A method, device, equipment and storage medium for monitoring the operating state of a device

Abstract

The present invention provides a method, device, equipment and storage medium for monitoring the operating state of equipment, including: obtaining audio data of the equipment to be monitored collected by a sound collector within a preset time period, and calling an edge conversion algorithm to process the audio data to generate a frequency division feature quantity to be judged; obtaining a reference data set of the equipment to be monitored, and calculating the reference data set to generate a reference threshold; using the conversion edge algorithm to calculate the frequency division feature quantity to be judged and the reference data set to generate a maximum similarity value, the maximum similarity value is calculated by the Euclidean distance after normalization processing, and the Euclidean distance is used to measure the similarity between the frequency division feature quantity to be judged and the reference data set; comparing the maximum similarity value with the pre-saved reference threshold to generate a device state curve corresponding to the frequency division feature quantity to be judged. The aim is to solve the problems that the existing state recognition scheme has inaccurate judgment results and limitations for equipment lacking historical data.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a method, device, equipment and storage medium for monitoring the operating state of equipment. Background Art

[0002] In the field of industrial production, it is necessary to monitor the operating state of industrial equipment. Among them, the monitoring of the operating state of industrial equipment is a key technology required in the industrial field and is a prerequisite for judging whether the industrial equipment is in a normal operating state and realizing predictive maintenance. Currently, there are usually two methods for monitoring the operating state of industrial equipment on the market: one is a state recognition method based on thresholds, and the other is a state recognition method based on models.

[0003] The method of state recognition based on thresholds needs to extract features from the monitored original data and compare the feature values with the preset operating state thresholds for state judgment. Due to the singularity of threshold setting, the method of state judgment based on thresholds is difficult to make accurate judgments on the operating states of industrial equipment with large fluctuations in feature values during operation and industrial equipment with frequent feature value offsets. And setting the threshold to a lower value to achieve accurate state judgment results is likely to lead to missed reports of abnormal equipment operation.

[0004] Although the method of state recognition based on models can construct a model using the historical operating data of a large number of industrial equipment and extract key information on operation from the equipment to achieve accurate state recognition, this method has certain application limitations for industrial equipment lacking historical operating data.

[0005] In view of this, the present application is proposed. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a method, device, equipment and storage medium for monitoring the operating state of equipment, which can effectively solve the problems that the existing equipment operating state recognition solutions have inaccurate state judgment results and limitations for industrial equipment lacking historical operating data.

[0007] The present invention provides a method for monitoring the operating state of equipment, including:

[0008] Obtain the audio data of the equipment to be monitored collected by the sound collector within a preset time period, and call the edge conversion algorithm to process the audio data to generate a frequency division feature quantity to be judged;

[0009] Obtain the reference data set of the equipment to be monitored, and calculate the reference data set to generate a reference threshold;

[0010] The conversion edge algorithm is used to calculate the to-be-determined frequency division feature quantity and the reference data set to generate a maximum similarity value, where the maximum similarity value is calculated by the normalized Euclidean distance, and the Euclidean distance is used to measure the similarity between the to-be-determined frequency division feature quantity and the reference data set;

[0011] Compare the maximum similarity value with a pre-stored reference threshold to generate a device status curve corresponding to the to-be-determined frequency division feature quantity.

[0012] Preferably, before obtaining the reference data set of the device to be monitored, it further includes:

[0013] Obtain the historical production audio data of the device to be monitored collected by the sound collector;

[0014] Use the conversion edge algorithm to process the historical production audio data to generate historical frequency division feature quantities;

[0015] Use the Gaussian multi-envelope fitting algorithm to label the historical frequency division feature quantities to generate multiple sub-reference data sets in different operating states;

[0016] Randomly select a preset number of data points from each of the sub-reference data sets to generate a reference data set.

[0017] Preferably, calculating a reference threshold for the reference data set, specifically:

[0018] According to the formula and the formula Calculate the Euclidean distance of the reference data of two shutdown states of the reference data set to generate a standard value, where X is the reference data of any operating state of the reference data set, t is the to-be-determined frequency division feature quantity, and 0 is the reference data of the shutdown state of the reference data set;

[0019] According to the formula and the formula Calculate the Euclidean distance of the reference data of the operating states other than the shutdown state in the reference data set respectively, and compare it with the standard value to generate a similarity value and a generation histogram corresponding to the similarity value and the standard value;

[0020] Process the histogram to generate a reference threshold corresponding to the histogram.

