating status data of industrial equipment.

[0004] Therefore, the urgent technical problem to be solved is how to provide an artificial intelligence-based method and system for identifying abnormal data of industrial equipment, so as to achieve accurate and efficient identification of abnormal data of industrial equipment and improve the accuracy of assessment of abnormal conditions of industrial equipment operating status data. Summary of the Invention

[0005] The purpose of this application is to provide an artificial intelligence-based method and system for identifying abnormal data of industrial equipment, so as to achieve accurate and efficient identification of abnormal data of industrial equipment and improve the accuracy of assessment of abnormal conditions of industrial equipment operating status data.

[0006] To achieve the above objectives, as a first aspect of this application, this application provides an artificial intelligence-based method for identifying abnormal data of industrial equipment. The method includes: acquiring operating parameters and historical fault record data of different industrial equipment; calculating the sampling frequency of different industrial equipment based on the operating parameters and historical fault record data; collecting operating status data, working condition data, and environmental data of the corresponding industrial equipment according to the determined sampling frequency, and storing them in an industrial database; preprocessing the operating status data of the industrial equipment in the industrial database; slicing and labeling the operating status data, working condition data, and environmental data of the industrial equipment in the industrial database according to the working condition data to obtain normal segment data and abnormal segment data; inputting the normal segment data and abnormal segment data into a pre-constructed basic network model for training to obtain a multi-working-condition industrial equipment abnormal data identification model; and inputting the real-time collected operating status data, working condition data, and environmental data of the industrial equipment into the multi-working-condition industrial equipment abnormal data identification model for identification to obtain the abnormal data results output by the model.

[0007] The artificial intelligence-based industrial equipment abnormal data identification method described above further includes: calculating the abnormality index of individual abnormal operating status data and the deviation score of overall abnormal data based on the abnormal data results output by the model, as well as the corresponding industrial equipment's operating standard data, working condition data, and environmental data.

[0008] The above-described artificial intelligence-based industrial equipment anomaly data identification method includes, in which, based on the operating parameters and historical fault record data of the industrial equipment, the calculation of the sampling frequency of different industrial equipment includes: determining whether the industrial equipment contains a motor based on the operating parameters, and dividing the industrial equipment containing a motor and the industrial equipment not containing a motor into a first group of industrial equipment and a second group of industrial equipment, respectively; and calculating the sampling frequency of the industrial equipment using the first method and the second method for the first group of industrial equipment and the second group of industrial equipment, respectively.

[0009] The AI-based industrial equipment anomaly data identification method described above involves, based on operating condition data, slicing and labeling the operating status data, operating condition data, and environmental data of industrial equipment in the industrial database to obtain normal and abnormal segment data. This includes: obtaining operating status data, operating condition data, and environmental data of industrial equipment under normal operating conditions based on the operating condition data, adding a normal label to each data point, and using this as normal segment data; and obtaining operating status data, operating condition data, and environmental data of industrial equipment under abnormal operating conditions based on the operating condition data, adding a normal label to each data point, and using this as abnormal segment data.

[0010] The above-described AI-based method for identifying abnormal data in industrial equipment includes a first method as follows: ; in, This indicates the sampling frequency of industrial equipment containing motors; This indicates the predetermined basic sampling frequency for industrial equipment containing motors; This indicates the weighting of the impact of historical fault record data on the sampling frequency of industrial equipment. This represents the number of faults within the total sampling time T of the industrial equipment; Indicates the first industrial equipment Duration of the fault; This represents the weighting of the influence of the operating parameters of industrial equipment on the sampling frequency of the industrial equipment. Indicates the safety factor of industrial equipment; This represents the maximum value of the frequency coefficient characteristic of industrial equipment failures; This indicates the number of revolutions per minute of industrial equipment.

[0011] The above-described AI-based method for identifying abnormal data in industrial equipment includes a second method as follows: ; in, This indicates the sampling frequency of industrial equipment that does not include a motor; This indicates the predetermined base sampling frequency for industrial equipment that does not include a motor; This indicates the weighting of the impact of historical fault record data on the sampling frequency of industrial equipment. This represents the number of faults within the total sampling time T of the industrial equipment; Indicates the first industrial equipment Duration of the fault.

