Method, device, equipment and medium for determining boundary value of equipment working state

Abstract

The present invention relates to the field of artificial intelligence technology, and discloses a method, device, equipment and medium for determining the boundary value of the working state of a device, including: obtaining description data of the working state of the device in a preset time period; clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device; and determining the boundary value of the description data of each working state of the device according to the at least one cluster. It is not necessary to know the fixed parameters of the working current of the device, and the working current range of the device can be calculated by the threshold value obtained by the algorithm.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and medium for determining a boundary value of a device working state. Background Art

[0002] Equipment leasing refers to the purchase of medium and heavy equipment that small and medium-sized enterprises are temporarily unable to purchase by leasing service providers, and the provision of equipment use rights to equipment leasing companies in the form of rent. Through the continuous tracking of the business status of enterprises through the Internet of Things technology, more precisely, the monitoring of the working status of leased equipment, as well as other enterprise and equipment information, data analysis can grasp the current business status of enterprises and various industries in the market, and even predict the trends of enterprises and various industries. It is also convenient for risk control to make reasonable judgments and the market to make more accurate promotions. The effectiveness of working status identification comes from the accuracy of standby and power-on current judgment.

[0003] The operating rate indicator based on the equipment standby and power-on current thresholds is the most basic indicator for business-related business analysis. The accuracy of the operating rate indicator is directly determined by the accuracy of the threshold. The traditional method is to obtain it based on the standby and power-on currents marked on the equipment. However, due to many factors such as the age of the equipment and the operating load, the marked standby and power-on currents cannot accurately reflect the actual situation. And the default value is 0.2 to 1 (unit, ampere) if there is no marking.

[0004] The inventors realize that devices vary greatly, and the default values ​​certainly cannot reflect the actual operating status of the devices. Therefore, the traditional implementation method will be quite different from the actual situation, which will cause information delay and the problem of not being able to know the device status changes in time. Summary of the invention

[0005] The present invention provides a method, device, computer device and medium for determining the boundary value of the working state of an equipment to solve the technical problem that when predicting the working states of different types of equipment, the boundary values ​​(threshold values) of the working states of all equipment are fixed values, which are inconsistent with the actual conditions of the equipment, and the subsequent predicted working states of the equipment cannot reflect the actual operating states of the equipment.

[0006] In a first aspect, a method for determining a boundary value of a device working state is provided, comprising:

[0007] Obtaining descriptive data of the working status of the device during a preset time period;

[0008] Clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device;

[0009] According to the at least one cluster, a boundary value of the description data of each working state of the device is determined.

[0010] In a second aspect, a device for determining a boundary value of a device working state is provided, comprising:

[0011] An input unit, used to obtain descriptive data of the working status of the device during a preset time period;

[0012] A clustering unit, configured to cluster the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device;

[0013] The output unit is used to determine the boundary value of the description data of each working state of the device according to the at least one cluster.

[0014] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the step of determining the boundary value of the working state of the above-mentioned device is implemented.

[0015] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the step of determining the boundary value of the working state of the above-mentioned device is implemented.

[0016] In the scheme implemented by the method, device, computer device and storage medium for determining the boundary value of the working state of the above-mentioned equipment, the description data of the working state of the equipment in a preset time period can be obtained through the client; the server clusters the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the equipment; and the boundary value of the description data of each working state of the equipment is determined according to at least one cluster. In the present invention, the density clustering algorithm is used for each device to solve the clustering of irregular shapes, and the processing of noise data is also better. The algorithm requires a large amount of space resources and budget time complexity, but the cluster resources of big data, dynamic memory, core number allocation, and memory-based calculation method can be used to quickly calculate the results. The model algorithm can basically be operated on a daily basis to update the boundary value (threshold) of the description data of the working state of the equipment. The improved algorithm threshold can replace the traditional fixed boundary value, and there is no need to know the fixed parameters of the working current of the equipment. The working current range of the equipment can be calculated by the threshold obtained by the algorithm, which greatly improves the accuracy of equipment startup. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0018] Figure 1 It is a schematic diagram of an application environment for determining the boundary value of the working state of a device in one embodiment of the present invention;

