Industrial equipment fault prediction algorithm based on deep learning
Through deep learning-based algorithms, combined with multiple data sources of industrial cutting machines, the environmental impact coefficient and power loss index are calculated, and the problem of low fault prediction accuracy in the existing technology is solved, achieving higher prediction accuracy and reliability.
Patent Information
- Application Number
- CN202510356511.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-25
- Publication Date
- 2025-06-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, the accuracy of fault prediction of industrial cutting machines is low and it is prone to misjudgment.
The industrial equipment fault prediction algorithm based on deep learning is adopted to collect the operating status data, environmental data and appearance data of the industrial cutting machine, calculate the environmental impact coefficient and power loss index, analyze the power loss fluctuations, and predict the possibility of failure.
It improves the accuracy of industrial cutting machine fault prediction, reduces misjudgment, and enhances the guarantee of cutting efficiency and accuracy.
Smart Images
Figure CN120178845A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fault prediction, and specifically to an industrial equipment fault prediction algorithm based on deep learning. Background Art
[0002] Industrial equipment refers to mechanical equipment and devices used in various links such as processing, manufacturing, assembly, inspection, and transportation during the industrial production process. They are the infrastructure of industrial production and play a crucial role in improving production efficiency, ensuring product quality, and reducing production costs.
[0003] Among them, industrial cutting machines, as common industrial equipment, are widely used in various fields. Common industrial cutting machines generally drive the cutting disc to rotate at high speed through a motor during operation to achieve the cutting effect. Since industrial cutting machines require high cutting quality, the faults of industrial cutting machines are generally predicted. Traditional prediction algorithms generally analyze the cutting accuracy of industrial cutting machines and the possibility of future faults based on the cutting quality of all cutting materials within a batch, so as to realize the fault prediction of industrial cutting machines.
[0004] In the prior art, the fault prediction of industrial cutting machines generally analyzes based on the cutting quality of all cutting materials within a batch. However, since the cutting quality of cutting materials is related to multiple factors and it is not the case that the cutting quality is low only when the industrial cutting machine has a fault or abnormality, predicting the fault of the industrial cutting machine according to the high or low cutting quality of the cutting materials has a low prediction accuracy and is prone to misjudgment. Summary of the Invention
[0005] The purpose of the present invention is to provide an industrial equipment fault prediction algorithm based on deep learning, and solve the following technical problems: How to improve the fault prediction accuracy of industrial cutting machines.
[0006] The purpose of the present invention can be achieved by the following technical solutions: An industrial equipment fault prediction algorithm based on deep learning, the algorithm includes the following steps: S1: Respectively collect the operation state data and environmental data of the industrial cutting machine during each cutting process through the data acquisition module; S2: Collect the appearance data of each industrial cutting machine through a camera, and extract the important features in the appearance data; S3: Calculate the environmental influence coefficient of the industrial cutting machine during each cutting process by combining the appearance data and environmental data of different industrial cutting machines, and perform a secondary assignment to the calculation result; S4: Analyze the power loss situation of the industrial cutting machine during a cutting process by combining the operating state data and the assignment result of the environmental impact coefficient in each cutting process of the industrial cutting machine; S5: Analyze the fluctuation of the power loss of the industrial cutting machine by combining the power loss situation in any cutting process of the industrial cutting machine, and further predict the possibility of the occurrence of faults in the industrial cutting machine.
[0007] Furthermore, the operating state data collected in S1 includes: The supply voltage value, operating current value, and supply resistance of the industrial cutting machine in each cutting process; The environmental data collected in S1 includes: The ambient temperature and dust content in the air of the industrial cutting machine in each cutting process; The important features of the appearance data in S2 include: The area of oil stains and the thickness of dust on the outer surface of the industrial cutting machine before cutting operations.
[0008] Furthermore, the calculation process in S3 includes: Obtain the environmental impact coefficient of the industrial cutting machine in the a-th cutting process through the formula ; ; where a is any cutting of the industrial cutting machine in a cutting operation, is the area of oil stains on the outer surface of the industrial cutting machine before cutting operations, is the preset area of oil stains, is the thickness of dust on the outer surface of the industrial cutting machine before cutting operations, is the preset thickness of dust, is the dust content in the air of the industrial cutting machine in the a-th cutting process, is the preset dust content in the air, and are weight coefficients, is the cutting machine temperature of the industrial cutting machine in the a-th cutting process, is the preset cutting machine temperature, is the standard value of, is the ambient temperature of the industrial cutting machine in the a-th cutting process, is the preset ambient temperature, is the standard value of, is a defined function. If , then let , otherwise, let .
