Equipment start-stop state judgment method and device, equipment and medium

Abstract

The invention discloses an equipment start-stop state judgment method and device, equipment and a medium, and relates to the technical field of computers. Comprising the following steps: acquiring multi-dimensional data, and preprocessing the multi-dimensional data to obtain target data; the multi-dimensional data comprises mechanical data, electrical data, temperature data, flow data and pressure data; performing feature extraction operation on the target data to obtain a data feature value of the target data, and performing multi-modal data fusion operation on the data feature value to obtain target fusion data; and generating a target threshold value based on the target fusion data by using a preset dynamic threshold value algorithm, and judging the start-stop state of the target equipment based on the target threshold value. Therefore, the start-stop state of the equipment can be judged by adapting to the state change of the equipment under the condition of not depending on an additional sensor.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment and medium for determining the start / stop status of a device. Background Art

[0002] Existing technologies for determining equipment start / stop status have the following limitations: single data dimensions, reliance on single parameters such as speed and vibration, and a lack of collaborative analysis of multimodal data, resulting in limited diagnostic accuracy. This approach uses a minimum speed threshold and fluctuation range, combined with duration, to determine equipment start / stop status, addressing the problem of misjudgment in low-load, long-term operation scenarios. Its advantages include a simple algorithm and high real-time performance. However, its reliance on speed sensors makes it unsuitable for equipment without speed signals (such as dumb equipment). Furthermore, its static threshold settings cannot adapt to equipment aging or load fluctuations. The thresholds are rigid; using static thresholds or semi-dynamic adjustments (such as exponential smoothing) cannot respond in real time to equipment aging, load fluctuations, and environmental interference. This approach generates start / stop thresholds through cluster analysis of vibration signal RMS values, eliminating the need for speed sensors. Driven by non-invasive vibration data, it reduces hardware modification costs. It is well-suited for dumb equipment and complex scenarios such as machinery in mining, steel, and nonferrous metals. However, its threshold generation relies on long-term historical data accumulation, requiring manual intervention for initial setting. It also fails to consider the dynamic impact of environmental parameters (such as temperature and load).

[0003] As can be seen from the above, how to adapt to changes in equipment status and determine the equipment start and shutdown status without relying on additional sensors is an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method, apparatus, device, and medium for determining the startup and shutdown status of a device, which can adapt to changes in the device status and determine the startup and shutdown status of the device without relying on additional sensors. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a method for determining the start / stop status of a device, comprising:

[0006] Acquiring multi-dimensional data and preprocessing the multi-dimensional data to obtain target data; the multi-dimensional data includes mechanical data, electrical data, temperature data, flow data, and pressure data;

[0007] Performing a feature extraction operation on the target data to obtain data feature values of the target data, and performing a multimodal data fusion operation on the data feature values to obtain target fused data;

[0008] A preset dynamic threshold algorithm is used to generate a target threshold based on the target fusion data, and the start / stop status of the target device is determined based on the target threshold.

[0009] Optionally, preprocessing the multi-dimensional data to obtain target data includes:

[0010] The multi-dimensional data of different sampling frequencies are time-aligned, the multi-dimensional data are normalized, and abnormal data in the multi-dimensional data are eliminated to obtain target data.

[0011] Optionally, performing a feature extraction operation on the target data to obtain a data feature value of the target data includes:

[0012] Extracting vibration information characteristic values of the target data through time domain, frequency domain, and time-frequency domain analysis;

[0013] Extracting the current harmonic characteristic value of the target data by calculating the harmonic distortion rate and the fundamental amplitude offset;

[0014] The temperature change characteristic value of the target data is extracted by calculating the temperature gradient change rate.

[0015] Optionally, performing a multimodal data fusion operation on the data feature values to obtain target fused data includes:

[0016] Performing a multimodal data fusion operation on the data feature values based on a machine learning fusion method to obtain target fused data;

[0017] Alternatively, a multimodal data fusion operation is performed on the data feature values based on a weighted average method to obtain target fused data.