[0019] Figure 2 It is a schematic diagram of a process for determining the boundary value of the working state of a device in one embodiment of the present invention;

[0020] Figure 3 yes Figure 1 A schematic flow chart of a specific implementation of step S30;

[0021] Figure 4 yes Figure 1 A schematic flow chart of a specific implementation of step S301;

[0022] Figure 5 It is a flowchart of a specific implementation method for determining parameters of a density clustering model;

[0023] Figure 6 It is a structural schematic diagram of a device working state apparatus in one embodiment of the present invention;

[0024] Figure 7 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0025] Figure 8 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0027] The device working status method based on artificial intelligence provided by the embodiment of the present invention can be applied to Figure 1In an application environment, the client communicates with the server through a network. The server can obtain the descriptive data of the working state of the device in a preset time period through the client, cluster the descriptive data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device; and determine the boundary value of the descriptive data of each working state of the device according to at least one cluster. In the present invention, the density clustering algorithm is used to solve the clustering of irregular shapes, and the processing of noise data is also better. The principle is simply to draw a circle, in which two parameters are defined, one is the maximum radius of the circle, and the other is the minimum number of points that should be contained in a circle. As long as the density (the number of objects or data points) of the adjacent area exceeds a certain threshold, clustering continues, and the last one in a circle is a class. By adjusting the maximum radius and the parameter configuration of the core point, the accuracy will become higher and higher with the richness of historical data. The algorithm requires a lot of space resources and budget time complexity, but the cluster resources of big data, dynamic memory, core number allocation, and memory-based calculation methods can be used to quickly calculate the results. The model algorithm can basically be operated on a daily basis to update the device threshold. The improved algorithm threshold can replace the traditional fixed threshold, and there is no need to know the fixed parameters of the working current of the device. The working current range of the device can be calculated by the threshold obtained by the algorithm, which greatly improves the accuracy of the device startup. Among them, the client can be but not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices. The server can be implemented with an independent server or a server cluster composed of multiple servers. The present invention is described in detail below through specific embodiments.

[0028] The present invention is described in detail below through specific embodiments.

[0029] See also Figure 2 As shown, Figure 2 A flow chart of a method for determining a boundary value of a device working state provided in an embodiment of the present invention includes the following steps:

[0030] S10: Obtaining description data of the working status of the device in a preset time period.

[0031] The determination of the boundary value of the device working state provided by the present invention can be applied to the determination of the threshold of the working state of the device in various application scenarios. The boundary value of the description data of each working state is usually realized through the server, which can receive the description data of the working state in real time.

[0032] The working status data may refer to the data generated synchronously during the operation of the leased equipment, such as the temperature data, amplitude data and current data of the leased equipment. The working status description information may refer to the status information of the lessee of the leased equipment performing production operations by operating the leased equipment, such as the time the lessee uses the leased equipment within a week. The working status description information may include at least one of the following: the operating rate, operating time and operating peak time period under at least one time unit. The time unit may be a day, a week, a month, etc.

[0033] In an application scenario, the preset working state may include device shutdown, device standby and device startup; in step S10, that is, obtaining the description data of the working state of the device in the preset time period, may include:

[0034] Obtain the current data and / or amplitude data uploaded by the device, divide the description data according to the preset time window, select the largest current data and / or amplitude data in each time window as the description data obtained in the time window, and use the time when the current data and / or amplitude data was last uploaded in the time window as the acquisition time of the time window.

[0035] The equipment to be judged is manufacturing equipment, including semi-automatic equipment and low-power equipment. The current to be processed refers to the data obtained by collecting the operating current of the manufacturing equipment through an external collection device, which can be data collected at a lower frequency, for example, data generated by collecting once a minute. The current can specifically refer to current signal data.

[0036] It can be understood that real-time collection of working status description data of each device includes: real-time collection of amplitude data of each device through a device bracelet set on each device, and / or real-time collection of current data of each device through a current acquisition device configured on the power grid line where each device is located.