[0009] Further, the assignment process in S3 includes: By assigning a value to the environmental impact coefficient during the ath cutting process of the industrial cutting machine a value is generated that is between 1 and 2 and increases as the environmental impact coefficient during the ath cutting process increases, which is the environmental impact weight value during the ath cutting process; Among them, the environmental impact weight value corresponding to the environmental impact coefficient during the ath cutting process of the industrial cutting machine is set as .
[0010] Further, the analysis process in S4 includes: By using the formula the power loss index during the ath cutting process of the industrial cutting machine is calculated as ; Among them, is the time point of one data acquisition at a fixed time interval, is the total number of data acquisitions during the ath cutting process of the industrial cutting machine, is the power supply voltage value at the ith time point during the ath cutting process of the industrial cutting machine, is the preset voltage value, is the standard value of, is the operating current value at the ith time point during the ath cutting process of the industrial cutting machine, is the preset operating current value, is the standard value of, is the power supply resistance at the ith time point during the ath cutting process of the industrial cutting machine, is all the average value of.
[0011] Further, the analysis process in S4 also includes: By comparing the power loss index during the ath cutting process of the industrial cutting machine with the preset power loss index threshold ; If , it is determined that the power loss during this cutting process of the industrial cutting machine is high, which means that the operating state of the industrial cutting machine during this cutting is abnormal and will affect the subsequent cutting efficiency and cutting accuracy; If , determining that the power loss of the industrial cutting machine during this cutting process is low means that the operating state of the industrial cutting machine during this cutting is normal and will not affect the subsequent cutting efficiency and cutting accuracy.
[0012] Further, the prediction process in S5 includes: During a cutting operation, when it is determined that the operating state of the industrial cutting machine is abnormal during any cutting process, it is determined that the probability of the industrial cutting machine malfunctioning in the future is high, and a fault warning is issued; When it is determined that the operating state of the industrial cutting machine is normal during each cutting process, by combining the power loss index of the industrial cutting machine during the a-th cutting process A deep learning model is established; and based on the set deep learning model The power loss dispersion coefficient of the industrial cutting machine during a cutting operation is calculated ; Among them, is the total number of cuts of the industrial cutting machine during a cutting operation, is for all average value, is the total time consumed by the industrial cutting machine during the a-th cutting process, is the preset time consumption.
[0013] Further, the prediction process in S5 also includes: By comparing the power loss dispersion coefficient of the industrial cutting machine after the a-th cutting with the preset dispersion coefficient threshold ; If , it is determined that during a cutting operation, the power loss index of the industrial cutting machine fluctuates greatly, which means that the stability of the operating state of the industrial cutting machine is poor, indicating that the industrial cutting machine is abnormal, and the probability of malfunctioning in the future is high; If , it is determined that during a cutting operation, the power loss index of the industrial cutting machine fluctuates little, which means that the stability of the operating state of the industrial cutting machine is high, indicating that the industrial cutting machine has no obvious abnormality, and the probability of malfunctioning in the future is low.
[0014] Advantages of the present invention: (1) By combining the appearance data and environmental data of different industrial cutting machines, the present invention can reflect the influence of the cutting environment on the power of the industrial cutting machine. Then, by combining the operating state data of the industrial cutting machine during each cutting process, the power loss of the industrial cutting machine is analyzed. Further, by combining the analysis results of the power loss during each cutting process of the industrial cutting machine, the stability of the operating state of the industrial cutting machine is judged. Finally, by combining this data, the fault of the industrial cutting machine can be predicted, and the diversified data sources can improve the accuracy of the analysis results, thereby improving the accuracy of the fault prediction of the industrial cutting machine and avoiding misjudgment.