[0018] Optionally, the machine learning-based fusion method performs a multimodal data fusion operation on the data feature values to obtain target fused data, including:

[0019] generating a composite input vector based on the vibration information characteristic value, the current harmonic characteristic value, and the temperature change characteristic value, and training the original neural network model using the composite input vector to obtain a trained neural network model;

[0020] Weights are dynamically allocated based on the attention mechanism, and a multimodal data fusion operation is performed on the data feature values using the trained neural network model to obtain target fused data.

[0021] Optionally, generating a target threshold based on the target fusion data using a preset dynamic threshold algorithm includes:

[0022] Calculating the historical mean deviations corresponding to the real-time operating parameters and historical operating parameters of the target device to determine the adjustment ranges of the upper and lower thresholds, and generating a target threshold based on the target fusion data and the adjustment ranges;

[0023] Alternatively, the mean and variance of the target fusion data acquired within a preset time period are calculated to determine a target threshold interval, and the target threshold is generated based on the target threshold interval.

[0024] Optionally, the method for determining the start / stop status of a device further includes:

[0025] Updating the target threshold based on a preset update time;

[0026] Alternatively, the target threshold is updated based on sudden changes in the environment and abnormal equipment load.

[0027] In a second aspect, the present application provides a device for determining the start-up and shutdown status of a device, comprising:

[0028] a target data acquisition module, configured to acquire multi-dimensional data and pre-process the multi-dimensional data to obtain target data; the multi-dimensional data includes mechanical data, electrical data, temperature data, flow data, and pressure data;

[0029] a target data fusion module, configured to perform a feature extraction operation on the target data to obtain data feature values of the target data, and perform a multimodal data fusion operation on the data feature values to obtain target fused data;

[0030] The target threshold determination module is used to generate a target threshold based on the target fusion data using a preset dynamic threshold algorithm, and to determine the start and stop status of the target device based on the target threshold.

[0031] In a third aspect, the present application provides an electronic device, comprising:

[0032] Memory, used to store computer programs;

[0033] The processor is used to execute the computer program to implement the aforementioned method for determining the start and stop status of the equipment.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned method for determining the start and stop status of a device.

[0035] The present application provides a method for determining the start-up and shutdown status of equipment. First, multi-dimensional data is acquired and pre-processed to obtain target data; the multi-dimensional data includes mechanical data, electrical data, temperature data, flow data, and pressure data; then, a feature extraction operation is performed on the target data to obtain data characteristic values of the target data, and a multimodal data fusion operation is performed on the data characteristic values to obtain target fusion data; finally, a preset dynamic threshold algorithm is used to generate a target threshold based on the target fusion data, and the start-up and shutdown status of the target equipment is determined based on the target threshold.

[0036] As can be seen from the above, this application overcomes the limitations of a single data source and improves judgment accuracy by acquiring multi-dimensional data and preprocessing it to obtain target data. By utilizing a preset dynamic threshold algorithm to generate a target threshold based on the target fusion data, it can adapt to changes in device status. This allows for the ability to adapt to changes in device status and determine the device's start and stop status without relying on additional sensors. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0038] Figure 1 This is a flow chart of a method for determining the start and stop status of a device disclosed in this application;

[0039] Figure 2 This is a flow chart of a specific method for determining the start and stop status of equipment disclosed in this application;

[0040] Figure 3 This is a flow chart of a process for generating and updating a device start / stop threshold value disclosed in this application;

[0041] Figure 4 This is a schematic diagram of a device for determining the start and stop status of equipment disclosed in this application;

[0042] Figure 5 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] Existing technologies for determining equipment startup and shutdown status suffer from limitations including a single data dimension and rigid thresholds. Threshold generation relies on long-term historical data accumulation, requiring manual intervention for initial setting. The dynamic impact of environmental parameters (such as temperature and load) is not considered. To address this, this application provides a solution for determining equipment startup and shutdown status that adapts to changes in equipment status without relying on additional sensors.

[0045] See also Figure 1 As shown, the embodiment of the present application discloses a method for determining the start-up and shutdown status of a device, including:

[0046] Step S11: Acquire multi-dimensional data, and pre-process the multi-dimensional data to obtain target data.

[0047] In this embodiment, multi-dimensional data is collected from lower-level systems and pre-processed to obtain target data. The multi-dimensional data includes, but is not limited to, mechanical data, electrical data, temperature data, flow data, and pressure data; the lower-level systems include, but are not limited to, mechanical monitoring systems, electrical monitoring systems, and MES (Manufacturing Execution System) systems.