[0037] Among them, the device bracelet can collect the amplitude data of each device in real time. Specifically, since each device will generate vibration amplitude when working, the device bracelet can be used to further determine whether each device is in working state.

[0038] Accordingly, the current acquisition device can collect the current data of each device in real time. Specifically, since each industrial device will generate corresponding current data when working, the current data collected by the current acquisition device can be used to further determine whether each device is in working state.

[0039] Furthermore, the prediction data preparation includes active and passive types of wristbands. The frequency of active wristbands is relatively high, once every 7 seconds. In many cases, the current value will be relatively stable in a short period of time, so the selected data is not representative. It is necessary to sample and take values, specifically in units of hours, grouped every 15 minutes, take the latest collection time in the group as the collection time, and the maximum current value as the current value to obtain the sampled current data.

[0040] According to the above principles, after sampling the data from the DWD layer of the Hive data warehouse, obtain the latest 3000 current data of the device in reverse order of collection time (the latest historical data can better reflect the recent working status of the device), and the number of days for current collection should be at least 7 days. If the number of days is too short, it will affect the accuracy of the algorithm model.

[0041] S20: Clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device.

[0042] Among them, when the device has multiple working states, the distribution of the corresponding obtained working state description data should be a distribution span of multiple working states. Therefore, the description data is clustered to obtain at least one cluster. For example, the preset working states may include device shutdown, device standby and device startup, so the description data is clustered to obtain at least three clusters.

[0043] It should be noted that density clustering is a density-based clustering algorithm that assumes that the clustering structure can be determined by the compactness of the sample distribution, and performs clustering based on the density of the data set in spatial distribution. That is, as long as the sample density in an area is greater than a certain threshold, it will be classified into a cluster similar to it.

[0044] Take the basic DBSCAN algorithm as an example:

[0045] The DBSCAN algorithm uses a set of parameters about the concept of "neighborhood" to describe the compactness of sample distribution, divides areas with sufficient density into clusters, and can discover clusters of arbitrary shapes under noisy conditions.

[0046] The idea of ​​the algorithm is to derive a sample set (cluster) connected by the maximum density from the density reachable relationship. There are one or more core objects in such a set. If there is only one core object, then all other non-core objects in the cluster are in the ε neighborhood of this core object; if there are multiple core objects, then the ε neighborhood of any core object must contain another core object (otherwise it is not density reachable). These core objects and all samples contained in their ε neighborhood constitute a class.

[0047] Of course, there are more advanced density clustering algorithms, many of which are based on the DBSCAN algorithm, so the DBSCAN algorithm can represent the density clustering binning algorithm.

[0048] The basic concepts are explained as follows:

[0049] Neighborhood: For any given sample x and distance ε, the ε neighborhood of x refers to the set of samples whose distance to x does not exceed ε;

[0050] Core object: If the ε neighborhood of sample x contains at least minPts samples, then x is a core object;

[0051] Density direct access: If sample b is in the ε neighborhood of a and a is a core object, then sample b is said to be density directly accessible from sample x;

[0052] Density reachable: For samples a and b, if there exist samples p1, p2, ..., pn, where p1 = a, pn = b, and each sample in the sequence is density directly reachable to its previous sample, then samples a and b are said to be density reachable;

[0053] Density connection: For samples a and b, if there exists a sample k such that a and k are density reachable, and k and b are density reachable, then a and b are density connected.

[0054] The main features of density clustering are: (1) discovering clusters of arbitrary shapes; (2) insensitive to noisy data; (3) one-time scanning; (4) requiring density parameters as stopping conditions; and (5) large amount of computation and high complexity.

[0055] In the field of detecting the working status of equipment, it is not necessary to calculate high-dimensional data, only one-dimensional data is needed; because it is one-dimensional data, the density parameter is easy to set. Therefore, with the help of these excellent characteristics of density clustering, irregular clustering can be solved, and noise data can be processed better. The principle is simply to draw a circle, in which two parameters must be defined, one is the maximum radius of the circle, and the other is the minimum number of points that should be contained in a circle. As long as the density of the neighboring area (the number of objects or data points) exceeds a certain threshold, clustering will continue, and the last one in a circle is a class. By adjusting the maximum radius and the parameter configuration of the core point, the accuracy will become higher and higher with the richness of historical data. The algorithm requires a lot of space resources and budget time complexity, but the cluster resources of big data, dynamic memory, core number allocation, and memory-based computing methods can quickly calculate the results.

