Trend analysis and assessment method for industrial time-series data

Abstract

Disclosed in the present application is a trend analysis and assessment method for industrial time-series data. The method comprises: collecting industrial data over a period of time, and verifying and pre-processing the industrial data, so as to obtain an industrial time-series data set; performing feature extraction on the industrial time-series data set, so as to acquire feature data of the industrial time-series data set; forming a plurality of data intervals in view of the number of intervals preset by a target user; assigning approximate values to industrial time-series data, wherein each approximate value is the upper limit value or lower limit value of a data interval within which the industrial time-series data falls, so as to form an approximate data set; and in view of the magnitude relationship between consecutive pieces of data in the approximate data set, acquiring linear encoded data sets, and determining a data trend on the basis of adjacent data in adjacent linear encoded data sets. When using the method, a target user only needs to preset the number of intervals, and the precision of trend analysis and assessment can be controlled by means of the number of intervals, thereby reducing the impact of data noise on a trend analysis and assessment result, and improving the stability of the method with regard to trend analysis and assessment.

Description

An industrial time series data trend judgment method

[0001] Cross-reference to Related Applications

[0002] The present application claims priority to the Chinese patent application No. 202411244873.7, filed on September 6, 2024, and entitled "An industrial time series data trend judgment method", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The present application belongs to the field of electronic digital data processing, and specifically relates to an industrial time series data trend judgment method. BACKGROUND

[0004] In industrial time series data analysis, the application of trend judgment is particularly crucial, because the upward trend or downward trend of industrial time series data usually represents changes in working conditions or abnormal states, etc. Industrial time series data judgment can help us understand and predict the changing trends and laws in industrial production processes.

[0005] The trend judgment of industrial time series data can be subdivided into four main types: upward trend, downward trend, horizontal trend, and oscillation trend. However, existing industrial time series data trend judgment methods can usually only identify several of the upward trend, downward trend, or horizontal trend, and lack effective identification means for oscillation trend.

[0006] Chinese patent CN202111192904.5 proposes a data trend detection method, which divides historical data in chronological order and performs curve fitting to obtain the slope, and then uses distribution symmetry detection to accurately and quickly judge the overall trend of the data. However, the trend judgment relies on multiple preset parameters (such as threshold), which requires higher professional knowledge and experience of the target user, not only increasing the difficulty of the target user using the method for trend judgment, but also reducing the stability of the method for trend judgment. Different target users may have different trend prediction results for the same set of industrial time series data. SUMMARY

[0007] The present application provides an industrial time series data trend judgment method to solve the problems of lack of effective identification means for oscillation trend and misjudgment of trend judgment results, as well as the problems of increased operation difficulty and poor stability of trend judgment results caused by setting parameters.

[0008] The technical solution adopted by the present application is as follows:

[0009] An industrial time series data trend judgment method, comprising:

[0010] S1: According to the set start time and end time, the industrial data in this time period is collected, and the collected industrial data is verified and preprocessed to obtain an industrial time series data set;

[0011] S2: Feature extraction is performed on the industrial time series data set to obtain feature data of the industrial time series data set, and the feature data includes maximum value and minimum value in the industrial time series data set; in combination with a target user preset interval segment number, a plurality of data intervals are formed;

[0012] S3: The industrial time series data set is processed, and according to the data interval where the data in the industrial time series data set is located, the industrial time series data is sequentially approximately valued, and the approximate value is the upper limit value or the lower limit value of the data interval, to form an approximately valued data set;

[0013] S4: According to a preset rule, in combination with the size relationship of adjacent data in the approximately valued data set, a linear state coding data set is obtained, and the data trend is determined according to adjacent data in the adjacent linear coding data set, and the data trend includes rising trend, falling trend, stable trend and oscillation trend.

[0014] In some embodiments of the application, in step S4, according to a preset rule, in combination with the size relationship of adjacent data in the approximately valued data set, a linear state coding data set is obtained, specifically:

[0015] An initial array of the linear state coding data set is established, and an initial value is set, and the data amount of the initial array is one less than that of the approximately valued data set;

[0016] According to the numerical size relationship of adjacent data in the approximately valued data set, the initial array is valued to form the linear state coding data set.