[0048] In this embodiment, target data is obtained by performing time synchronization and normalization on the multidimensional data and removing abnormal data from the multidimensional data. The time synchronization process involves time-aligning data with different sampling frequencies. The specific operation method can be determined based on the actual application and is not specifically limited here. For example, in one embodiment, a sliding window interpolation method can be used to time-align data with different sampling frequencies. The normalization process eliminates dimensional differences in the multidimensional data, such as normalizing vibration acceleration and temperature to the range [0, 1]. Abnormal data in the multidimensional data includes instantaneous outliers. The specific method for removing outliers can be determined based on the exemplary application and is not specifically limited here. For example, in one embodiment, a sliding standard deviation method can be used to filter out instantaneous spike noise in the current signal. Specifically, preprocessing the multidimensional data to obtain the target data can include: using a sliding window interpolation method to time-align data with different sampling frequencies and eliminate dimensional differences in the multidimensional data; and filtering out instantaneous spike noise in the current signal using a sliding standard deviation method to obtain the target data. That is, by preprocessing the multi-dimensional data to obtain target data, the influence of abnormal data on the judgment of the start and stop status of the equipment can be reduced, and the efficiency and quality of the judgment of the start and stop status of the equipment can be improved.

[0049] Step S12: performing a feature extraction operation on the target data to obtain data feature values of the target data, and performing a multimodal data fusion operation on the data feature values to obtain target fused data.

[0050] In this embodiment, feature extraction is performed based on the characteristics of the target data itself. For example, in a specific embodiment, vibration information features are extracted for mechanical vibration data through time domain, frequency domain, and time-frequency domain analysis; current harmonic features are extracted for current signal data by calculating the harmonic distortion rate and the fundamental amplitude offset; and temperature change features are extracted for temperature data by calculating the temperature gradient change rate. Specifically, the feature extraction operation performed on the target data to obtain the data characteristic values of the target data may include: extracting the vibration information characteristic values of the target data through time domain, frequency domain, and time-frequency domain analysis; extracting the current harmonic characteristic values of the target data by calculating the harmonic distortion rate and the fundamental amplitude offset; and extracting the temperature change characteristic values of the target data by calculating the temperature gradient change rate. That is, by performing feature extraction on different data characteristics, data fusion operations can be performed in subsequent steps.

[0051] In this embodiment, a multimodal data fusion operation is performed on the data feature values based on a multimodal data fusion strategy to obtain target fused data. The multimodal data fusion strategy includes but is not limited to a fusion strategy based on machine learning and a weighted average calculation strategy. Specifically, the multimodal data fusion operation is performed on the data feature values to obtain target fused data, which may include: performing a multimodal data fusion operation on the data feature values based on a fusion method based on machine learning to obtain target fused data; or performing a multimodal data fusion operation on the data feature values based on a weighted average method to obtain target fused data.

[0052] It is understandable that the specific type of the multimodal data fusion strategy can be determined based on the actual application situation and is not specifically limited here. For example, in a specific embodiment, a fusion method based on machine learning can be used as a multimodal data fusion strategy. Specifically, the fusion method based on machine learning performs a multimodal data fusion operation on the data feature values to obtain target fusion data, which may include: splicing the vibration spectrum, current harmonics, and temperature gradient into a composite input vector for neural network training; dynamically allocating weights based on the attention mechanism (for example, increasing the weight of temperature features under high temperature conditions); independently training vibration, current, and temperature sub-models, and comprehensively determining the start and stop status through weighted voting.

[0053] Step S13: Generate a target threshold based on the target fusion data using a preset dynamic threshold algorithm, and determine the start / stop status of the target device based on the target threshold.