[0056] In some application scenarios, density clustering is performed on the acquired current data according to time information describing the acquisition of the data (such as device current) to obtain at least one cluster.

[0057] Specifically, time information refers to the order in which current is acquired, such as the order in which current is acquired. For convenience, only the order in which current is acquired and the time interval may be required (for example, in some scenarios, the current data and / or amplitude data uploaded by the device are acquired, the description data is divided according to a preset time window, the largest current data and / or amplitude data in each time window is selected as the description data acquired in the time window, and the time of the last upload of current data and / or amplitude data in the time window is the acquisition time of the time window). In this way, density clustering is performed on the current terminals to be judged, that is, the time information is fully combined to calculate the density of the current to be judged, so as to cluster the current terminals to be judged according to the density of the current to be judged in the order of time.

[0058] Specifically, here we add a dimension to the current to be clustered. The data in this dimension is a sequence value. The simplest sequence value is to take a value from 0 to n, but the discreteness of this sequence value needs to be selected according to the size of the current value. Assuming that the current value is 1, the sequence value is preferably increased or decreased at intervals of 0.5. (Add time information) Density clustering is performed on this two-dimensional data after the dimension is increased. Density clustering is different from hierarchical clustering. It tends to cluster data together according to the proximity of the data. After adding the time series dimension, density clustering will also cluster points that are close in time together, so that the effect of vertical classification can be achieved. For example, there are 5 currents to be judged, and the time information of the 5 currents to be judged is 8 o'clock, 8 o'clock 1 minute, 8 o'clock 2 minutes, 8 o'clock 3 minutes and 8 o'clock 4 minutes, that is, the intervals are the same, so it is only necessary to judge according to the value of the current to be judged. Assume that the current values ​​of the 5 currents to be judged are 1 ampere, 1.1 amperes, 1.2 amperes, 1.4 amperes and 1.6 amperes respectively. When density clustering is performed, since the densities of 1.1 amps and 1 amps, and 1.2 amps and 1.1 amps are equal, these parts are clustered into one cluster. However, the densities of 1.4 amps and 1.2 amps, and 1.6 amps and 1.4 amps are equal, so 1.4 amps and 1.6 amps are clustered into another cluster.

[0059] S30: Determine, according to at least one cluster, a boundary value of description data of each working state of the device.

[0060] In some application scenarios, in step S30, that is, determining the boundary value of the description data of the device in each working state according to at least one cluster, may include:

[0061] S301: Determine, according to the number of preset working states, a cluster in at least one cluster corresponding to each working state;

[0062] In some application scenarios, for example, the equipment is manufacturing equipment, it may include multiple working states. For example, for semi-automatic processing equipment, it may include processing state, non-processing state, standby state, etc., or it may also include rough processing state, fine processing state, or processing process one state, processing process two state, etc. This application does not impose any restrictions on this.

[0063] In this embodiment, the current data collected by the acquisition device refers to the data of the processing equipment in the continuous operation time interval, that is, the collected current data may include the current data of the processing state, and may also include the current data of the non-processing state, and may include the current data of the rough processing state, and may also include the current data of the fine processing state, or may also include the current data of the processing process one state, and may also include the current data of the processing process two state. It can be seen that since the working state of the equipment is generally a plurality of working states, a plurality of clusters can be obtained after clustering. In step S301, according to the preset working state, a cluster corresponding to each working state is determined from at least one cluster, including:

[0064] S301a: Determine whether the number of obtained clusters is greater than the number of preset working states.

[0065] S301b: If it is greater than, sort the average values ​​of the descriptive data in the clusters, and select a preset number of clusters at the beginning or end of the sorting result as the target cluster, wherein the preset number is the number of preset working states, and the average value of the descriptive data of the target cluster is greater than the average value of the descriptive data in any cluster other than the target cluster in at least two clusters.