[0015] (2) The present invention compares the power loss index of the industrial cutting machine during the a-th cutting process with a preset power loss index threshold . Since this data is calculated based on diversified data and has high accuracy and reliability, through this comparison method, an accurate judgment can be made on the level of power loss of the industrial cutting machine during this cutting process, and further, it can be judged whether there is an abnormality in the operating state of the industrial cutting machine during this cutting and whether it will affect the subsequent cutting efficiency and cutting accuracy, thereby providing accurate data support for the subsequent fault prediction of the industrial cutting machine to improve the accuracy of the fault prediction of the industrial cutting machine.
[0016] (3) By combining the analysis results of whether there are abnormalities in the operating state of the industrial cutting machine during each cutting process, the present invention can preliminarily judge the likelihood of future faults of the industrial cutting machine. When it is judged that there are no abnormalities in the operating state of the industrial cutting machine during each cutting process, by calculating the power loss dispersion coefficient of the industrial cutting machine during a cutting operation, the fluctuation size of the power loss index of the industrial cutting machine during a cutting operation is analyzed, thereby analyzing the working stability of the industrial cutting machine, and further predicting the fault of the industrial cutting machine based on this data, so as to improve the accuracy and reliability of the prediction results.
[0017] (4) The present invention compares the power loss dispersion coefficient of the industrial cutting machine after the a-th cutting is completed with a preset dispersion coefficient threshold . Through this comparison method, the fluctuation size of the power loss index of the industrial cutting machine during a cutting operation can be judged, and based on the fluctuation size of the power loss index, the stability of the operating state of the industrial cutting machine can be further judged, thereby realizing the analysis of the probability of future faults. And since the comparison data is calculated based on diversified data, the accuracy and reliability of the fault prediction of the industrial cutting machine can be further improved, and misjudgment can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The present invention will be further described below in conjunction with the accompanying drawings.
[0019] Figure 1 It is a flowchart of the steps of the industrial equipment fault prediction algorithm based on deep learning in the present invention. Specific embodiments
[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0021] Please refer to Figure 1 As shown, in one embodiment, the present application provides an industrial equipment fault prediction algorithm based on deep learning, and the algorithm includes the following steps: S1: The data acquisition module respectively acquires the operation state data and environmental data of the industrial cutting machine during each cutting process; S2: The appearance data of each industrial cutting machine is acquired by a camera, and the important features in the appearance data are extracted; S3: By combining the appearance data and environmental data of different industrial cutting machines, the environmental impact coefficient of the industrial cutting machine during each cutting process is calculated, and the calculation result is re-assigned; S4: By combining the operation state data of the industrial cutting machine during each cutting process and the assignment result of the environmental impact coefficient, the power loss situation of the industrial cutting machine during one cutting process is analyzed; S5: By combining the power loss situation of the industrial cutting machine during any cutting process, the fluctuation of the power loss of the industrial cutting machine is analyzed, and further the possibility of the occurrence of faults of the industrial cutting machine is predicted; Through the above technical solution, this example provides an industrial equipment fault prediction algorithm based on deep learning, including the following steps. First, the data acquisition module collects the operation status data and environmental data of the industrial cutting machine during each cutting process respectively, and uses a camera to collect the appearance data of each industrial cutting machine, and extracts the important features in the appearance data. Then, by combining the appearance data and environmental data of different industrial cutting machines, calculate the environmental impact coefficient of the industrial cutting machine during each cutting process, and perform a secondary assignment to the calculation result. And by combining the operation status data of the industrial cutting machine during each cutting process and the assignment result of the environmental impact coefficient, analyze the power loss situation of the industrial cutting machine during a cutting process. Finally, by combining the power loss situation of the industrial cutting machine during any cutting process, analyze the fluctuation of the power loss of the industrial cutting machine, and further predict the possibility of the occurrence of industrial cutting machine faults; By setting it like this, during the cutting operation, by combining the appearance data and environmental data of different industrial cutting machines, it can reflect the influence of the cutting environment on the power of the industrial cutting machine. Then, combine the operation status data of the industrial cutting machine during each cutting process to analyze the power loss of the industrial cutting machine, and further combine the analysis results of the power loss of the industrial cutting machine during each cutting process to judge the operation status stability of the industrial cutting machine. Finally, combining this data can realize the prediction of the faults of the industrial cutting machine, and the diversified data sources can improve the accuracy of the analysis results, thereby improving the fault prediction accuracy of the industrial cutting machine and avoiding misjudgment.