[0054] In this embodiment, an adaptive dynamic threshold algorithm is used to adjust the target fusion data to obtain a threshold target threshold. The adaptive dynamic threshold algorithm may adopt a reinforcement learning algorithm or a sliding window statistical method. Specifically, the use of a preset dynamic threshold algorithm to generate a target threshold based on the target fusion data may include: calculating the deviation between the real-time operating parameters (speed, load, temperature) of the device and the historical mean to set the adjustment range of the upper and lower thresholds, and introducing a reward mechanism to generate a target threshold based on the target fusion data; if the misjudgment rate decreases after adjustment, a positive reward is given; if the misjudgment rate increases after adjustment, a penalty is imposed; or, dynamically calculating the mean and variance based on the data of the past 24 hours to set an initial threshold interval, and generating a target threshold based on the target threshold interval.

[0055] Furthermore, adaptively updating the threshold ensures that the device startup / shutdown status determination method can adapt to environmental factors. Specifically, the device startup / shutdown status determination method may also include: updating the target threshold based on a preset update time; or updating the target threshold based on sudden environmental changes and abnormal device load conditions. For example, in one embodiment, if a sudden load change (e.g., greater than 20%) or a sudden change in ambient temperature (e.g., greater than 5 degrees Celsius) is detected, a threshold adjustment is immediately triggered.

[0056] As can be seen from the above, the embodiments of the present application use a lower-level system to collect data, without relying on additional hardware devices such as speed sensors. At the same time, an adaptive judgment threshold is obtained through multimodal data fusion and a dynamic threshold adjustment mechanism. The fusion of mechanical, electrical and other multi-dimensional data breaks through the limitations of a single data source and improves judgment accuracy. Dynamic threshold adjustment can adapt to changes in device status. It fully utilizes existing data and reduces the cost of installing additional sensors. This makes it possible to adapt to changes in device status and make judgments on the start and shutdown status of the device without relying on additional sensors.

[0057] See also Figure 2 As shown, the embodiment of the present application provides a specific method for determining the start and stop status of a device, including:

[0058] In this embodiment, multi-dimensional data is collected and aggregated from each lower system, such as acceleration data from the mechanical monitoring system, current data from the electrical monitoring system, and process data such as rolling force from the MES system; the multi-dimensional data is preprocessed to obtain preprocessed data, and the processing method includes time synchronization, normalization processing, and outlier removal operations. Specifically, due to the large difference in the acquisition frequency of acceleration, current, and rolling force, time alignment is required through sliding window interpolation. The acceleration (unit: m / s²), current (unit: A), and rolling force (unit: N) are normalized to the [0,1] interval to eliminate dimensional differences. The instantaneous spike noise in the current signal is filtered out based on the sliding standard deviation method, and outliers in the acceleration are removed by filtering. Preprocessing multi-dimensional data can reduce the impact of abnormal data.

[0059] Furthermore, feature extraction is performed on the preprocessed data to obtain feature data. Feature extraction requires different methods depending on the data type. Specifically, the effective value of acceleration is obtained through time domain processing, and the calculation formula is as follows:

[0060] ;

[0061] Where n is the number of vibration acceleration values collected within the preset time period, is the i-th vibration acceleration value.

[0062] By calculating the harmonic distortion (THD) and the fundamental amplitude offset, the current harmonic characteristics are extracted to characterize electrical load anomalies. The calculation formula is as follows:

[0063] ;

[0064] in, is the effective value of the fundamental current, 、 、... They are the effective values of the 2nd, 3rd, ..., nth harmonic currents respectively.

[0065] By calculating the temperature gradient change rate, abnormal heat dissipation or local overheating can be monitored.

[0066] Furthermore, the feature data is fused to obtain fused data. This embodiment adopts a weighted average method, and the formula is:

[0067] ;

[0068] in, is the characteristic data, For each feature data in the fused data The weights can be assigned based on the confidence level. Feature fusion is performed through weighted averaging, which does not rely on machine learning and does not require secondary training. Therefore, the fused data can be directly used as the start / stop judgment threshold.

[0069] The fused data is compared with the obtained threshold. If the fused data is not less than the threshold, the start / stop status value is set to 1; otherwise, it is set to 0. The specific comparison method and data setting method can be determined according to the actual application situation and are not specifically limited here.

