Equipment failure early warning method, device, medium and system based on industrial Internet of Things

Abstract

The present application discloses a method, device, medium and system for warning of equipment failure based on industrial Internet of Things, including: collecting multiple operating data of equipment; determining the initial attenuation factor of each operating data according to the volatility of each operating data; for any operating data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data; using an exponentially weighted average algorithm to predict based on the adaptive attenuation factor of each operating data to obtain several predicted values ​​of each operating data; and warning of equipment failure according to several predicted values ​​of each operating data. The present application can set different attenuation factors according to the volatility of the data so that each operating data can obtain a more accurate predicted value.

Description

Technical Field

[0001] The present application relates to the field of equipment failure warning, and in particular to equipment failure warning methods, devices, media and systems based on industrial Internet of Things. Background Art

[0002] In the industrial IoT environment, equipment failure warning is crucial. By monitoring the performance data of the equipment in real time, the failure warning system can identify potential failures before problems occur, thereby performing preventive maintenance. Failure warning for equipment can not only reduce unexpected downtime and ensure production continuity, but also reduce maintenance costs and extend equipment life.

[0003] The current method for predicting time series usually uses the exponential weighted average method, which is based on the past data of the time series and gives a higher weight to the latest discrete data and a lower weight to the earlier discrete data. This weight value is called the decay factor. Therefore, the exponential weighted average method can respond to the latest data changes more quickly, making it particularly suitable for data analysis in dynamic environments, such as early warning of equipment failures.

[0004] When using the exponential weighted average method for equipment failure warning, the selection of the attenuation factor has a greater impact on the prediction results, but the attenuation factor is usually preset manually. If the attenuation factor is not selected properly, the accuracy of equipment failure prediction will be affected.

[0005] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0006] The main purpose of this application is to provide an equipment failure warning method, device, medium and system based on the industrial Internet of Things, aiming to solve the technical problem that when using the exponential weighted average method for equipment failure warning, the attenuation factor is usually preset manually, and when the attenuation factor is improperly selected, it will affect the accuracy of equipment failure prediction.

[0007] To achieve the above-mentioned purpose, the present application provides an equipment failure warning method based on the industrial Internet of Things, including: collecting multiple operating data of the equipment; determining the initial attenuation factor of each operating data according to the volatility of each operating data; for any operating data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data; using an exponentially weighted average algorithm to predict based on the adaptive attenuation factor of each operating data to obtain several predicted values ​​of each operating data; and performing equipment failure warning according to the several predicted values ​​of each operating data.

[0008] Optionally, determining the initial attenuation factor of each operating data according to the volatility of each operating data includes: calculating the degree of fluctuation of each operating data; and determining the initial attenuation factor of each operating data based on the numerical difference of the degree of fluctuation of each operating data.

[0009] Optionally, determining the initial attenuation factor of each operating data based on the numerical difference of the fluctuation degree of each operating data includes: using the following formula (1) to determine the initial attenuation factor of each operating data:

[0010]

[0011] In the formula, Indicates The initial attenuation factor of the running data, It represents the set of fluctuation degree of each operation data. Indicates The degree of fluctuation of the operating data, It means to find the maximum value in a set. It means to find the minimum value in a set.

[0012] Optionally, the adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain the adaptive attenuation factor of the operating data includes: calculating the noise possibility of each discrete data in the operating data; adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain the adaptive attenuation factor of the operating data.

[0013] Optionally, calculating the noise possibility of each discrete data in the operating data includes: calculating the numerical approximation of each discrete data in the operating data; calculating the positional difference of each discrete data in the operating data; and determining the noise possibility of each discrete data in the operating data based on the numerical approximation and positional difference of each discrete data in the operating data.

[0014] Optionally, the calculating of the numerical approximation of each discrete data in the running data includes: for the running data, respectively calculating the absolute value of the difference between each discrete data and other data in turn, each of the absolute values ​​of the difference constituting a difference sequence of the corresponding discrete data, using a clustering algorithm to perform clustering processing on the difference sequence of each discrete data to obtain a clustering result of each discrete data; calculating the number of values ​​contained in each cluster in the clustering result of each discrete data, and determining the difference between the maximum number and the minimum number as the numerical approximation of the corresponding discrete data.