[0022] The operation status data collected in S1 includes: The power supply voltage value, operating current value and power supply resistance of the industrial cutting machine during each cutting process; The environmental data collected in S1 includes: The surrounding temperature and dust content in the air of the industrial cutting machine during each cutting process; The important features of the appearance data in S2 include: The area of oil stains on the outer surface and the thickness of dust on the industrial cutting machine before the cutting operation; Through the above technical solution, this example provides important features of operating state data, environmental data, and appearance data. The operating state data includes the power supply voltage value, operating current value, and power supply resistance during each cutting process of the industrial cutting machine. This data directly reflects whether there are problems with the power during the cutting process of the industrial cutting machine. The environmental data includes the surrounding temperature and dust content in the air during each cutting process of the industrial cutting machine. Since the temperature and dust content in the air will affect the working state of the industrial cutting machine, this data can also reflect the influence of environmental factors on the power of the industrial cutting machine. Finally, the important features of the appearance data include the oil stain area and dust thickness on the outer surface of the industrial cutting machine before the cutting operation. The oil stain area and dust thickness on the outer surface of the industrial cutting machine will affect the heat dissipation of the cutting machine and also affect the power of the industrial cutting machine. Therefore, by collecting diversified data, accurate data can be provided for subsequent analysis of the power loss situation of the industrial cutting machine during the cutting process, thereby ensuring the improvement of the fault prediction accuracy of the industrial cutting machine.
[0023] The calculation process in S3 includes: Through the formula Calculate the environmental impact coefficient of the industrial cutting machine during the a-th cutting process ; Among them, a is any cutting of the industrial cutting machine during a cutting operation, is the oil stain area on the outer surface of the industrial cutting machine before the cutting operation, is the preset oil stain area, the dust thickness on the outer surface of the industrial cutting machine before the cutting operation, is the preset dust thickness, is the dust content in the air during the a-th cutting process of the industrial cutting machine, is the preset dust content in the air, and are weight coefficients, set by empirical fitting, is the cutting machine temperature during the a-th cutting process of the industrial cutting machine, is the preset cutting machine temperature, is the standard value of, and the above standard value can be selected and set according to the allowable error in the empirical data, is the environmental temperature during the a-th cutting process of the industrial cutting machine, is the preset environmental temperature, is the standard value of, and the above standard value can be selected and set according to the allowable error in the empirical data, is a defined function. If , then let , otherwise, let ; Through the above technical solution, this embodiment provides the environmental impact coefficient of the industrial cutting machine during the a-th cutting process , which can be obtained by the formula . Obviously, when the dust thickness and oil stain area on the outer surface of the industrial cutting machine before cutting operation are larger, and the cutting machine temperature, environmental temperature and dust content in the air during the a-th cutting process of the industrial cutting machine are higher, then the environmental impact coefficient of the industrial cutting machine during the a-th cutting process is larger, indicating that during this cutting process, environmental factors will affect the power of the industrial cutting machine. Specifically, the higher the dust content in the air during the a-th cutting process of the industrial cutting machine, the thicker the dust layer on the outer surface of the industrial cutting machine will be increased. When the dust thickness and oil stain area on the outer surface of the industrial cutting machine before cutting operation are larger and the cutting machine temperature and environmental temperature are higher, it will accelerate the equipment aging and affect the performance of electrical components and mechanical parts, resulting in a power drop; Therefore, when the dust thickness and oil stain area on the outer surface of the industrial cutting machine before cutting operation are smaller, and the cutting machine temperature, environmental temperature and dust content in the air during the a-th cutting process of the industrial cutting machine are lower, then the environmental impact coefficient of the industrial cutting machine during the a-th cutting process is smaller, indicating that during this cutting process, environmental factors will not affect the power of the industrial cutting machine. Through this calculation method, accurate data can be provided for subsequent judgment of the power loss level of the industrial cutting machine to improve the accuracy of the judgment result.