[0066] For example, the working state includes three working states: shutdown, standby and power on, but four clusters are obtained through clustering. In this case, one cluster needs to be deleted. Specifically, in this embodiment, the cluster with the smallest average working state value in the cluster is deleted. This is because the description data of the working state in this embodiment is current, so the cluster with the smallest average working state value in the smallest cluster is deleted.

[0067] In this embodiment, the clusters are sorted in ascending order according to the meam (sample average) values ​​in the clusters, and the first three clusters are selected.

[0068] S302: Determine the boundary value of the description data of the current working state and the previous working state according to the first description data in the cluster corresponding to the current working state and the second description data in the cluster corresponding to the previous working state, wherein the first description data is the minimum description data in the cluster and the second description data is the maximum description data in the cluster.

[0069] It is understandable that the current working state and the previous working state only represent the relationship between two adjacent working states, not the working state that the characteristic device is in. Take the working state of the device including three working states: shutdown, standby and power on, and the description data of the working state of the device is current as an example, where culs_maxs represents the maximum value of the cluster set, culs_mins represents the minimum value of the cluster set, and the threshold value of the standby current idle_ele = round((culs_maxs[0]+culs_mins[1]) / 2, 4), culs_max[0] corresponds to the maximum value in the shutdown state cluster, culs_mins[1] corresponds to the minimum value in the standby state cluster, 4 represents the value is accurate to four decimal places, and the following has the same meaning, that is, the standby current threshold is the maximum value in the shutdown state cluster and Half of the minimum value in the standby state cluster, that is, the threshold of the standby current, that is, the current of the device is less than the standby current, then the device is in the shutdown state; the threshold of the working current work_ele = round((culs_maxs[1]+culs_mins[2]) / 2, 4), culs_max[1] corresponds to the maximum value of the standby state, culs_mins[2] corresponds to the minimum value of the working state, that is, the threshold of the working current is half of the maximum value in the standby state cluster and the minimum value in the working state cluster, that is, when the current of the device is between the standby current and the working current, the device is in the standby state; when the current of the device is greater than the working current, the device is in the working state. If the working state of the device is more than one, the same principle applies. The standby, power-on, sample variance (clu_std) and evaluation coefficient are obtained by this averaging method and returned as the result.

[0070] In some implementations, the method for determining the boundary value of the device working state further includes the step of determining parameters of a density clustering model (DBSCAN), wherein the parameters include a minimum number of samples and a domain distance; determining the parameters of the density clustering model (DBSCAN) includes:

[0071] S01: Obtain training description data of a preset working state of a device and a plurality of candidate parameters of a density clustering model (DBSCAN).

[0072] The training description data is the historical data of the device in the historical time period, for example, in units of hours, grouped every 15 minutes, and the latest collection time in the group is taken as the collection time, and the maximum current value is taken as the current value to obtain the sampled current data. The latest 3000 current data of the device are obtained in reverse order of collection time (the latest historical data can better reflect the recent working status of the device), and the number of days for current collection is at least 7 days. If the number of days is too short, it will affect the accuracy of the algorithm model.

[0073] The candidate parameters include the minimum number of samples (min_samples) = [2, 3] and the neighborhood distance (eps). For example, the specific values ​​can be the minimum number of samples (min_samples) = [2, 3], the neighborhood distance (eps) = [0.01, 0.02, 0.05, 0.1, 0.001, 0.002, 0.005, 0.0001, 0.0002, 0.0005], that is, the clustering algorithm of the same device runs 2*10 times. Of course, the number and specific values ​​of the minimum number of samples and the neighborhood distance can also be other. For example, if the minimum number of samples is 3 and the neighborhood distance is 10, the clustering algorithm of the same device runs 3*10 times. The steps of each operation are the same as the above steps S10 to S30, which will not be repeated here.