[0015] Optionally, the calculation of the position difference of each discrete data in the running data includes: for the running data, determining the data with the smallest numerical difference with each discrete data in turn, obtaining the position associated data corresponding to the discrete data, and determining the absolute value of the difference between the position of each discrete data and the position of the corresponding position associated data as the position difference of the discrete data.

[0016] In addition, to achieve the above-mentioned purpose, the present application also provides an equipment failure warning system based on the industrial Internet of Things, including: a data acquisition module, used to collect various operating data of the equipment; an initial attenuation factor acquisition module, used to determine the initial attenuation factor of each operating data according to the volatility of each operating data; an initial attenuation factor adjustment module, used to adjust the initial attenuation factor for any operating data based on the noise possibility of each discrete data in the operating data, so as to obtain an adaptive attenuation factor of the operating data; a data prediction module, used to predict based on the adaptive attenuation factor of each operating data using an exponentially weighted average algorithm to obtain several predicted values ​​of each operating data; an equipment failure warning module, used to perform equipment failure warning based on several predicted values ​​of each operating data.

[0017] The present application also provides an equipment failure warning device based on the industrial Internet of Things, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned equipment failure warning method based on the industrial Internet of Things.

[0018] The present application also provides a computer-readable storage medium, comprising: a computer program stored therein, and when the computer program is executed by a processor, the above-mentioned equipment failure early warning method based on industrial Internet of Things is implemented.

[0019] The present application proposes a method, device, medium and system for early warning of equipment failure based on the industrial Internet of Things, which collects a variety of operating data of the equipment, wherein each operating data is used to characterize the operating status of the equipment from different dimensions, and then determines the initial attenuation factor of each operating data according to the volatility of each operating data, so that different initial attenuation factors are set for different operating data, and the attenuation factor is set to be smaller for operating data with relatively chaotic data, so that the short-term fluctuation of the operating data can be ignored and its long-term changes can be paid attention to. For operating data with large data fluctuations, the attenuation factor is set to be larger, so that more attention can be paid to its latest data, and the changes in the data can be quickly reflected, thereby achieving a relatively accurate prediction value for different operating data.

[0020] In addition, the initial attenuation factor is adjusted based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data. Considering that the operating data are mostly data collected by sensors, and thus the possibility of noise in each operating data is relatively high, and the noise points have a strong interference with the prediction, the present application further adjusts the attenuation factor of each operating data according to the characteristics of the noise, reduces the impact of noise on the predicted value, and thereby improves the reliability of equipment fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of an equipment failure early warning method based on the industrial Internet of Things according to an implementation of the present application;

[0022] Figure 2 This is a structural block diagram of an equipment failure early warning system based on the industrial Internet of Things according to an implementation mode of the present application;

[0023] Figure 3 The present invention is a schematic diagram of the structure of an industrial Internet of Things-based equipment failure warning device according to an embodiment of the present application.

[0024] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0025] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0026] In the prior art, when the exponential weighted average method is used for equipment failure early warning, the selection of the attenuation factor has a great influence on the prediction result. However, the attenuation factor is usually preset manually. If the attenuation factor is not selected properly, the accuracy of equipment failure prediction will be affected.

[0027] To solve the above problems, the present application provides an equipment failure warning method, device, medium and system based on industrial Internet of Things. The present application scheme is introduced in detail below.

[0028] Figure 1 FIG. 1 is a flow chart of an industrial IoT-based equipment failure early warning method according to an embodiment of the present application. The industrial IoT-based equipment failure early warning method can be executed by an early warning device with data processing capability. The early warning device can be, for example, an industrial IoT-based equipment failure early warning device. Figure 1 , the equipment failure early warning method based on industrial Internet of Things may include the following steps:

[0029] Step S10, collecting various operating data of the equipment.

[0030] It should be noted that since equipment failure prediction is performed based on only one type of operating data of the equipment, the accuracy is low. Therefore, this embodiment collects multiple operating data of the equipment, predicts each of the multiple operating data separately, and then issues equipment failure warning based on multiple prediction values ​​to improve the accuracy of the prediction.

[0031] In the specific implementation process, various operation data of any device are collected on the IoT platform. The various operation data may be temperature data, vibration data, pressure data, and humidity data collected by sensors installed on the device, as well as current data and voltage data of the device. Among them, the various operation data may be in a time series format and arranged by timestamp. It is understandable that the operation data may reflect the working status of the device.

[0032] Step S20: determining an initial attenuation factor of each operating data according to the volatility of each operating data.