[0024] The assignment process in S3 includes: By assigning a value to the environmental impact coefficient of the industrial cutting machine during the a-th cutting process, an environmental impact weight value during the a-th cutting process that is between 1 and 2 and increases with the increase of the environmental impact coefficient during the a-th cutting process is generated; Among them, the environmental impact weight value during the a-th cutting process corresponding to the environmental impact coefficient of the industrial cutting machine during the a-th cutting process is set as ; Through the above technical solution, this example provides a process for assigning a value to the environmental impact coefficient of the industrial cutting machine during the a-th cutting process; As an example, the value-taking standard is as follows: It should be noted that the environmental impact coefficient The larger it is, the corresponding environmental impact weight value in the a-th cutting process is larger. Then, because the environmental impact coefficient of the industrial cutting machine in the a-th cutting process reflects the impact of environmental factors on the power of the industrial cutting machine during the a-th cutting process. When this data is larger, it means that the impact of environmental factors on the power of the industrial cutting machine during the a-th cutting process is greater. Therefore, based on this situation, the environmental impact weight value in the a-th cutting process needs to increase with the increase of the environmental impact coefficient of the industrial cutting machine in the a-th cutting process to ensure that this data can truly reflect the impact degree of environmental factors on the power of the cutting machine.
[0025] The analysis process in S4 includes: Obtain the power loss index of the industrial cutting machine in the a-th cutting process through the formula ; ; Among them, is a data acquisition time point at a fixed time interval, is the total number of data acquisitions of the industrial cutting machine in the a-th cutting process, is the power supply voltage value of the industrial cutting machine at the i-th time point in the a-th cutting process, is a preset voltage value, is the standard value of, and the above standard value can be selected and set according to the allowable error in the empirical data, is the operating current value of the industrial cutting machine at the i-th time point in the a-th cutting process, is a preset operating current value, is the standard value of, and the above standard value can be selected and set according to the allowable error in the empirical data, is the power supply resistance of the industrial cutting machine at the i-th time point in the a-th cutting process, is all the average value of; Through the above technical solution, this example provides the power loss index of the industrial cutting machine in the a-th cutting process, which can be obtained through the formula . Among them, the formula can calculate the power supply resistance fluctuation value in the a-th cutting process. Obviously, when the power supply voltage of the industrial cutting machine at the i-th time point in the a-th cutting process is lower than the preset value, the difference between the operating current and the preset value is larger, and the power supply resistance fluctuation value is larger, then the power loss index of the industrial cutting machine in the a-th cutting process The larger it is, it indicates that during the ath cutting process, there is a high power loss in the industrial cutting machine, which greatly affects the cutting quality. Moreover, the high power loss means that there is an abnormality in the industrial cutting machine. On the contrary, when the supply voltage of the industrial cutting machine at the ith time point during the ath cutting process is higher than the preset value, the operating current has a smaller difference from the preset value, and the fluctuation value of the supply resistance is smaller, then the power loss index of the industrial cutting machine during the ath cutting process is smaller, indicating that during the ath cutting process, the power loss of the industrial cutting machine is small, and the impact on the cutting quality is small. The small loss means that there is no abnormality in the industrial cutting machine. Through this calculation method, the power loss of the industrial cutting machine during any cutting process can be analyzed, thereby providing accurate data for subsequent judgment of the state of the industrial cutting machine.
[0026] The analysis process in S4 also includes: By comparing the power loss index of the industrial cutting machine during the ath cutting process with the preset power loss index threshold ; If , it is determined that the power loss of the industrial cutting machine during this cutting process is high, which means that there is an abnormality in the operating state of the industrial cutting machine during this cutting, and it will affect the subsequent cutting efficiency and cutting accuracy. If , it is determined that the power loss of the industrial cutting machine during this cutting process is low, which means that there is no abnormality in the operating state of the industrial cutting machine during this cutting, and it will not affect the subsequent cutting efficiency and cutting accuracy. Through the above technical solution, in this embodiment, by comparing the power loss index of the industrial cutting machine during the ath cutting process with the preset power loss index threshold , since this data is obtained based on diversified data calculations and has high accuracy and reliability, through this comparison method, an accurate judgment can be made on the high or low power loss of the industrial cutting machine during this cutting process, and further determine whether there is an abnormality in the operating state of the industrial cutting machine during this cutting and whether it will affect the subsequent cutting efficiency and cutting accuracy, thereby providing accurate data support for the subsequent fault prediction of the industrial cutting machine to improve the accuracy of the fault prediction of the industrial cutting machine.