[0074] S02: Substitute each candidate parameter into the density clustering model (DBSCAN) respectively, and cluster the training description data through the density clustering model (DBSCAN) to obtain at least one candidate cluster and a clustering evaluation coefficient;

[0075] S03: Determine the sample variance of each candidate cluster;

[0076] S04: Determine the parameters of the density clustering model (DBSCAN) according to the clustering evaluation coefficient and the sample variance of each cluster.

[0077] It should be noted that the clustering evaluation coefficients include the silhouette coefficient (silhouette_score) and the CH index (calinski_harabasz_score). Of course, other clustering evaluation coefficients can also be used in other application scenarios as long as the parameters of the density clustering model can be optimized. The result sets obtained by multiple passes of the model algorithm are sorted in ascending order according to the sample variance (clu_std) (the smaller the cluster variance value, the closer it is).

[0078] The CH index measures the compactness within a class by calculating the sum of the squares of the distances between each point in the class and the center of the class, and measures the separation of the data set by calculating the sum of the squares of the distances between the center points of each class and the center point of the data set. The CH index is obtained by the ratio of separation to compactness. Therefore, the larger the CH, the tighter the class itself, the more dispersed the classes are, and the better the clustering results.

[0079] The silhouette coefficient combines cohesion and separation, and its calculation steps are as follows:

[0080] 1. For the i-th object, calculate the average distance from it to all other objects in its cluster, record it as ai (reflecting the cohesion);

[0081] 2. For the i-th object and any cluster that does not contain this object, calculate the average distance from this object to all objects in the given cluster, and denote it as bi (reflecting the separation degree).

[0082] 3. The silhouette coefficient of the i-th object is si = (bi - ai) / max(ai, bi) / / Need to study the formula plugin of WordPress later.

[0083] As can be seen from the above, the value range of the silhouette coefficient is [-1, 1], and the larger the value, the better. When the value is negative, it indicates that ai < bi, and the sample is assigned to the wrong cluster, and the clustering result is unacceptable. For results close to 0, it indicates that there is an overlapping situation in the clustering result.

[0084] The result sets obtained multiple times through the model algorithm are sorted in ascending order according to the sample variance (clu_std) (the smaller the cluster variance value, the closer they are), and sorted in ascending order according to the silhouette_score and calinski_harabasz_score (the larger these two coefficient values, the closer the same-class samples are and the farther the different-class samples are, and the better the clustering effect). After sorting, add the values of clu_std, silhouette_score, and calinski_harabasz_score, and select the data with the smallest final score value.

[0085] Based on the above, step S04 can specifically include:

[0086] Sum the clustering evaluation coefficients corresponding to each of the candidate parameters and the sample variances of all clusters corresponding to the candidate parameters to obtain a sum value.

[0087] Select the candidate parameter corresponding to the smallest sum value as the parameter of the density clustering model.

[0088] Furthermore, the present invention can also optimize the model algorithm, mainly by reasonably adjusting the empirical values of the parameters of eps (neighborhood) and min_sample (minimum number of samples in a cluster). For example, increasing the minimum number of samples can improve the accuracy of clustering. At the same time, the size of the dataset of current sampling points can be increased, as well as the range of the number of acquisition days, to obtain a more abundant dataset and a more accurate prediction and judgment.

[0089] It can be seen from the above scheme that the description data of the working status of the device in a preset time period can be obtained through the client; the server clusters the description data to obtain at least one cluster, wherein each cluster corresponds to a working status of the device; and the boundary value of the description data of each working status of the device is determined according to at least one cluster. In the present invention, a density clustering algorithm is used for each device to solve the clustering of irregular shapes, and the processing of noise data is also better. The algorithm requires a large amount of space resources and budget time complexity, but the results can be calculated quickly by using the cluster resources of big data, dynamic memory, core number allocation, and memory-based calculation methods. The model algorithm can basically be operated on a daily basis to update the device threshold. The improved algorithm threshold can replace the traditional fixed threshold, and there is no need to know the fixed parameters of the working current of the device. The working current range of the device can be calculated by the threshold obtained by the algorithm, which greatly improves the accuracy of equipment startup.