[0033] In this exemplary embodiment, an exponentially weighted average algorithm can be used to predict equipment failures. The size of the attenuation factor in the exponentially weighted average algorithm directly affects the sensitivity of the calculation and the accuracy of the prediction results. The attenuation factor is usually pre-set manually, but since the values ​​of each operating data are different, setting the same attenuation factor for each operating data will result in low prediction accuracy. In addition, directly setting the attenuation factor in advance without considering the data characteristics will result in poor reliability of the predicted value.

[0034] Based on this, in the embodiment of the present application, the pre-set attenuation factor is improved in combination with the numerical characteristics of each operating data to obtain an initial attenuation factor that can better characterize the numerical characteristics of each operating data. It can be understood that the initial attenuation factor is a further improvement of the pre-set attenuation factor. In the embodiment of the present application, for operating data with small numerical fluctuations, a smaller attenuation factor is set (i.e., more attention is paid to historical data), thereby ignoring short-term fluctuations in the operating data of the equipment and paying attention to its long-term changes. For operating data with large numerical fluctuations, a larger attenuation factor is set (i.e., more attention is paid to the latest data), so that changes in the latest data can be quickly reflected when using the exponentially weighted average algorithm for prediction.

[0035] In one embodiment, in step S20, determining the initial attenuation factor of each operating data according to the volatility of each operating data may specifically include:

[0036] S21, calculating the fluctuation degree of each operation data;

[0037] S22. Determine an initial attenuation factor for each operating data based on the numerical difference in the fluctuation degree of each operating data.

[0038] In the specific implementation process, the variance of each running data is recorded as the fluctuation degree of each running data. It should be noted that the variance can represent the numerical difference of the sequence.

[0039] Further, the Take the running data as an example to determine the Initial attenuation factor of the running data The calculation formula is as follows:

[0040]

[0041] In the formula, It represents the set of fluctuation degree of each operation data. Indicates The degree of fluctuation of the operating data, It means to find the maximum value in a set. It means to find the minimum value in a set.

[0042] It should be noted that this embodiment determines the initial attenuation factor of each operating data based on the numerical difference in the degree of fluctuation of each operating data, so that for the operating data with a smaller numerical change, the attenuation factor is set to be smaller, so that the short-term fluctuation of the operating data can be ignored and the long-term change can be focused on, thereby more comprehensively considering the changing trend of the data and improving the prediction accuracy; and for the operating data with a larger numerical fluctuation, the attenuation factor is set to be larger, so that the prediction process pays more attention to the latest data, quickly responds to the change of the data and is not affected by the earlier data, which can also improve the prediction accuracy.

[0043] It should be further explained that, taking the temperature data of a certain device over a period of time (in °C), 42.5, 43, 43.5, 42, 43, 42.5 and the vibration data of a certain device over a period of time (in mm / s), 0.3, 0.7, 1.0, 2.0, 2.8, 2.9 as examples, the smaller attenuation factor is preset to 0.2, and the larger attenuation factor is preset to 0.8.

[0044] The first five temperature data, 42.5, 43, 43.5, 42, and 43, are used to predict the temperature data at the next moment. The temperature data at the next moment predicted by the exponential weighted average algorithm using a smaller attenuation factor is 42.81, and the temperature data at the next moment predicted by the exponential weighted average algorithm using a larger attenuation factor is 42.94. Among them, the result 42.81 predicted by using a smaller attenuation factor is less different from the actual temperature data 42.5 at the next moment.

[0045] The vibration data of the next moment is predicted using the first five data of the vibration data, 0.3, 0.7, 1.0, 2.0, and 2.8. The vibration data of the next moment is predicted to be 1.88 using a smaller attenuation factor and the exponential weighted average algorithm using a larger attenuation factor and the exponential weighted average algorithm using a larger attenuation factor. The vibration data of the next moment is predicted to be 2.64. Among them, the difference between the result 2.64 predicted using a larger attenuation factor and the actual vibration data of the next moment 2.9 is small. Therefore, for temperature data, that is, the changes between each data are small, a smaller attenuation factor needs to be set so that the predicted value is more accurate. For vibration data, that is, the changes between each data are large, a larger attenuation factor needs to be set so that the predicted value is more accurate.

[0046] Step S30: for any operation data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor of the operation data.