[0027] The prediction process in S5 includes: During a cutting operation, when it is determined that there is an abnormality in the operating state of the industrial cutting machine during any cutting process, it is determined that the probability of the industrial cutting machine having a future fault is high, and a fault warning is issued. When it is determined that the operating status during each cutting process of the industrial cutting machine is normal, a deep learning model is established by combining the power loss index during the a-th cutting process of the industrial cutting machine ; And based on the set deep learning model The power loss dispersion coefficient of the industrial cutting machine during a cutting operation is calculated ; Wherein, is the total number of cuts of the industrial cutting machine during a cutting operation, is all The average value of, is the total time consumed by the industrial cutting machine during the a-th cutting process, is the preset time consumption; Through the above technical solution, this example provides the power loss dispersion coefficient of the industrial cutting machine during a cutting operation , which can be calculated according to the set deep learning model . Through the above technical solution, by combining the analysis results of whether there are abnormalities in the operating status of the industrial cutting machine during each cutting process, it is possible to initially judge the likelihood of future failures of the industrial cutting machine. And when it is determined that the operating status during each cutting process of the industrial cutting machine is normal, by calculating the power loss dispersion coefficient of the industrial cutting machine during a cutting operation , analyze the fluctuation magnitude of the power loss index of the industrial cutting machine during a cutting operation, thereby analyzing the working stability of the industrial cutting machine, and further predicting the faults of the industrial cutting machine based on this data, so as to improve the accuracy and reliability of the prediction results.
[0028] The prediction process in S5 further includes: By comparing the power loss dispersion coefficient after the a-th cutting of the industrial cutting machine with the preset dispersion coefficient threshold ; If , it is judged that during a cutting operation, the power loss index of the industrial cutting machine fluctuates greatly, which means that the operating status of the industrial cutting machine is unstable, indicating that the industrial cutting machine has abnormalities and predicting a high probability of future failures; If , it is judged that during a cutting operation, the power loss index of the industrial cutting machine fluctuates little, which means that the operating status of the industrial cutting machine is highly stable, indicating that the industrial cutting machine has no obvious abnormalities and predicting a low probability of future failures; Through the above technical solution, in this example, by comparing the power loss dispersion coefficient Compare with a preset coefficient of variation threshold Through this comparison method, the fluctuation magnitude of the power loss index of the industrial cutting machine during a single cutting operation can be judged, and the stability of the operating state of the industrial cutting machine can be further judged based on the fluctuation magnitude of the power loss index, so as to analyze the probability of future failures. And because the comparison data is obtained based on diversified data calculations, the fault prediction accuracy and reliability of the industrial cutting machine can be further improved, and the situation of misjudgment can be avoided.
[0029] The above has described an embodiment of the present invention in detail, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present invention application should still fall within the scope covered by the patent of the present invention.
Claims
1. Industrial equipment fault prediction algorithm based on deep learning, characterized by: The algorithm comprises the following steps: S1: Collect the operation status data and environmental data of the industrial cutting machine during each cutting process through the data acquisition module; S2: Collect the appearance data of each industrial cutting machine through a camera and extract important features from the appearance data; S3: By combining the appearance data and environmental data of different industrial cutting machines, the environmental impact coefficient of the industrial cutting machine in each cutting process is calculated, and the calculation result is re-assigned; S4: By combining the operating status data of the industrial cutting machine in each cutting process and the assignment result of the environmental impact coefficient, the power loss of the industrial cutting machine in a cutting process is analyzed; S5: By combining the power loss of the industrial cutting machine during any cutting process, the fluctuation of the power loss of the industrial cutting machine is analyzed, and the possibility of failure of the industrial cutting machine is further predicted.
2. The deep learning-based industrial equipment fault prediction algorithm according to claim 1, characterized in that: The operating status data collected in S1 includes: The power supply voltage, operating current and power supply resistance of the industrial cutting machine during each cutting process; The environmental data collected in S1 includes: The ambient temperature and dust content in the air of the industrial cutting machine during each cutting process; The important features of the appearance data in S2 include: The oil stain area and dust thickness on the outer surface of the industrial cutting machine before cutting operation.