[0017] According to the numerical size relationship of adjacent data in the approximately valued data set, the initial array is valued, specifically:

[0018] If the numerical size of adjacent data in the approximately valued data set is equal, the corresponding data in the initial array is valued as 0;

[0019] If the next data value of the adjacent data in the approximately valued data set is larger, the corresponding data in the initial array is valued as 1;

[0020] If the previous data value of the adjacent data in the approximately valued data set is larger, the corresponding data in the initial array is valued as -1.

[0021] In some embodiments of the application, the data trend is determined according to adjacent data in the adjacent linear coding data set, specifically:

[0022] If the product of the adjacent data in the adjacent linear coding data set is -1, it is determined that the corresponding two intervals in the industrial time series data set are oscillation trends;

[0023] If the product of the adjacent data in the adjacent linear coding data set is not -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is a linear trend, and the linear trend includes an upward trend, a downward trend, and a stable trend.

[0024] If the product of the adjacent data in the adjacent linear coding data set is not -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is a linear trend, and according to a preset rule, the upward trend, the downward trend, and the stable trend are determined, specifically:

[0025] If the previous data of the adjacent data in the adjacent linear coding data set is 1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is an upward trend.

[0026] If the previous data of the adjacent data in the adjacent linear coding data set is -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is a downward trend.

[0027] If the previous data of the adjacent data in the adjacent linear coding data set is 0, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is a stable trend.

[0028] In some embodiments of the present application, an industrial time series data trend determination method further comprises:

[0029] S5: According to the determined data trend, the time interval statistics corresponding to the upward trend, the downward trend, the stable trend, and the oscillation trend are performed, and data statistical analysis is performed, including the time range length corresponding to each data trend and the proportion in the corresponding time period of the industrial data.

[0030] In some embodiments of the present application, in step S3, the approximate value is the upper limit value or the lower limit value of the data interval in which the approximate value is located, so as to form an approximate value data set, specifically:

[0031] The approximate value is the lower limit value of the data interval in which the approximate value is located;

[0032] Alternatively, the approximate value is the upper limit value of the interval in which the approximate value is located;

[0033] Alternatively, the proximity between the data in the industrial time series data set and the upper limit value and the lower limit value of the corresponding data interval is determined, and if the data is more close to the upper limit value, the approximate value is the upper limit value, otherwise the approximate value is the lower limit value.

[0034] In some embodiments of the present application, in step S2, the target user preset interval segment number is an integer and is greater than or equal to 2.

[0035] In some embodiments of this application, step S1 involves verifying the industrial data, specifically as follows:

[0036] Perform integrity verification on industrial data to verify its continuity;

[0037] And / or, perform volatility verification on industrial data to verify the validity of the industrial data.

[0038] In some embodiments of this application, step S1 involves preprocessing the collected industrial data, specifically as follows:

[0039] If the industrial data is found to be missing, interpolation processing is performed on the industrial data;

[0040] If outliers are found in the industrial data, smooth the data and / or remove the outliers.

[0041] Some embodiments of this application also disclose an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0042] Memory, used to store computer programs;

[0043] A processor, when executing a program stored in memory, implements the method as described in the embodiments of this application.

[0044] Some embodiments of this application also disclose a computer non-volatile readable storage medium having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the methods of the embodiments described above in this application.

[0045] Due to the adoption of the above technical solution, the beneficial effects achieved by this application are as follows:

[0046] 1. In this application, within the range of the maximum and minimum values ​​of the industrial time-series dataset, the dataset is divided into multiple data intervals based on a preset number of intervals by the target user. Approximate values ​​are then taken from the data intervals in which the data falls, with the approximate values ​​being the upper or lower limit of the corresponding data interval, thus forming an approximate value dataset. Based on preset rules and the size relationship between adjacent data in the approximate value dataset, a linear state-coded dataset is obtained. The data trend is then determined based on adjacent data in the adjacent linearly coded datasets.