[0012] The above-described AI-based method for identifying abnormal data in industrial equipment includes preprocessing of the operating status data of industrial equipment in the industrial database, which includes: data cleaning; data standardization and normalization; and data alignment and resampling.

[0013] The above-described AI-based method for identifying abnormal data in industrial equipment includes the following data: vibration data, temperature data, electrical parameter data, mechanical parameter data, operating time, motor speed, and / or safety factor of the industrial equipment.

[0014] As a second aspect of this application, this application provides an artificial intelligence-based industrial equipment anomaly data identification system, which executes the aforementioned artificial intelligence-based industrial equipment anomaly data identification method. The system includes: The acquisition module is used to acquire operating parameters and historical fault records of different industrial equipment. The sampling frequency acquisition module is used to calculate the sampling frequency of different industrial equipment based on the operating parameters and historical fault record data of the industrial equipment. The data acquisition device is used to collect operating status data, working condition data and environmental data of the corresponding industrial equipment according to a determined sampling frequency, and store them in an industrial database. The first data processor is used to preprocess the operating status data of industrial equipment in the industrial database; The first data processor is also used to slice and label the operating status data, operating condition data and environmental data of industrial equipment in the industrial database based on the operating condition data to obtain normal segment data and abnormal segment data. The model training module is used to input normal and abnormal fragment data into a pre-built basic network model for training, so as to obtain a multi-condition industrial equipment abnormal data recognition model. The abnormal data identification module is used to input real-time collected operating status data, working condition data and environmental data of industrial equipment into the multi-working-condition industrial equipment abnormal data identification model for identification, and obtain the abnormal data results output by the model.

[0015] The AI-based industrial equipment anomaly data identification system described above further includes: The second data processor is used to calculate the anomaly index of individual abnormal operating status data of industrial equipment and the deviation score of overall abnormal data based on the abnormal data results output by the model, as well as the corresponding operating standard data, operating condition data and environmental data of industrial equipment.

[0016] The beneficial effects achieved by this application are as follows: (1) Based on the operating parameters and historical fault record data of industrial equipment, this application calculates the sampling frequency of different industrial equipment, sets a suitable sampling frequency for different industrial equipment, matches the probability and duration of equipment failure, and achieves accurate and efficient identification of abnormal data of industrial equipment.

[0017] (2) Based on the abnormal data results output by the model, as well as the corresponding industrial equipment operation standard data, working condition data and environmental data, this application calculates the abnormal index of a single abnormal operating status data of industrial equipment and the deviation score of the overall abnormal data, so as to realize the assessment of abnormal conditions of industrial equipment operating status data and improve the accuracy of the assessment of abnormal conditions of industrial equipment operating status data. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart illustrating an artificial intelligence-based method for identifying abnormal data in industrial equipment, as described in an embodiment of this application.

[0020] Figure 2 This is a schematic diagram of the structure of an artificial intelligence-based industrial equipment abnormal data identification system according to an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0022] Example 1

[0023] like Figure 1 As shown, this application provides a method for identifying abnormal data in industrial equipment based on artificial intelligence. The method includes: Step S1: Obtain the operating parameters and historical fault records of different industrial equipment.

[0024] The operating parameters of industrial equipment include: the motor speed and safety factor. Historical fault records include: the number of historical faults and the duration of each fault.

[0025] Step S2: Calculate the sampling frequency of different industrial equipment based on the operating parameters and historical fault record data of the industrial equipment.

[0026] Step S2 includes: Step S210: Based on the operating parameters of the industrial equipment, determine whether the industrial equipment includes a motor, and divide the industrial equipment including the motor and the industrial equipment not including the motor into the first group of industrial equipment and the second group of industrial equipment, respectively.

[0027] Step S220: For the first group of industrial equipment and the second group of industrial equipment, the sampling frequency of the industrial equipment is calculated using the first method and the second method respectively.

[0028] This application considers the probability, duration, or frequency of abnormal data occurrence in industrial equipment to determine the data acquisition frequency, thereby avoiding missed abnormal data acquisition during equipment malfunctions, accurately capturing abnormal data, and preventing information loss. For high-frequency anomalies (such as bearing impact vibration, high-frequency discharge): a high sampling frequency is required to capture transient signals and avoid missed detections. For low-frequency anomalies (such as temperature drift, slow wear): a low sampling frequency (such as 1Hz) is sufficient to meet the requirements, avoiding redundant data. This significantly reduces storage and transmission costs and minimizes the waste of computing resources.