[0047] Among them, in this step, the early warning device can first calculate the noise possibility of each discrete data in the operating data, and then adjust the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data.

[0048] It should be noted that since the operating data are mostly collected by sensors, and the sensors are easily interfered by other factors such as electromagnetics when collecting data, the obtained data will contain noise points. When the attenuation factor corresponding to the noise point is large (that is, when the noise point is the latest data, the weight corresponding to the noise point is large), it will cause the final predicted value to have a large difference. Based on this, the embodiment of the present application first analyzes the noise approximation of each discrete data in the operating data, and then adjusts the attenuation factor according to the noise approximation to improve the prediction accuracy.

[0049] In one embodiment, calculating the noise likelihood of each discrete data in the operation data may specifically include:

[0050] S31, calculating the numerical approximation of each discrete data in the operation data;

[0051] S32, calculating the position difference of each discrete data in the operation data;

[0052] S33. Determine the noise possibility of each discrete data in the operating data based on the numerical similarity and position difference of each discrete data in the operating data.

[0053] In an exemplary embodiment, for any running data, the absolute values ​​of the differences between each discrete data and other data are first calculated in turn, and each absolute value of the difference constitutes a difference sequence of the corresponding discrete data. A clustering algorithm is used to perform clustering processing on the difference sequence of each discrete data to obtain a clustering result of each discrete data. Then, the number of values ​​contained in each cluster in the clustering result of each discrete data is calculated, and the difference between the maximum number and the minimum number is determined as the numerical approximation of the corresponding discrete data.

[0054] Specifically, taking any running data as an example, starting from the first discrete data, the absolute value of the difference between the first discrete data and other discrete data is calculated respectively to form a difference sequence of the first discrete data. The difference sequence of other discrete data is calculated in the same way. The difference sequence of each discrete data is clustered using a clustering algorithm to obtain the clustering result of each discrete data. Taking the first discrete data as an example, the number of values ​​contained in each cluster in the clustering result of the first discrete data is calculated, and the difference between the maximum number and the minimum number is recorded as the numerical approximation of the first discrete data. Among them, when the difference sequence of each discrete data is clustered using a clustering algorithm, the preset number of clusters of the clustering algorithm is the same.

[0055] It should be noted that, taking the vibration data of a certain equipment over a period of time as an example (the unit is mm / s), 0.2, 0.3, 0.2, 0.4, 1.3, 1.2, 2.5, 1.4, 1.0, at this time, the absolute value of the difference between each vibration data and other vibration data is clustered. Since the preset number of clusters of the clustering algorithm is the same, for the vibration data 0.3, the absolute value of the difference between it and other vibration data is calculated: 0.1, 0.1, 0.1, 1, 0.9, 2.2, 1.1, 0.7, which constitutes the difference sequence of the vibration data 0.3. Since the numerical difference between 2.2 and other values ​​in the difference sequence is large, the difference sequence is clustered, and the number of clusters obtained is quite different between each cluster. For the vibration data 2.5, the absolute value of the difference between it and other vibration data is calculated: 2.3, 2, 2.3, 2.1, 1.2, 1.3, 1.1, 1.5, forming a difference sequence of vibration data 2.5. Since the difference between 2.3, 2, 2.3 and 2.1 in the difference sequence is small, and the difference between 1.2, 1.3, 1.1 and 1.5 is small, the difference sequence is clustered, and the difference in the number of clusters among the obtained clusters is also small. Therefore, for possible normal discrete data, the difference in the number of values ​​between clusters is large (the clusters formed by the difference with normal data and the clusters formed by the difference with abnormal discrete data), while for possible abnormal discrete data, the difference in the number of values ​​between clusters is small. Then, according to the numerical approximation of each discrete data, the possibility that each discrete data belongs to possible abnormal discrete data, that is, noise discrete data, can be obtained.

[0056] Furthermore, for any operating data, the data with the smallest numerical difference with each discrete data is determined in turn to obtain the position-related data corresponding to the discrete data, and the absolute value of the difference between the position of each discrete data and the position of the corresponding position-related data is determined as the position difference of the discrete data.

[0057] Specifically, taking any running data as an example, starting from the first discrete data, obtain the data with the same value and the closest position as the discrete data, record it as the position associated data of the first discrete data, calculate the absolute value of the difference between the position of the first discrete data and the position associated data corresponding to the first discrete data, record it as the position difference of the first discrete data. If there is no data with the same value as a discrete data in the running data, the data with the smallest value difference with the discrete data is recorded as the position associated data of the discrete data.