[0090] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0091] In one embodiment, a device for determining the boundary value of the device working state is provided, and the device for determining the boundary value of the device working state corresponds one-to-one to the method for determining the boundary value of the device working state in the above embodiment. Figure 6 As shown, the device working status apparatus includes an input unit 101, a clustering unit 102 and an output unit 103.

[0092] The detailed description of each functional module is as follows:

[0093] An input unit 101 is used to obtain description data of the working status of the device in a preset time period;

[0094] A clustering unit 102, configured to cluster the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device;

[0095] The output unit 103 is configured to determine a boundary value of the description data of each working state of the device according to the at least one cluster.

[0096] In one embodiment, the clustering unit 102 is specifically configured to:

[0097] The description data is clustered by a density-based clustering model (DBSCAN) to obtain at least one cluster.

[0098] In one embodiment, the clustering unit 102 is further configured to:

[0099] Determining parameters of the density clustering model (DBSCAN), wherein the parameters include a minimum number of samples and a domain distance; determining the parameters of the density clustering model (DBSCAN), including:

[0100] Obtaining training description data of the preset working state of the device and multiple candidate parameters of the density clustering model (DBSCAN)

[0101] Substituting each of the candidate parameters into the density clustering model (DBSCAN), clustering the training description data by the density clustering model (DBSCAN) to obtain at least one candidate cluster and a clustering evaluation coefficient;

[0102] Determining the sample variance of each of the candidate clusters;

[0103] The parameters of the density clustering model (DBSCAN) are determined according to the clustering evaluation coefficient and the sample variance of each cluster.

[0104] In one embodiment, the clustering unit 102 is further configured to:

[0105] Determine the candidate parameter that corresponds to the minimum sum of the clustering evaluation coefficient and the sample variance of each cluster as the parameter of the density clustering model (DBSCAN).

[0106] In one embodiment, the clustering unit 102 is specifically configured to:

[0107] According to the number of preset working states, determining a cluster in the at least one cluster corresponding to each working state one by one;

[0108] According to the first description data in the cluster corresponding to the current working state and the second description data corresponding to the previous working state, the boundary values ​​of the description data of the current working state and the previous working state are determined, wherein the first description data is the minimum description data in the cluster and the second description data is the maximum description data in the cluster.

[0109] In one embodiment, the clustering unit 102 is specifically configured to:

[0110] Determine whether the number of clusters obtained is greater than the number of preset working states,

[0111] If it is greater than, the average values ​​of the descriptive data in the clusters are sorted, and a preset number of clusters at the beginning or end of the sorting result are selected as target clusters, wherein the preset number is the number of preset working states, and the average value of the descriptive data of the target cluster is greater than the average value of the descriptive data in any cluster other than the target cluster in the at least two clusters.

[0112] In one embodiment, the preset working state includes device shutdown, device standby and device startup;

[0113] The input unit 101 is specifically used for:

[0114] Acquire the current data and / or the amplitude data uploaded by the device, divide the description data according to preset time windows, select the largest current data and / or the amplitude data in each time window as the description data acquired in the time window, and use the time when the current data and / or the amplitude data were last uploaded in the time window as the acquisition time of the time window.

[0115] For the specific definition of the device working state device, please refer to the definition of the method for determining the boundary value of the device working state in the above text, which will not be repeated here. Each module in the above-mentioned device working state device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0116] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, the functions or steps of the service side of a method for determining a boundary value of a device working state are implemented.

[0117] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 8As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, the client-side functions or steps of a method for determining a boundary value of a device working state are implemented.

[0118] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0119] Obtaining descriptive data of the working status of the device during a preset time period;

[0120] Clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device;

[0121] According to the at least one cluster, a boundary value of the description data of each working state of the device is determined.

[0122] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0123] Obtaining descriptive data of the working status of the device during a preset time period;

[0124] Clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device;

[0125] According to the at least one cluster, a boundary value of the description data of each working state of the device is determined.