[0070] See also Figure 4 As shown, the threshold is recalculated and updated periodically. The specific time period can be set based on actual site conditions and historical experience. For steady-state devices, the threshold update period can be set longer. Here, the period is 24 hours, and a sliding window statistical method is used to dynamically calculate the mean, which is used as the new threshold. Similarly, other statistical values, such as the median, can also be used. It is important to note that during the threshold update process, the data collection and aggregation, data preprocessing, feature extraction, and multimodal data fusion processes must be consistent with the process for determining the device's startup and shutdown status.

[0071] As can be seen from the above, the embodiments of the present application can reduce the error rate through multimodal data fusion. Adaptive thresholds can respond in real time to device aging and load fluctuations, avoiding frequent misjudgments. This can fully utilize data already collected by lower-level systems, reducing the cost of new data acquisition hardware. This allows the system to adapt to changes in device status and determine device startup and shutdown status without relying on additional sensors.

[0072] See also Figure 4 As shown, the embodiment of the present application discloses a device for determining the start-up and shutdown status of a device, comprising:

[0073] The target data acquisition module 11 is used to acquire multi-dimensional data and pre-process the multi-dimensional data to obtain target data; the multi-dimensional data includes mechanical data, electrical data, temperature data, flow data and pressure data;

[0074] The target data fusion module 12 is configured to perform a feature extraction operation on the target data to obtain data feature values of the target data, and perform a multimodal data fusion operation on the data feature values to obtain target fused data;

[0075] The target threshold determination module 13 is configured to generate a target threshold based on the target fusion data using a preset dynamic threshold algorithm, and determine the start / stop status of the target device based on the target threshold.

[0076] As can be seen from the above, the embodiments of the present application overcome the limitations of a single data source and improve judgment accuracy by acquiring multi-dimensional data and preprocessing it to obtain target data. By utilizing a preset dynamic threshold algorithm to generate a target threshold based on the target fusion data, it can adapt to changes in device status. This allows for adaptive judgment of device startup and shutdown status without relying on additional sensors.

[0077] In some specific implementations, the target data acquisition module 11 may specifically include:

[0078] The data preprocessing unit is used to time-align the multi-dimensional data of different sampling frequencies, normalize the multi-dimensional data, and eliminate abnormal data in the multi-dimensional data to obtain target data.

[0079] In some specific implementations, the target data fusion module 12 may specifically include:

[0080] A first feature extraction unit is used to extract the vibration information feature value of the target data through time domain, frequency domain, and time-frequency domain analysis;

[0081] A second feature extraction unit is used to extract the current harmonic feature value of the target data by calculating the harmonic distortion rate and the fundamental amplitude offset;

[0082] The third feature extraction unit is configured to extract the temperature change feature value of the target data by calculating the temperature gradient change rate.

[0083] a first target fusion number generating submodule, configured to perform a multimodal data fusion operation on the data feature values based on a machine learning fusion method to obtain target fusion data;

[0084] a second target fusion number generating submodule, configured to perform a multimodal data fusion operation on the data feature values based on a weighted average method to obtain target fusion data;

[0085] Accordingly, the first target fusion number generation submodule may specifically include:

[0086] a neural network model training unit, configured to generate a composite input vector based on the vibration information characteristic value, the current harmonic characteristic value, and the temperature change characteristic value, and train the original neural network model using the composite input vector to obtain a trained neural network model;

[0087] The target fusion data generating unit is used to dynamically allocate weights based on the attention mechanism, and use the trained neural network model to perform a multimodal data fusion operation on the data feature values to obtain target fusion data.

[0088] In some specific implementations, the target threshold determination module 13 may specifically include:

[0089] a first target threshold generating unit, configured to calculate a historical mean deviation corresponding to the real-time operating parameters and the historical operating parameters of the target device to determine an adjustment range of the upper and lower thresholds, and generate a target threshold based on the target fusion data and the adjustment range;

[0090] The second target threshold generating unit is configured to calculate the mean and variance of the target fusion data acquired within a preset time period to determine a target threshold interval, and generate a target threshold based on the target threshold interval.

[0091] In some specific implementations, the device startup / shutdown status determination apparatus may further include:

[0092] a first target threshold updating unit, configured to update the target threshold based on a preset update time;

[0093] The second target threshold updating unit is configured to update the target threshold based on sudden environmental changes and abnormal equipment load conditions.