[0058] It should be noted that, taking the temperature data of a certain device over a period of time as an example (in °C), 25, 28, 60, 32, 38, 40, 58, 61, 62, for temperature data 62, the corresponding position-related data is 61, so the position difference of temperature data 62 is 1, that is, the position of the temperature data with the smallest difference from temperature data 62 is close to the position of temperature data 62, which further indicates that the possibility of temperature data 62 being normal data is relatively high. For temperature data 60, the corresponding position-related data is 61, so the position difference of temperature data 60 is 6, that is, the position of the temperature data with the smallest difference from temperature data 60 is far away from the position of temperature data 60, that is, temperature data 60 has a sudden change, and although the values ​​of temperature data 60 and temperature data 62 are similar, temperature data 62 may be normal data obtained by increasing the load of the device and increasing the temperature of the device, while temperature data 60 may be noise data obtained by the interference of various factors of the temperature sensor. Therefore, it is necessary to analyze the distance between each discrete data and the discrete data with the smallest difference in value, and further determine the possibility that each discrete data is a noise point.

[0059] Furthermore, the position differences of all discrete data in the running data are linearly normalized, and the numerical approximations of all discrete data in the running data are linearly normalized.

[0060] Furthermore, the ratio of the position difference to the numerical approximation of a discrete data in the operating data is determined as the noise possibility of the discrete data in the operating data.

[0061] It should be noted that, since the greater the position difference of a discrete data, the higher the possibility that the discrete data is noise, and the greater the numerical approximation of a discrete data, the lower the possibility that the discrete data is noise, the ratio of the position difference to the numerical approximation of a discrete data in the running data is determined as the noise possibility of the discrete data in the running data.

[0062] Step S40: predicting using an exponentially weighted average algorithm based on the adaptive attenuation factor of each operating data to obtain a plurality of predicted values ​​of each operating data.

[0063] In a specific implementation process, an exponentially weighted average algorithm is used to predict future values ​​based on the adaptive attenuation factor of each operating data, and each type of operating data obtains a preset number of predicted values.

[0064] It should be noted that the traditional exponential weighted average algorithm is used to predict equipment failures, and its attenuation factor needs to be set in advance, and the fluctuation of each sequence and the noise that is prone to occur during sensor acquisition are not considered during the setting. This embodiment considers the difference in the numerical fluctuation of different operating data while considering the problem that the data collected by the sensor is prone to noise points, thereby affecting the prediction effect. It solves the problem that when using the exponential weighted average method for equipment failure warning, the attenuation factor usually needs to be manually set in advance, and the improper selection of the attenuation factor will affect the accuracy of equipment failure prediction, thereby improving the accuracy of the prediction.

[0065] Step S50: Provide equipment failure warning based on a number of predicted values ​​of each operating data.

[0066] In the specific implementation process, the early warning threshold of each operating data is preset, and several predicted values ​​of each operating data are compared with the preset early warning threshold. If the predicted value exceeds the early warning threshold, fault early warning information is automatically generated and relevant data is recorded for subsequent analysis. At the same time, an alarm notification is sent to relevant maintenance personnel to ensure that necessary maintenance measures are taken in time, thereby reducing the risk of equipment downtime and ensuring the safe and stable operation of the equipment.

[0067] Based on the above embodiments, Figure 2 FIG. 1 is a structural block diagram of an equipment failure early warning system based on the industrial Internet of Things according to an embodiment of the present application, such as Figure 2 As shown, the equipment failure early warning system based on the industrial Internet of Things may include: a data acquisition module 210, an initial attenuation factor acquisition module 220, an initial attenuation factor adjustment module 230, a data prediction module 240, and an equipment failure early warning module 250, wherein:

[0068] The data collection module 210 is used to collect various operation data of the equipment;

[0069] The initial attenuation factor acquisition module 220 is used to determine the initial attenuation factor of each operation data according to the volatility of each operation data;

[0070] The initial attenuation factor adjustment module 230 is used to adjust the initial attenuation factor for any operation data based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor for the operation data;

[0071] The data prediction module 240 is used to perform prediction based on the adaptive attenuation factor of each operation data using an exponential weighted average algorithm to obtain a number of predicted values ​​of each operation data;

[0072] The equipment failure warning module 250 is used to provide equipment failure warning according to a number of predicted values ​​of each operation data.