[0029] Specifically, for the first group of industrial equipment (including industrial equipment with motors), the first method is used to calculate the sampling frequency of the industrial equipment. The first method is as follows: ; in, This indicates the sampling frequency of industrial equipment containing motors; This indicates the predetermined basic sampling frequency for industrial equipment containing motors; This represents the weighting of the impact of historical fault record data on the sampling frequency of industrial equipment; T represents the total sampling duration. This indicates the number of failures within the total sampling time T of the industrial equipment (e.g., time T = 365 days). Indicates the first industrial equipment Duration of the fault; This represents the weighting of the influence of the operating parameters of industrial equipment on the sampling frequency of the industrial equipment. Indicates the safety factor of industrial equipment; This represents the maximum value of the fault characteristic frequency coefficient of industrial equipment (provided by the manufacturer); This indicates the number of revolutions per minute of industrial equipment.

[0030] For the second group of industrial equipment (industrial equipment excluding motors), the sampling frequency of the industrial equipment is calculated using the second method, as follows: ; in, This indicates the sampling frequency of industrial equipment that does not include a motor; This indicates the predetermined base sampling frequency for industrial equipment that does not include a motor; This indicates the weighting of the impact of historical fault record data on the sampling frequency of industrial equipment. This indicates the number of failures within the total sampling time T of the industrial equipment (e.g., time T = 365 days). Indicates the first industrial equipment Duration of the fault.

[0031] Step S3: Based on the determined sampling frequency, collect the operating status data, working condition data, and environmental data of the corresponding industrial equipment and store them in the industrial database.

[0032] This involves collecting operational status data, operating condition data, and environmental data from various sensors. The operational status data reflects the physical condition of the industrial equipment. This data includes vibration data, temperature data, electrical parameter data, mechanical parameter data, and operating time. Operating condition data includes load rate, start / stop, loading, stable operation, fault alarm operation, or abnormal shutdown. Environmental data includes ambient temperature and humidity.

[0033] Vibration data: Vibration acceleration and velocity of bearings and motors, commonly collected by piezoelectric sensors; Temperature data: Temperatures of the motor windings, gearbox, and hydraulic system are collected using thermocouples and infrared sensors; Electrical parameter data: voltage, current, power factor, collected via PLC or smart meter; Mechanical parameter data: speed, pressure, flow rate, collected by encoder or flow meter.

[0034] Step S4: Preprocess the operating status data of industrial equipment in the industrial database.

[0035] The preprocessing of the operating status data of industrial equipment in the industrial database includes: Step S410: Data cleaning and processing. For a small amount of missing data, interpolation methods (linear interpolation, spline interpolation, suitable for time series data) or mean or median are used to fill the gaps. For a large amount of missing data, the data collected by the corresponding sensor is directly deleted.

[0036] Step S420: Data standardization and normalization processing.

[0037] As a specific embodiment of the present invention, data standardization adopts the following formula: ; Where x′ represents the standardized value of the data; x represents the input data; μ represents the mean of the training set; and σ represents the standard deviation of the training set.

[0038] As a specific embodiment of the present invention, data normalization is performed using the following formula: ; Where y′ represents the normalized value of the data; y represents the input data, y min Indicates the minimum limit of the input data; y max This indicates the maximum limit of the input data.

[0039] Step S430: Data alignment and resampling.

[0040] As a specific embodiment of the present invention, data alignment: the collection frequencies of multiple sensors may be different (e.g., temperature 10s / time, vibration 1s / time), and the timestamps need to be aligned according to the "highest frequency" to ensure that the data at the same moment corresponds to the state of the same industrial equipment.

[0041] As a specific embodiment of the present invention, resampling: if the amount of data is too large (e.g., data collected at 1 kHz for one month reaches TB level), the amount of computation can be reduced by downsampling (e.g., from 1 kHz to 10 Hz) while retaining key fault features (it needs to be verified that no features are lost after downsampling).

[0042] Step S5: Based on the operating condition data, the operating status data, operating condition data and environmental data of industrial equipment in the industrial database are sliced ​​and labeled to obtain normal segment data and abnormal segment data.