[0021] Preferably, calculating a reference threshold corresponding to the histogram, specifically:

[0022] Determine the first interval from right to left in the histogram that satisfies the condition that the quantity contained in its adjacent left interval is greater than the quantity of data contained in this interval and the quantity of data contained in its adjacent right interval is greater than the quantity of data contained in this interval;

[0023] Extract the similarity value corresponding to this interval in the histogram to generate a first reference threshold.

[0024] Preferably, calculate the histogram to generate a reference threshold corresponding to the histogram, specifically:

[0025] Determine the interval in the histogram that contains the most data, and calculate the simulated health score of the histogram based on this interval to generate a first score calculation result;

[0026] Screen the first score calculation result to generate a score value at 2.5%;

[0027] Extract the similarity value of this score value in the corresponding area of the histogram to generate a second reference threshold.

[0028] Preferably, calculate the histogram to generate a reference threshold corresponding to the histogram, specifically:

[0029] Determine the interval with the most data in the histogram, and calculate the simulated health score of the histogram based on this interval to generate a second score calculation result;

[0030] Process the second score calculation result using the interquartile range method to generate a lowest boundary score;

[0031] Extract the similarity value of the lowest boundary score in the corresponding area of the histogram to generate a third reference threshold.

[0032] Preferably, calculate the histogram to generate a reference threshold corresponding to the histogram, specifically:

[0033] Process the similarity value using the interquartile range method to generate a lowest boundary similarity value;

[0034] Extract the similarity value of the lowest boundary similarity value in the corresponding area of the histogram to generate a fourth reference threshold.

[0035] The present invention also provides a device operation status monitoring device, including:

[0036] A data acquisition unit, configured to acquire the audio data of the device to be monitored collected by a sound collector within a preset time period, and call an edge conversion algorithm to process the audio data to generate a frequency division feature quantity to be judged;

[0037] A reference threshold generation unit, configured to obtain a reference data set of the device to be monitored, and calculate the reference data set to generate a reference threshold;

[0038] A similarity acquisition unit, configured to calculate the to-be-judged frequency division feature quantity and the reference data set by using a transformed edge algorithm to generate a maximum similarity value, where the maximum similarity value is calculated by an Euclidean distance after normalization processing, and the Euclidean distance is used to measure the similarity between the to-be-judged frequency division feature quantity and the reference data set;

[0039] A state curve generation unit, configured to compare the maximum similarity value with a pre-stored reference threshold to generate a device state curve corresponding to the to-be-judged frequency division feature quantity.

[0040] The present invention further provides a device operation state monitoring device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a device operation state monitoring method as described in any one of the above.

[0041] The present invention further provides a readable storage medium storing a computer program, and the computer program can be executed by a processor of a device where the storage medium is located to implement a device operation state monitoring method as described in any one of the above.

[0042] In summary, a device operation state monitoring method, device, equipment, and storage medium provided in this embodiment. The device operation state monitoring method can determine the current operation state of an industrial device and generate a corresponding device state curve by calculating and comparing the Euclidean distance between the to-be-judged frequency division feature quantity and the reference threshold by using a transformed edge algorithm. Among them, during the use of the device operation state monitoring method, no parameter adjustment is required, and only by accumulating different operation state data of a small number of industrial devices to be detected, the current operation state of the industrial device can be obtained. Thereby, the problems that the state judgment result in the existing device operation state recognition solution is inaccurate and the existing device operation state recognition solution has limitations for industrial devices lacking historical operation data are solved. Description of the Drawings

[0043] Figure 1 is a schematic flowchart of a device operation state monitoring method provided by an embodiment of the present invention.

[0044] Figure 2 is a schematic diagram of similarity of a device operation state monitoring method provided by an embodiment of the present invention.

[0045] Figure 3 is a schematic diagram of the effect of a device operation state monitoring method provided by an embodiment of the present invention.

[0046] Figure 4 It is a schematic flowchart of a method for monitoring the operating state of a device in the application scenario of acoustic monitoring of industrial equipment provided by an embodiment of the present invention.

[0047] Figure 5 It is a schematic diagram of modules of a device operating state monitoring device provided by an embodiment of the present invention. Detailed implementation manners

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention. Therefore, the detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention.

[0049] The following will make a detailed description of specific embodiments of the present invention with reference to the accompanying drawings.