[0126] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0127] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0128] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0129] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for determining a boundary value of a device working state, characterized in that: include: Obtaining descriptive data of the working status of the device during a preset time period; Clustering the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device; Determining, according to the at least one cluster, a boundary value of the description data of each working state of the device; The step of determining, according to the at least one cluster, boundary values ​​of the description data of the device in each working state includes: According to the number of preset working states, determining a cluster in the at least one cluster corresponding to each working state one by one; The step of determining, according to the number of preset working states, a cluster in the at least one cluster corresponding to each working state one by one, comprises: Determine whether the number of clusters obtained is greater than the number of preset working states, If it is greater than, the average values ​​of the descriptive data in the clusters are sorted, and a preset number of clusters at the beginning or end of the sorted result are selected as target clusters, wherein the preset number is the number of preset working states, and the average value of the descriptive data of the target cluster is greater than the average value of the descriptive data in any cluster other than the target cluster in the at least two clusters.

2. The method for determining the boundary value of the device working state according to claim 1, characterized in that: The clustering of the description data to obtain at least one cluster includes: The description data is clustered by a density clustering model to obtain at least one cluster.

3. The method for determining the boundary value of the device working state according to claim 2, characterized in that: The method further comprises: determining parameters of the density clustering model, the parameters comprising a minimum number of samples and a domain distance; Determining the parameters of the density clustering model includes: Acquire training description data of a preset working state of a device and a plurality of candidate parameters of the density clustering model; Substituting each of the candidate parameters into the density clustering model, clustering the training description data using the density clustering model to obtain at least one candidate cluster and a clustering evaluation coefficient; Determining the sample variance of each of the candidate clusters; The parameters of the density clustering model are determined according to the clustering evaluation coefficient and the sample variance of all clusters obtained by clustering the density clustering model.

4. The method for determining the boundary value of the device working state according to claim 3, characterized in that: Determining the parameters of the density clustering model according to the clustering evaluation coefficient and the sample variance of all clusters obtained by clustering the density clustering model includes: Summing the clustering evaluation coefficient corresponding to each of the candidate parameters and the sample variances of all clusters corresponding to the candidate parameters to obtain a summed value; The candidate parameter corresponding to the smallest sum value is selected as the parameter of the density clustering model.

5. The method for determining the boundary value of the device working state according to claim 1, characterized in that: The step of determining the boundary value of the description data of the device in each working state according to the at least one cluster further includes: According to the first description data in the cluster corresponding to the current working state and the second description data corresponding to the previous working state, the boundary values ​​of the description data of the current working state and the previous working state are determined, wherein the first description data is the minimum description data in the cluster and the second description data is the maximum description data in the cluster.

6. The method for determining the boundary value of the device working state according to claim 1, characterized in that: The preset working states include device shutdown, device standby and device startup; The obtaining of description data of the working status of the device includes: Obtain the current data and / or amplitude data uploaded by the device, divide the description data according to preset time windows, select the maximum current data and / or amplitude data in each time window as the description data obtained in the time window, and use the time when the current data and / or the amplitude data were last uploaded in the time window as the acquisition time of the time window.

7. A device for determining a boundary value of a device working state, characterized in that: include: An input unit, used to obtain descriptive data of the working status of the device during a preset time period; A clustering unit, configured to cluster the description data to obtain at least one cluster, wherein each cluster corresponds to a working state of the device; an output unit, configured to determine a boundary value of the description data of each working state of the device according to the at least one cluster; The output unit is further used to determine, according to the number of preset working states, a cluster in the at least one cluster corresponding to each working state; The step of determining, according to the number of preset working states, a cluster in the at least one cluster corresponding to each working state one by one, comprises: Determine whether the number of clusters obtained is greater than the number of preset working states, If it is greater than, the average values ​​of the descriptive data in the clusters are sorted, and a preset number of clusters at the beginning or end of the sorted result are selected as target clusters, wherein the preset number is the number of preset working states, and the average value of the descriptive data of the target cluster is greater than the average value of the descriptive data in any cluster other than the target cluster in the at least two clusters.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method for determining the boundary value of the working state of the device as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for determining the boundary value of the working state of the device as claimed in any one of claims 1 to 6 are implemented.

Images