[0043] Step S5 includes: Step S510: Based on the operating condition data, obtain the operating status data, operating condition data and environmental data of the industrial equipment under normal operating conditions, and add a normal label to each data as normal segment data.

[0044] Step S520: Based on the operating condition data, obtain the operating status data, operating condition data and environmental data of industrial equipment under abnormal operating conditions, and add a normal label to each data point as abnormal segment data.

[0045] Step S6: Input normal segment data and abnormal segment data into the pre-built basic network model for training to obtain a multi-condition industrial equipment abnormal data recognition model.

[0046] The base network model can be a purely supervised network model or a semi-supervised network model. A purely supervised network model is, for example, a one-dimensional convolutional neural network model. A semi-supervised network model is, for example, the Deep SVDD model.

[0047] Step S7: Input the real-time collected operating status data, working condition data and environmental data of industrial equipment into the multi-condition industrial equipment abnormal data identification model for identification, and obtain the abnormal data results output by the model.

[0048] The model outputs operational status data for abnormal industrial equipment.

[0049] Step S8: Based on the abnormal data results output by the model, as well as the corresponding industrial equipment's operating standard data, working condition data, and environmental data, calculate the abnormality index of a single abnormal operating state data of the industrial equipment and the deviation score of the overall abnormal data.

[0050] Step S8 includes: Step S810: Based on the abnormal data results output by the model, as well as the corresponding industrial equipment's operating standard data, working condition data, and environmental data, calculate the abnormality index of a single abnormal operating state data of the industrial equipment.

[0051] Step S820: Based on the abnormal data results output by the model, as well as the corresponding industrial equipment's operating standard data, working condition data, and environmental data, calculate the deviation score of the overall abnormal data of the industrial equipment.

[0052] The formula for calculating the abnormality index of the h-th abnormal operating state data of industrial equipment is as follows: ; in, The abnormality index represents the h-th type of abnormal operating state data of industrial equipment; Indicates the first industrial equipment Measured values ​​of various abnormal operating status data; Indicates the current operating condition of industrial equipment. Standard values ​​for various abnormal operating status data; This indicates the total number of categories of environmental data; Indicates the first Weighting factors for environmental data; Indicates the first Measured values ​​of environmental data; Indicates the first Normal standard data for this type of environmental data; Indicates the first Environmental data deviating from normal standard data affects industrial equipment. Adjustment coefficient for the type of operating status data; if the first The measured values ​​of environmental data are higher than the normal standard data, causing the industrial equipment to be... If the value of the operating status data increases, then ;otherwise, ; Indicates the first Environmental data deviating from normal standard data affects industrial equipment. The impact value of various operational status data.

[0053] The formula for calculating the deviation score of the overall abnormal data of industrial equipment is as follows: ; in, This indicates the deviation score of overall abnormal data for industrial equipment; This indicates the number of types of abnormal operating status data for industrial equipment. The total number of categories representing the operating status data of industrial equipment; Indicates the first industrial equipment Weights of abnormal operating state data; Indicates the first industrial equipment Measured values ​​of various abnormal operating status data; Indicates the current operating condition of industrial equipment. Standard values ​​for various abnormal operating status data; This indicates the total number of categories of environmental data; Indicates the first Weighting factors for environmental data; Indicates the first Measured values ​​of environmental data; Indicates the first Normal standard data for this type of environmental data; Indicates the first Environmental data deviating from normal standard data affects industrial equipment. Adjustment coefficient for the type of operating status data; if the first The measured values ​​of environmental data are higher than the normal standard data, causing the industrial equipment to be... If the value of the operating status data increases, then ;otherwise, ; Indicates the first Environmental data deviating from normal standard data affects industrial equipment. The impact value of various operational status data.