[0050] Please refer to Figure 1 , a first embodiment of the present invention provides a method for monitoring the operating state of a device, including:

[0051] S101, obtaining audio data of a device to be monitored within a preset time period collected by a sound collector, and calling an edge conversion algorithm to process the audio data to generate a frequency division feature quantity to be judged;

[0052] Specifically, in this embodiment, the device to be monitored may be any industrial device, and the device to be monitored will emit sounds during the production process. The audio data generated by the device to be monitored during the production process is collected through the sound collector; wherein, the sound collector may be a collection device such as a microphone recorder. It should be noted that in other embodiments, other types of sound collectors may also be used, which are not specifically limited here, but these solutions are all within the protection scope of the present invention.

[0053] S102, obtaining a reference data set of the device to be monitored, and calculating the reference data set to generate a reference threshold;

[0054] Specifically, step S102 includes: according to the formula and the formula Calculate the Euclidean distance of the reference data of two shutdown states of the reference data set to generate a standard value, where X is the reference data of any operating state of the reference data set, t is the frequency division feature quantity to be judged, and 0 is the reference data of the shutdown state of the reference data set;

[0055] According to the formula and the formula Calculate the Euclidean distance of the reference data of the operating states other than the shutdown state in the reference data set respectively, and compare it with the standard value to generate a similarity value and a generation histogram corresponding to the similarity value and the standard value;

[0056] Process the histogram to generate a reference threshold corresponding to the histogram.

[0057] Specifically, in this embodiment, by calculating the Euclidean distance between the frequency division feature quantity and the reference data in different states of the reference data set, the reference threshold is calculated and the state is judged. The Euclidean distance measures the similarity between two frequency division feature quantities here. Due to the volatility of the frequency division feature quantity during the operation of industrial equipment, there is a certain deviation in the Euclidean distance values between the frequency division feature quantities belonging to the same state and the fixed deviation value cannot be determined. However, the Euclidean distance value between the frequency division feature quantities belonging to different operating states will be much larger than the Euclidean distance value between the frequency division feature values belonging to the same operating state. Therefore, in order to ensure the stability of calculating the similarity value, the Euclidean distance calculated according to the reference data set of the equipment shutdown state is used as the reference value, and the Euclidean distance values calculated according to the reference data sets of other operating states of the equipment are compared with the reference value respectively to obtain the similarity value. Since the order of magnitude of the frequency division feature quantities obtained during the operation of industrial equipment is large, and the order of magnitude of the Euclidean distance between the feature quantities is also large; therefore, in order to unify the dimension of the calculated Euclidean distance, the obtained reference data set is used to calculate the Euclidean distance between the frequency division feature quantities in each sub-data set, and Min-MaxScaler is trained according to the calculation results for normalization; the data after normalization processing will be used to calculate the similarity.

[0058] In this embodiment, four calculation methods for the reference threshold are provided. According to the information obtained from the histogram, the required reference threshold can be calculated according to these four methods.

[0059] Among them, the first calculation method is specifically:

[0060] Determine the interval in the histogram that satisfies the condition that the number contained in its adjacent left interval is greater than the number of data contained in this interval and the number contained in its adjacent right interval is greater than the number of data contained in this interval from right to left for the first time;

[0061] Extract the similarity value corresponding to the interval within the histogram to generate a first reference threshold.

[0062] Among them, the second calculation method is specifically as follows:

[0063] Determine the interval in the histogram that contains the most data, and calculate the simulated health score for the histogram based on this interval to generate a first score calculation result;

[0064] Screen the first score calculation result to generate a score value at the 2.5% percentile;

[0065] Extract the similarity value of the score value within the corresponding area of the histogram to generate a second reference threshold.

[0066] Among them, the third calculation method is specifically as follows:

[0067] Determine the interval with the most data in the histogram, and calculate the simulated health score for the histogram based on this interval to generate a second score calculation result;

[0068] Process the second score calculation result using the interquartile range method to generate a lowest boundary score;

[0069] Extract the similarity value of the lowest boundary score within the corresponding area of the histogram to generate a third reference threshold.

[0070] Among them, the fourth calculation method is specifically as follows:

[0071] Process the similarity value using the interquartile range method to generate a lowest boundary similarity value;

[0072] Extract the similarity value of the lowest boundary similarity value within the corresponding area of the histogram to generate a fourth reference threshold.

[0073] In this embodiment, the simulated health score calculation is a method of assigning 100 to this interval and the intervals to the right of this interval, and for each interval moved one interval to the left of this interval, decreasing by 99 points one by one, with the lowest score being 0. Among them, the above four calculation methods of the reference threshold are all methods of screening discrete points from the running frequency division characteristics. During the process of detecting and judging the running state of the device to be monitored, the smallest reference threshold can be selected from the first reference threshold, the second reference threshold, the third reference threshold, and the fourth reference threshold as the final reference threshold.