[0094] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 5 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram should not be considered as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the device start / stop status determination method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0095] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0096] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0097] The operating system 221 is used to manage and control the hardware devices on the electronic device 20 and the computer program 222, and can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of implementing the device startup and shutdown status determination method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs capable of performing other specific tasks.

[0098] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned method for determining the device startup and shutdown status. The specific steps of this method can be found in the corresponding content disclosed in the aforementioned embodiments and will not be repeated here.

[0099] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0100] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0101] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0102] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0103] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for determining the start-up and shutdown status of a device, characterized in that: include: Acquiring multi-dimensional data, and preprocessing the multi-dimensional data to obtain target data; The multi-dimensional data includes mechanical data, electrical data, temperature data, flow data and pressure data; Performing a feature extraction operation on the target data to obtain data feature values of the target data, and performing a multimodal data fusion operation on the data feature values to obtain target fused data; A preset dynamic threshold algorithm is used to generate a target threshold based on the target fusion data, and the start / stop status of the target device is determined based on the target threshold.

2. The method for determining the start and stop status of equipment according to claim 1, characterized in that: The preprocessing of the multi-dimensional data to obtain target data includes: The multi-dimensional data of different sampling frequencies are time-aligned, the multi-dimensional data are normalized, and abnormal data in the multi-dimensional data are eliminated to obtain target data.

3. The method for determining the start-up and shutdown status of equipment according to claim 1, characterized in that: The performing a feature extraction operation on the target data to obtain a data feature value of the target data includes: Extracting vibration information characteristic values of the target data through time domain, frequency domain, and time-frequency domain analysis; Extracting the current harmonic characteristic value of the target data by calculating the harmonic distortion rate and the fundamental amplitude offset; The temperature change characteristic value of the target data is extracted by calculating the temperature gradient change rate.

4. The method for determining the start and stop status of equipment according to claim 3, characterized in that: The performing a multimodal data fusion operation on the data feature values to obtain target fused data includes: Performing a multimodal data fusion operation on the data feature values based on a machine learning fusion method to obtain target fused data; Alternatively, a multimodal data fusion operation is performed on the data feature values based on a weighted average method to obtain target fused data.

5. The method for determining the start and stop status of equipment according to claim 4, characterized in that: The machine learning-based fusion method performs a multimodal data fusion operation on the data feature values to obtain target fused data, including: generating a composite input vector based on the vibration information characteristic value, the current harmonic characteristic value, and the temperature change characteristic value, and training the original neural network model using the composite input vector to obtain a trained neural network model; Weights are dynamically allocated based on the attention mechanism, and a multimodal data fusion operation is performed on the data feature values using the trained neural network model to obtain target fused data.

6. The method for determining the start and stop status of equipment according to claim 1, characterized in that: The generating of a target threshold based on the target fusion data by using a preset dynamic threshold algorithm includes: Calculating the historical mean deviations corresponding to the real-time operating parameters and historical operating parameters of the target device to determine the adjustment ranges of the upper and lower thresholds, and generating a target threshold based on the target fusion data and the adjustment ranges; Alternatively, the mean and variance of the target fusion data acquired within a preset time period are calculated to determine a target threshold interval, and the target threshold is generated based on the target threshold interval.

7. The method for determining the start-up and shutdown status of equipment according to any one of claims 1 to 6, characterized in that: Also includes: Updating the target threshold based on a preset update time; Alternatively, the target threshold is updated based on sudden changes in the environment and abnormal equipment load.

8. A device for determining the start and stop status of equipment, characterized in that: include: A target data acquisition module is used to acquire multi-dimensional data and pre-process the multi-dimensional data to obtain target data; The multi-dimensional data includes mechanical data, electrical data, temperature data, flow data and pressure data; a target data fusion module, configured to perform a feature extraction operation on the target data to obtain data feature values of the target data, and perform a multimodal data fusion operation on the data feature values to obtain target fused data; The target threshold determination module is used to generate a target threshold based on the target fusion data using a preset dynamic threshold algorithm, and to determine the start and stop status of the target device based on the target threshold.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the device start-up and shutdown status judgment method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program, wherein when the computer program is executed by a processor, the method for determining the start-up and shutdown status of an equipment according to any one of claims 1 to 7 is implemented.

Images