[0073] In an exemplary embodiment, the initial attenuation factor acquisition module 220 may also be used to calculate the fluctuation degree of each operating data; and determine the initial attenuation factor of each operating data based on the numerical difference of the fluctuation degree of each operating data.

[0074] In an exemplary embodiment, the initial attenuation factor acquisition module 220 may determine the initial attenuation factor of each operation data using the following formula (1):

[0075]

[0076] In the formula, Indicates The initial attenuation factor of the running data, It represents the set of fluctuation degree of each operation data. Indicates The degree of fluctuation of the operating data, It means to find the maximum value in a set. It means to find the minimum value in a set.

[0077] In an exemplary embodiment, the initial attenuation factor adjustment module 230 can also be used to calculate the noise possibility of each discrete data in the operating data; adjust the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data.

[0078] In an exemplary embodiment, the initial attenuation factor adjustment module 230 can also be used to calculate the numerical approximation of each discrete data in the operating data; calculate the position difference of each discrete data in the operating data; and determine the noise possibility of each discrete data in the operating data based on the numerical approximation and position difference of each discrete data in the operating data.

[0079] In an exemplary embodiment, the initial attenuation factor adjustment module 230 can also be used to calculate the absolute value of the difference between each discrete data and other data in turn for the operating data, each of the absolute values ​​of the difference constitutes a difference sequence of the corresponding discrete data, and use a clustering algorithm to cluster the difference sequence of each discrete data to obtain a clustering result of each discrete data; calculate the number of values ​​contained in each cluster in the clustering result of each discrete data, and determine the difference between the maximum number and the minimum number as the numerical approximation of the corresponding discrete data.

[0080] In an exemplary embodiment, the initial attenuation factor adjustment module 230 can also be used to determine, for the operating data, the data with the smallest numerical difference from each discrete data in turn, to obtain position-associated data corresponding to the discrete data, and determine the absolute value of the difference between the position of each discrete data and the position of the corresponding position-associated data as the position difference of the discrete data.

[0081] Those skilled in the art should understand that the division of the various modules in the embodiment is only a division of logical functions, and can be fully or partially integrated into one or more actual carriers in actual application, and these modules can be all implemented in the form of software called by the processing unit, or all implemented in the form of hardware, or implemented in the form of a combination of software and hardware. It should be noted that each module in the equipment failure warning system based on the industrial Internet of Things in this embodiment corresponds one by one to each step in the equipment failure warning method based on the industrial Internet of Things in the aforementioned embodiment. Therefore, the specific implementation method of this embodiment can refer to the implementation method of the aforementioned equipment failure warning method based on the industrial Internet of Things, which will not be repeated here.

[0082] Based on the above embodiments, Figure 3 Schematic diagram of a device failure warning device based on industrial Internet of Things according to an implementation of the present application. Figure 3As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330 and a communication bus 340, wherein the processor 310, the communication interface 320 and the memory 330 communicate with each other through the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute an equipment failure early warning method based on the industrial Internet of Things, the method comprising: collecting a variety of operation data of the equipment; determining the initial attenuation factor of each operation data according to the volatility of each operation data; for any operation data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor of the operation data; using an exponential weighted average algorithm to predict based on the adaptive attenuation factor of each operation data to obtain a number of predicted values ​​of each operation data; and performing equipment failure early warning according to the several predicted values ​​of each operation data.

[0083] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0084] On the basis of the above embodiments, on the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the equipment failure warning method based on the industrial Internet of Things provided by the above methods, the method including: collecting multiple operating data of the equipment; determining the initial attenuation factor of each operating data according to the volatility of each operating data; for any operating data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data; using an exponentially weighted average algorithm based on the adaptive attenuation factor of each operating data to perform prediction to obtain several predicted values ​​of each operating data; and performing equipment failure warning according to the several predicted values ​​of each operating data.

[0085] On the basis of the above embodiments, on another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the equipment failure warning method based on the industrial Internet of Things provided by the above methods, the method comprising: collecting multiple operating data of the equipment; determining an initial attenuation factor of each operating data according to the volatility of each operating data; for any operating data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operating data to obtain an adaptive attenuation factor of the operating data; predicting based on the adaptive attenuation factor of each operating data using an exponentially weighted average algorithm to obtain a number of predicted values ​​of each operating data; and performing equipment failure warning according to the several predicted values ​​of each operating data.