[0054] Example 2 like Figure 2 As shown, this application provides an artificial intelligence-based industrial equipment anomaly data identification system 100. This system executes the aforementioned artificial intelligence-based industrial equipment anomaly data identification method. The system includes: Module 10 is used to acquire operating parameters and historical fault record data of different industrial equipment; The sampling frequency acquisition module 20 is used to calculate the sampling frequency of different industrial equipment based on the operating parameters and historical fault record data of the industrial equipment. The data acquisition device 30 is used to collect the operating status data, working condition data and environmental data of the corresponding industrial equipment according to the determined sampling frequency, and store them in the industrial database. The first data processor 40 is used to preprocess the operating status data of industrial equipment in the industrial database; The first data processor 40 is also used to slice and label the operating status data, operating condition data and environmental data of industrial equipment in the industrial database based on the operating condition data to obtain normal segment data and abnormal segment data. The model training module 50 is used to input normal fragment data and abnormal fragment data into a pre-built basic network model for training, so as to obtain a multi-condition industrial equipment abnormal data recognition model. The abnormal data identification module 60 is used to input the real-time collected operating status data, working condition data and environmental data of industrial equipment into the multi-working condition industrial equipment abnormal data identification model for identification, and obtain the abnormal data results output by the model.

[0055] The second data processor 70 is used to calculate the anomaly index of individual abnormal operating status data of industrial equipment and the deviation score of overall abnormal data based on the abnormal data results output by the model, as well as the corresponding operating standard data, operating condition data and environmental data of industrial equipment.

[0056] This application also provides a computer storage medium storing computer instructions, which, when invoked, execute the address mapping method for the large-capacity solid-state drive. The computer storage medium includes one or more program instructions, which are executed by a processor to provide an artificial intelligence-based method for identifying abnormal data in industrial equipment.

[0057] The embodiments disclosed in this invention provide a computer-readable storage medium storing computer program instructions. When the computer program instructions are executed on a computer, the computer performs the aforementioned method for identifying abnormal data of industrial equipment based on artificial intelligence.

[0058] This invention provides a processor for processing the above-described artificial intelligence-based method for identifying abnormal data in industrial equipment.

[0059] In this embodiment of the invention, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0060] The various methods, steps, and logic diagrams disclosed in the embodiments of this invention can be implemented or executed. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor reads information from the storage medium and, in conjunction with its hardware, completes the steps of the above methods.

[0061] The storage medium can be memory, such as volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0062] The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EEPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).

[0063] The beneficial effects achieved by this application are as follows: (1) Based on the operating parameters and historical fault record data of industrial equipment, this application calculates the sampling frequency of different industrial equipment, sets a suitable sampling frequency for different industrial equipment, matches the probability and duration of equipment failure, and achieves accurate and efficient identification of abnormal data of industrial equipment.

[0064] (2) Based on the abnormal data results output by the model, as well as the corresponding industrial equipment operation standard data, working condition data and environmental data, this application calculates the abnormal index of a single abnormal operating status data of industrial equipment and the deviation score of the overall abnormal data, so as to realize the assessment of abnormal conditions of industrial equipment operating status data and improve the accuracy of the assessment of abnormal conditions of industrial equipment operating status data.