[0047] When collecting industrial data, due to the collection error of the data collection sensor or the occurrence of environmental noise, the industrial time series data obtained will inevitably have data noise jitter, which affects the accuracy of subsequent trend analysis. For example, due to data noise jitter, a smooth trend is misjudged as an oscillating trend, or multiple preset parameters (such as threshold) need to be set, and the preset parameters are subject to the professional knowledge and experience of the target user, which may result in different trend analysis accuracy for different industrial time series data or different target users, and increases the workload of the target user and the complexity of the trend analysis of the industrial time series data.

[0048] The present application can reduce the small jitter of the collected industrial time series data caused by the collection error of the data collection sensor or environmental noise through the approximate value processing of the industrial time series data set, facilitate the subsequent trend analysis, and improve the accuracy of the trend analysis.

[0049] When using the method in the present application to analyze the trend of industrial time series data, the target user only needs to preset the interval segment number, without inputting other parameters, and the accuracy of the trend analysis can be controlled by the interval segment number. For example, increasing the interval segment number will increase the accuracy of the trend analysis. The interval segment number is less affected by different industrial time series data, and the same interval segment number can achieve the same trend analysis accuracy for different industrial time series data. The recognition process of the user for the industrial time series data is reduced, and the interval segment number needs to be set according to the requirement of the trend analysis accuracy, which improves the accuracy of the method in the present application. The stability of the trend analysis of the method is improved, and the same trend analysis result can be obtained for the same industrial time series data by different target users, which reduces the dependence on the knowledge and experience of the target user.

[0050] 2. In the present application, according to the preset rule, the size relationship of the adjacent data in the approximate value data set is combined to obtain the linear state coding data set, and the trend of the data is judged according to the adjacent data in the adjacent linear coding data set. In the prior art, the trend of the industrial time series data is directly judged, which not only needs to compare the size relationship of the adjacent data in the industrial time series data, but also needs to calculate the difference between the adjacent data, and the difference is compared with the preset threshold to reduce the influence of data noise, otherwise the trend will be misjudged. That is, the prior art has the problems of complex operation, long calculation time, and high occupation of calculation resources. When judging the trend, the data in the approximate value data set is subjected to approximate value, which reduces the influence of data noise, and only the size relationship comparison of the adjacent data in the approximate value data set is needed to obtain the linear state coding data set, and the trend of the data is judged according to the adjacent data in the adjacent linear coding data set. The trend judgment step is more simple, the rate of trend judgment is improved, and the occupation of calculation resources is reduced.

[0051] Moreover, the present application can judge the rising trend, the falling trend, the stable trend and the oscillation trend, covering all the trend modes of the industrial time series data, facilitating the identification of the working condition abnormal state according to the trend of the industrial time series data, and helping to find and adjust the problems in the production process in time. BRIEF DESCRIPTION OF DRAWINGS

[0052] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The schematic embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0053] Fig. 1 is a flowchart of the industrial time series data trend judgment method according to some embodiments of the present application;

[0054] Fig. 2 is a contrast diagram of the industrial time series data curve and the approximate value data curve after the industrial time series data is subjected to approximate value in some embodiments of the present application;

[0055] Fig. 3 is the time length corresponding to the four data trends after the trend of the industrial time series data is judged in some embodiments of the present application;

[0056] Fig. 4 is the time proportion corresponding to the four data trends after the trend of the industrial time series data is judged in some embodiments of the present application;

[0057] Fig. 5 is a block diagram of an electronic device according to some embodiments of the present application. DETAILED DESCRIPTION

[0058] In order to more clearly explain the overall concept of the present application, the following will be described in detail in an exemplary manner with reference to the drawings.

[0059] Many specific details are set forth in the following description in order to provide a thorough understanding of the present application. However, the present application can be practiced according to other embodiments that can not be described in detail herein, and the present application is not limited to the embodiments described herein.