[0074] S103. Use the conversion edge algorithm to calculate the to-be-determined frequency division feature quantity and the reference data set to generate a maximum similarity value. The maximum similarity value is calculated by the normalized Euclidean distance, and the Euclidean distance is used to measure the similarity between the to-be-determined frequency division feature quantity and the reference data set.

[0075] S104. Compare the maximum similarity value with a pre-stored reference threshold to generate a device state curve corresponding to the to-be-determined frequency division feature quantity.

[0076] Please refer to Figures 2 to 4 , specifically, in this embodiment, the to-be-determined frequency division feature quantity is calculated according to the formula and the formula respectively calculate the similarity with the reference data of each state in the reference data set, and take the maximum similarity value obtained from the calculation for state judgment. To reduce the computational amount and running time of the algorithm, since the reference data set has strong representativeness for the data of each state, only 200 data are randomly selected from each subset for calculation each time the similarity is calculated. Compared with other operating states, the to-be-determined frequency division feature quantity is more similar to the operating state corresponding to the maximum similarity value. According to the formula and the formula and the above Euclidean distance calculation method, it can be known that the maximum similarity value calculated by the to-be-determined frequency division feature quantity belonging to the shutdown state and the reference data set will be less than 0, while the maximum similarity value calculated by the to-be-determined frequency division feature quantity belonging to the operating state and the reference data set will be greater than or equal to 0. Therefore, 0 is the reference value for distinguishing the shutdown state from the operating state.

[0077] In this embodiment, the maximum similarity value is compared with 0 to determine whether the frequency division feature quantity to be judged is in the running state or the shutdown state; if it is in the running state, the calculated reference threshold is then used to judge the transition state: if the similarity value calculated for a certain frequency division feature quantity to be judged is less than the reference threshold, then this frequency division feature quantity to be judged should belong to the transition state. The inventor found that during the operation of industrial equipment, the sound of industrial equipment may change greatly over time, so the fluctuation of the characteristic values obtained during the operation of this industrial equipment is relatively large. Therefore, it is far from enough to always use the reference threshold and reference data set obtained from the historical operation data at the beginning, and it is necessary to continuously update the reference threshold and reference data set. The device operation state monitoring method adopts the method of continuously updating the reference data set to handle the possible drift of the frequency division feature quantity during the monitoring of the device operation. According to the similarity calculation method described above, the shutdown state and the running state judged to be in the non-transition state are respectively added to the corresponding reference sub-data sets; among them, the size of the sub-data set after the update should be maintained at 2000 data points. And the device operation state monitoring method adopts the method of continuously updating the reference threshold to handle the situation where the reference threshold changes due to the possible drift of the frequency division feature quantity during the monitoring of the device operation. According to the above reference threshold calculation method, using the similarity value calculated during the state judgment process, the reference threshold is recalculated and updated every set period of time; among them, if a sufficient number of running states or shutdown states are not captured within this set period of time, the reference threshold is not recalculated.

[0078] In a possible embodiment of the present invention, before obtaining the reference data set of the device to be monitored, it further includes:

[0079] Obtain the historical production audio data of the device to be monitored collected by the sound collector;

[0080] Process the historical production audio data by using the conversion edge algorithm to generate historical frequency division feature quantities;

[0081] Use the Gaussian multi-envelope fitting algorithm to perform marking processing on the historical frequency division feature quantities to generate sub-reference data sets in multiple different running states;

[0082] Randomly select a preset number of data points from each of the sub-reference data sets to generate a reference data set.

[0083] Specifically, in this embodiment, by means of Gaussian fitting of multiple envelopes, the unlabeled historical frequency-divided feature quantities are divided into sub-reference data sets corresponding to different operating states of the device according to different operating states of the device; 2000 points are randomly selected from each sub-reference data set as the reference data set for this operating state for storage. Among them, when dividing the sub-reference data sets of different operating states, if a certain operating state is not captured by the Gaussian multi-envelope fitting algorithm, the frequency-divided feature quantities corresponding to this state are manually labeled, and 2000 points are randomly selected from the labeled frequency-divided feature quantities as the reference data set for this operating state for storage. Among them, most of the data in the historical frequency-divided feature quantities belong to the main operating states of industrial equipment. The transition state only occurs during the conversion between the shutdown state and the operating state and accounts for a very small part in the operating data.