[0086] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A device failure early warning method based on industrial Internet of Things, characterized in that: include: Collect various operating data of equipment; Determine the initial attenuation factor of each operating data according to the volatility of each operating data; For any operation data, adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor of the operation data; Based on the adaptive attenuation factor of each operating data, an exponentially weighted average algorithm is used to perform prediction to obtain a number of predicted values ​​of each operating data; Provide equipment failure warning based on several predicted values ​​of each operating data; The step of adjusting the initial attenuation factor based on the noise possibility of each discrete data in the operation data to obtain the adaptive attenuation factor of the operation data includes: For the running data, the absolute values ​​of the differences between each discrete data and other data are calculated in sequence, each of which constitutes a difference sequence of the corresponding discrete data, and a clustering algorithm is used to perform clustering processing on the difference sequence of each discrete data to obtain a clustering result of each discrete data; Calculate the number of values ​​contained in each cluster in the clustering results of each discrete data, and determine the difference between the maximum number and the minimum number as the numerical approximation of the corresponding discrete data; For the operation data, the data with the smallest numerical difference from each discrete data is determined in sequence to obtain position-related data corresponding to the discrete data, and the absolute value of the difference between the position of each discrete data and the position of the corresponding position-related data is determined as the position difference of the discrete data; Determine the noise possibility of each discrete data in the operation data based on the numerical similarity and position difference of each discrete data in the operation data; The initial attenuation factor is adjusted based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor of the operation data.

2. The equipment failure early warning method based on industrial Internet of Things according to claim 1 is characterized in that: Determining the initial attenuation factor of each operating data according to the volatility of each operating data includes: Calculate the fluctuation degree of each operating data; The initial attenuation factor of each operation data is determined based on the numerical difference of the fluctuation degree of each operation data.

3. The equipment failure early warning method based on industrial Internet of Things according to claim 2 is characterized in that: The determining of the initial attenuation factor of each operating data based on the numerical difference of the fluctuation degree of each operating data includes: The initial attenuation factor for each operating data is determined using the following formula (1): In the formula, Indicates The initial attenuation factor of the running data, It represents the set of fluctuation degree of each operation data. Indicates The degree of fluctuation of the operating data, It means to find the maximum value in a set. It means to find the minimum value in a set.

4. An equipment failure early warning system based on industrial Internet of Things, characterized in that: The equipment failure early warning system based on industrial Internet of Things includes: Data acquisition module, used to collect various operating data of equipment; An initial attenuation factor acquisition module, used to determine the initial attenuation factor of each operating data according to the volatility of each operating data; An initial attenuation factor adjustment module, configured to adjust the initial attenuation factor for any operation data based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor for the operation data; A data prediction module, used to perform prediction based on the adaptive attenuation factor of each operating data using an exponential weighted average algorithm to obtain a number of predicted values ​​of each operating data; Equipment failure warning module, used to provide equipment failure warning based on several predicted values ​​of various operating data; The initial attenuation factor adjustment module is also used for: For the running data, the absolute values ​​of the differences between each discrete data and other data are calculated in sequence, each of which constitutes a difference sequence of the corresponding discrete data, and a clustering algorithm is used to perform clustering processing on the difference sequence of each discrete data to obtain a clustering result of each discrete data; Calculate the number of values ​​contained in each cluster in the clustering results of each discrete data, and determine the difference between the maximum number and the minimum number as the numerical approximation of the corresponding discrete data; For the operation data, the data with the smallest numerical difference from each discrete data is determined in sequence to obtain position-related data corresponding to the discrete data, and the absolute value of the difference between the position of each discrete data and the position of the corresponding position-related data is determined as the position difference of the discrete data; Determine the noise possibility of each discrete data in the operation data based on the numerical similarity and position difference of each discrete data in the operation data; The initial attenuation factor is adjusted based on the noise possibility of each discrete data in the operation data to obtain an adaptive attenuation factor of the operation data.

5. An equipment failure early warning device based on industrial Internet of Things, characterized in that: The equipment failure early warning device based on the industrial Internet of Things includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the equipment failure warning method based on the industrial Internet of Things as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the equipment failure early warning method based on the industrial Internet of Things as described in any one of claims 1 to 3 is implemented.

Images