3. The industrial equipment fault prediction algorithm based on deep learning according to claim 1 is characterized in that: The calculation process in S3 includes: By formula Calculate the environmental impact coefficient of the industrial cutting machine during the ath cutting process ; Where a is any cutting operation of the industrial cutting machine in a cutting operation. It is the oil stain area on the outer surface of the industrial cutting machine before cutting. is the preset oil stain area, The dust thickness on the surface of the industrial cutting machine before cutting. is the preset dust thickness, is the dust content in the air during the a-th cutting process of the industrial cutting machine, The dust content in the air is preset. and is the weight coefficient, is the temperature of the industrial cutting machine during the a-th cutting process, is the preset cutting machine temperature, for The standard value of is the ambient temperature of the industrial cutting machine during the ath cutting process, is the preset ambient temperature, for The standard value of To define a function, if , then let , otherwise, let .
4. The deep learning-based industrial equipment fault prediction algorithm according to claim 3 is characterized in that: The assignment process in S3 includes: By analyzing the environmental impact coefficient of the industrial cutting machine during the a-th cutting process Assign a value, and the environmental impact coefficient is between 1 and 2, and the environmental impact coefficient increases with the a-th cutting process. The environmental impact weight value in the a-th cutting process increases with the increase of ; Among them, the environmental impact coefficient of the industrial cutting machine during the ath cutting process is The corresponding environmental impact weight value in the a-th cutting process is set as .
5. The deep learning-based industrial equipment fault prediction algorithm according to claim 4 is characterized in that: The analysis process in S4 includes: By formula Calculate the power loss index of the industrial cutting machine during the ath cutting process ; in, is a data collection point at a fixed time interval. is the total number of data collection times during the a-th cutting process of the industrial cutting machine, is the power supply voltage value of the industrial cutting machine at the i-th time point during the a-th cutting process, is the preset voltage value, for The standard value of is the operating current value of the industrial cutting machine at the i-th time point during the a-th cutting process, is the preset operating current value, for The standard value of is the power supply resistance of the industrial cutting machine at the i-th time point during the a-th cutting process, For all The average value of .
6. The deep learning-based industrial equipment fault prediction algorithm according to claim 5, characterized in that: The analysis process in S4 further includes: By calculating the power loss index of the industrial cutting machine during the a-th cutting process With the preset power loss index threshold Make a comparison; like , it is judged that the power loss of the industrial cutting machine during this cutting process is high, which means that the operating state of the industrial cutting machine during this cutting is abnormal, which will affect the subsequent cutting efficiency and cutting accuracy; like , it is judged that the power loss of the industrial cutting machine during this cutting process is low, which means that there is no abnormality in the operating state of the industrial cutting machine during this cutting, and it will not affect the subsequent cutting efficiency and cutting accuracy.
7. The deep learning-based industrial equipment fault prediction algorithm according to claim 6, characterized in that: The prediction process in S5 includes: In a cutting operation, when it is determined that the operating state of the industrial cutting machine is abnormal during any cutting process, it is determined that the possibility of failure of the industrial cutting machine in the future is high, and a fault warning is issued; When it is judged that there is no abnormality in the operation state of the industrial cutting machine during each cutting process, the power loss index of the industrial cutting machine during the ath cutting process is combined Building deep learning models; And according to the set deep learning model Calculate the power loss dispersion coefficient of an industrial cutting machine in a cutting operation ; in, is the total number of cuts made by the industrial cutting machine in one cutting operation, For all The average value of is the total time consumed by the industrial cutting machine during the ath cutting process, The preset time consumption.
8. The deep learning-based industrial equipment fault prediction algorithm according to claim 7, characterized in that: The prediction process in S5 further includes: By calculating the power loss dispersion coefficient of the industrial cutting machine after the ath cutting is completed The preset coefficient of dispersion threshold Make a comparison; like , it is judged that in a cutting operation, the power loss index of the industrial cutting machine fluctuates greatly, which means that the stability of the operating state of the industrial cutting machine is poor, indicating that there is an abnormality in the industrial cutting machine, and predicting that the probability of failure in the future is high; like , it is judged that in a cutting operation, the power loss index of the industrial cutting machine fluctuates little, which means that the operating state of the industrial cutting machine is highly stable, indicating that there is no obvious abnormality in the industrial cutting machine, and predicting that the probability of future failures is small.