[0084] In this embodiment, it is only necessary to generate and save reference data sets for the main operating states of the device, such as the shutdown state, operating state 1, operating state 2, etc. The transition state generated during the switching between the shutdown state and the operating state does not need to generate and save a reference data set. Moreover, to ensure that the subsequent monitoring of the device's operating health is not affected by the transition state, the device operating state monitoring method will separately identify all non-main operating states during the device operation and uniformly identify them into a new state.

[0085] In summary, a device operating state monitoring method, device, equipment, and storage medium provided in this embodiment. The device operating state monitoring method can determine the current operating state of the industrial device and generate a corresponding device state curve by calculating and comparing the Euclidean distance between the to-be-determined frequency-divided feature quantity and the reference threshold by using the conversion edge algorithm; among them, during the use of the device operating state monitoring method, no parameter adjustment is required. Only by accumulating the different operating state data of a small number of industrial devices to be detected, the current operating state of the industrial device can be obtained. This method has a high degree of automation in the monitoring process, saves a large amount of time and labor costs, and at the same time, can adapt to different industrial devices and the requirements of different operating forms of the same industrial device, and provides an accurate state judgment on the premise of the existence of a small amount of historical operating data.

[0086] Please refer to Figure 5 , the second embodiment of the present invention provides a device operating state monitoring device, including:

[0087] A data acquisition unit 201, configured to acquire the audio data of the device to be monitored within a preset duration collected by the sound collector, and call the edge conversion algorithm to process the audio data to generate a to-be-determined frequency-divided feature quantity;

[0088] A reference threshold generation unit 202, configured to obtain a reference data set of the device to be monitored, and calculate the reference data set to generate a reference threshold;

[0089] A similarity acquisition unit 203, configured to calculate the to-be-determined frequency division feature quantity and the reference data set by using a transformed edge algorithm to generate a maximum similarity value, where the maximum similarity value is calculated by an Euclidean distance after normalization processing, and the Euclidean distance is used to measure the similarity between the to-be-determined frequency division feature quantity and the reference data set;

[0090] A state curve generation unit 204, configured to compare the maximum similarity value with a pre-stored reference threshold to generate a device state curve corresponding to the to-be-determined frequency division feature quantity.

[0091] A third embodiment of the present invention provides a device operation state monitoring device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the device implements a device operation state monitoring method as described in any one of the above.

[0092] A fourth embodiment of the present invention provides a readable storage medium storing a computer program, and the computer program can be executed by a processor of the device where the storage medium is located to implement a device operation state monitoring method as described in any one of the above.

[0093] Exemplarily, the computer program described in the third and fourth embodiments of the present invention can be divided into one or more modules. The one or more modules are stored in the memory and executed by the processor to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in implementing a device operation state monitoring device. For example, the device described in the second embodiment of the present invention.

[0094] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the method for monitoring the operating state of a device, and uses various interfaces and circuits to connect the whole to implement various parts of the method for monitoring the operating state of a device.

[0095] The memory can be used to store the computer programs and / or modules. The processor realizes various functions of the method for monitoring the operating state of a device by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, a text conversion function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, text message data, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0096] Among them, if the implemented module is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiment methods of the present invention can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0097] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0098] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention.

Claims

1. A method for monitoring the operating state of a device, characterized in that, comprising: Obtaining audio data of the device to be monitored collected by a sound collector within a preset time period, and invoking an edge conversion algorithm to process the audio data to generate a frequency division feature quantity to be judged; Obtaining a reference data set of the device to be monitored, and calculating the reference data set to generate a reference threshold; Using a conversion edge algorithm to calculate the frequency division feature quantity to be judged and the reference data set to generate a maximum similarity value, wherein the maximum similarity value is calculated by the Euclidean distance after normalization processing, and the Euclidean distance is used to measure the similarity between the frequency division feature quantity to be judged and the reference data set; Comparing the maximum similarity value with a pre-saved reference threshold to generate a device state curve corresponding to the frequency division feature quantity to be judged; Before obtaining the reference data set of the device to be monitored, further comprising: Obtaining historical production audio data of the device to be monitored collected by a sound collector; Using a conversion edge algorithm to process the historical production audio data to generate historical frequency division feature quantities; Using a Gaussian multi-envelope fitting algorithm to perform marking processing on the historical frequency division feature quantities to generate sub-reference data sets in multiple different operating states; Randomly selecting a preset number of data points from each of the sub-reference data sets to generate a reference data set; Calculating the reference data set to generate a reference threshold, specifically: According to the formula and the formula calculate the Euclidean distance of the reference data of two shutdown states of the reference data set to generate a standard value, where X is the reference data of any operating state of the reference data set, t is the frequency division feature quantity to be judged, and 0 is the reference data of the shutdown state of the reference data set; According to the formula and the formula calculate the Euclidean distance of the reference data of the operating states other than the shutdown state in the reference dataset respectively, and compare it with the standard value to generate a similarity value and a generated histogram corresponding to the similarity value and the standard value; Processing the histogram to generate a reference threshold corresponding to the histogram.