[0065] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0066] In the description of this application, the word "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0067] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. An artificial intelligence-based industrial equipment abnormal data identification method, characterized by, The method comprises: obtaining running parameters and historical fault record data of different industrial equipment; calculating sampling frequencies of different industrial equipment according to the running parameters and the historical fault record data of the industrial equipment; collecting running state data, working condition data and environmental data of the corresponding industrial equipment according to the determined sampling frequencies and storing the data into an industrial database; preprocessing the running state data of the industrial equipment in the industrial database; slicing and labeling the running state data, the working condition data and the environmental data of the industrial equipment in the industrial database according to the working condition data to obtain normal segment data and abnormal segment data; inputting the normal segment data and the abnormal segment data into a pre-constructed basic network model for training to obtain a multi-working-condition industrial equipment abnormal data identification model; inputting the real-time collected running state data, working condition data and environmental data of the industrial equipment into the multi-working-condition industrial equipment abnormal data identification model for identification to obtain abnormal data results output by the model. 2.The AI-based industrial equipment abnormal data recognition method of claim 1, wherein The method further comprises: calculating an abnormal index of a single abnormal running state data and a deviation score of overall abnormal data of the industrial equipment according to the abnormal data results output by the model and the running standard data, the working condition data and the environmental data of the corresponding industrial equipment. 3.The AI-based industrial equipment abnormal data recognition method of claim 1, wherein, The calculation of the sampling frequencies of different industrial equipment according to the running parameters and the historical fault record data of the industrial equipment comprises: judging whether the industrial equipment contains a motor according to the running parameters of the industrial equipment, and dividing the industrial equipment containing the motor and the industrial equipment not containing the motor into a first group of industrial equipment and a second group of industrial equipment respectively; calculating the sampling frequencies of the industrial equipment by using a first method and a second method respectively for the first group of industrial equipment and the second group of industrial equipment. 4.The AI-based industrial equipment abnormal data recognition method of claim 1, wherein The slicing and labeling of the running state data, the working condition data and the environmental data of the industrial equipment in the industrial database according to the working condition data to obtain the normal segment data and the abnormal segment data comprises: obtaining the running state data, the working condition data and the environmental data of the industrial equipment under normal working conditions according to the working condition data, adding a normal label to each data as normal segment data; obtaining the running state data, the working condition data and the environmental data of the industrial equipment under abnormal working conditions according to the working condition data, adding a normal label to each data as abnormal segment data. 5.The AI-based industrial equipment abnormal data recognition method of claim 3, wherein, The first method is as follows: ； wherein, represents a sampling frequency of the industrial equipment comprising the electric machine; represents a predetermined base sampling frequency of the industrial equipment comprising the electric machine; represents an influence weight of historical failure record data on the sampling frequency of the industrial equipment; represents a number of failures within a total sampling duration T of the industrial equipment; represents a duration of the i-th failure of the industrial equipment; represents a duration of the i-th failure of the industrial equipment; represents an influence weight of an operating parameter of the industrial equipment on the sampling frequency of the industrial equipment; represents a safety factor of the industrial equipment; represents a maximum value of a failure characteristic frequency coefficient of the industrial equipment; represents a number of revolutions per minute of the industrial equipment. 6.The AI-based industrial equipment abnormal data recognition method of claim 3, wherein, The second method is as follows: ； wherein, represents a sampling frequency of the industrial equipment without the motor; represents a predetermined base sampling frequency of the industrial equipment without the motor; represents an influence weight of the historical failure record data on the sampling frequency of the industrial equipment; represents a number of failures within a total sampling duration T of the industrial equipment; represents a duration of the i-th failure of the industrial equipment; represents a duration of the i-th failure of the industrial equipment. 7.The AI-based industrial equipment abnormal data recognition method of claim 1, wherein The preprocessing of the running state data of the industrial equipment in the industrial database comprises: data cleaning processing; data standardization and normalization processing; data alignment and resampling processing. 8.The AI-based industrial equipment abnormal data recognition method of claim 1, wherein The running state data of the industrial equipment comprises: vibration data, temperature data, electrical parameter data, mechanical parameter data, running duration, speed of a motor of the industrial equipment and / or safety factor of the industrial equipment. 9.An artificial intelligence-based industrial equipment abnormal data identification system, characterized by, The system executes the method of any one of claims 1-8, and the system comprises: an obtaining module for obtaining running parameters and historical fault record data of different industrial equipment; a sampling frequency obtaining module for calculating sampling frequencies of different industrial equipment according to the running parameters and the historical fault record data of the industrial equipment; The data acquisition device is configured to acquire operation state data, working condition data and environment data corresponding to the industrial equipment according to a determined sampling frequency, and store the data into an industrial database; The first data processor is configured to pre-process the operation state data of the industrial equipment in the industrial database; The first data processor is further configured to slice and label the operation state data, the working condition data and the environment data of the industrial equipment in the industrial database according to the working condition data, to obtain normal segment data and abnormal segment data; The model training module is configured to input the normal segment data and the abnormal segment data into a pre-constructed basic network model for training, to obtain a multi-working condition industrial equipment abnormal data identification model; The abnormal data identification module is configured to input the real-time acquired operation state data, working condition data and environment data of the industrial equipment into the multi-working condition industrial equipment abnormal data identification model for identification, to obtain an abnormal data result output by the model. 10.The AI-based industrial equipment abnormal data recognition system of claim 9, characterized by, The system further comprises: The second data processor is configured to calculate an abnormal index of a single abnormal operation state data and a deviation degree score of overall abnormal data of the industrial equipment according to the abnormal data result output by the model, and operation standard data, working condition data and environment data corresponding to the industrial equipment.

Images