2. The method for monitoring the operating state of a device according to claim 1, characterized in that, Calculating the histogram to generate a reference threshold corresponding to the histogram, specifically: Determining an interval in the histogram that satisfies, from right to left, that the number contained in its adjacent left interval is greater than the number of data contained in this interval and the number contained in its adjacent right interval is greater than the number of data contained in this interval; Extracting the similarity value corresponding to the interval in the histogram to generate a first reference threshold.

3. The method for monitoring the operating state of a device according to claim 1, characterized in that, Calculating the histogram to generate a reference threshold corresponding to the histogram, specifically: Determining the interval in the histogram that contains the most data, and calculating a simulated health score for the histogram based on this interval to generate a first score calculation result; Screening the first score calculation result to generate a score value at 2.5%; Extracting the similarity value corresponding to the score value in the corresponding area of the histogram to generate a second reference threshold.

4. The method for monitoring the operating state of a device according to claim 1, characterized in that, Calculating the histogram to generate a reference threshold corresponding to the histogram, specifically: Determining the interval with the most data in the histogram, and calculating a simulated health score for the histogram based on this interval to generate a second score calculation result; Using the interquartile range method to process the second score calculation result to generate a lowest boundary score; Extract the similarity value of the lowest boundary score within the corresponding region of the histogram to generate a third reference threshold.

5. The device operation status monitoring method according to claim 1, characterized in that calculating the histogram to generate a reference threshold corresponding to the histogram, specifically: processing the similarity values using the interquartile range method to generate the lowest boundary similarity value; extracting the similarity value of the lowest boundary similarity value within the corresponding region of the histogram to generate a fourth reference threshold.

6. A device operation status monitoring device for implementing the device operation status monitoring method according to claim 1, characterized in that comprising: a data acquisition unit for acquiring audio data of the device to be monitored collected by a sound collector within a preset time period, and invoking an edge transformation algorithm to process the audio data to generate a frequency division feature quantity to be judged; a reference threshold generation unit for acquiring a reference data set of the device to be monitored and calculating the reference data set to generate a reference threshold; a similarity acquisition unit for calculating the frequency division feature quantity to be judged and the reference data set using a transformed edge algorithm to generate a maximum similarity value, wherein the maximum similarity value is calculated by a normalized Euclidean distance, and the Euclidean distance is used to measure the similarity between the frequency division feature quantity to be judged and the reference data set; a state curve generation unit for comparing the maximum similarity value with a pre-stored reference threshold to generate a device state curve corresponding to the frequency division feature quantity to be judged; Before acquiring the reference data set of the device to be monitored, it further includes: acquiring historical production audio data of the device to be monitored collected by a sound collector; processing the historical production audio data using a transformed edge algorithm to generate historical frequency division feature quantities; performing a labeling process on the historical frequency division feature quantities using a Gaussian multi-envelope fitting algorithm to generate multiple sub-reference data sets in different operation states; randomly selecting a preset number of data points from each of the sub-reference data sets to generate a reference data set; calculating the reference data set to generate a reference threshold, specifically: According to the formula and the formula calculate the Euclidean distance of the reference data of two shutdown states of the reference data set to generate a standard value, where X is the reference data of any operating state of the reference data set, t is the frequency division feature quantity to be judged, and 0 is the reference data of the shutdown state of the reference data set; According to the formula and the formula Calculate the Euclidean distance of the reference data of the operating states other than the shutdown state in the reference dataset respectively, and compare it with the standard value to generate a similarity value and a generation histogram corresponding to the similarity value and the standard value; processing the histogram to generate a reference threshold corresponding to the histogram.

7. A device operation status monitoring device, characterized in that comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements a device operation status monitoring method according to any one of claims 1 to 5.

8. A readable storage medium, characterized in that storing a computer program, and the computer program can be executed by the processor of the device where the storage medium is located to implement a device operation status monitoring method according to any one of claims 1 to 5.

Images