[0060] As shown in FIG. 1, an industrial time series data trend analysis method comprises the following steps:

[0061] S1: According to the set start time and end time, collect the industrial data in this time period, and verify and preprocess the collected industrial data to obtain an industrial time series data set;

[0062] This step aims to process the industrial time series data and analyze the trend of the industrial time series data in different time intervals. The industrial data recorded by the industrial sensor is generally stored in the form of industrial data and corresponding time, and the time interval between adjacent two points of industrial data is the sampling interval. In order to facilitate the calculation of the method, the order of the data in the industrial data is generally coded in sequence, and the sampling interval is recorded separately.

[0063] In this step, the industrial data is first verified to determine the completeness and validity of the industrial data. Avoiding the trend analysis error caused by incomplete industrial data, or the error of the time interval corresponding to each trend after trend analysis. In addition, by determining the validity of the industrial data, the valid industrial data is analyzed for trend to provide accurate trend analysis results, so as to accurately distinguish the industrial abnormality. When the industrial data is invalid, it does not occupy the trend analysis resources, and the target user can be reminded to check and maintain the industrial data sensor. Avoiding the invalidation of industrial data caused by sensor abnormality, avoiding the error prediction of trend and the error prediction of industrial operation condition based on invalid industrial data.

[0064] Secondly, this step will clear the time data, encode the order of each data in the industrial data, and record the sampling time interval. In this way, the data can be traversed and processed by the method, so as to facilitate the trend prediction of industrial time series data. Avoiding the need to change the time interval in the method when processing different industrial time series data, improving the universality of the method and reducing the change operation of the method.

[0065] Specifically, the industrial time series data set is recorded as X={x1,…,x t ,…,x m}, t∈[1,m], wherein x t is any data in the industrial time series data set, and t is the time interval between adjacent two points of x tThe arrangement order coding in the industrial time series dataset, and m is the total number of data in the industrial time series dataset.

[0066] It should be noted that the industrial data in the present application includes but is not limited to temperature, pressure, flow, speed, power, current, voltage, content and the like. Through the measurement and monitoring of these parameters, the trend of the industrial production process can be more accurately mastered, thereby improving the production efficiency, optimizing the product quality, reducing the energy consumption and the like.

[0067] S2: Feature extraction is performed on the industrial time series dataset to obtain feature data of the industrial time series dataset, and the feature data includes the maximum value and the minimum value in the industrial time series dataset; in combination with a target user preset interval segment number, a plurality of data intervals are formed. As shown in FIG. 2, the target user preset interval segment number is 6.

[0068] This step aims to form a plurality of data intervals according to the feature data in the industrial time series dataset, i.e. the maximum value and the minimum value in the industrial time series dataset, through the target user preset interval segment number. The method in the present application automatically extracts the feature data without the need for user input.

[0069] The user only needs to preset the interval segment number without inputting other parameters, and the precision of the trend research and judgment can be controlled through the interval segment number, for example, the higher the interval segment number, the higher the trend research and judgment precision. The interval segment number is less affected by different industrial time series data, and the same trend research and judgment precision can be achieved by using the same interval segment number for different industrial time series data. The identification process of the user on the industrial time series data is reduced, and the interval segment number is set according to the requirement of the trend research and judgment precision, thereby improving the precision of the use of the method in the present application. Moreover, the stability of the trend research and judgment of the method is improved, and the same trend research and judgment result can be obtained for the same industrial time series data by different target users, thereby reducing the dependence on the knowledge and experience of the target user.

[0070] S3: The industrial time series dataset is processed, and the industrial time series data is approximately valued according to the data interval where the data in the industrial time series dataset is located, and the approximate value is the upper limit value or the lower limit value of the data interval, so as to form an approximate value dataset, as shown in the approximate value data in FIG. 2.

[0071] This step aims to approximately value the industrial time series data according to the data interval where the data in the industrial time series dataset is located according to the plurality of data intervals formed in step S2. The approximate value processing of the industrial time series dataset can reduce the slight jitter of the industrial time series data caused by the collection error of the data collection sensor or the environmental noise, facilitate the trend research and judgment in the subsequent steps, and improve the accuracy of the trend research and judgment.

[0072] It should be noted that according to the data interval in which the industrial time series data is located, the upper limit value of the data interval can be taken at the same time, or the lower limit value of the data interval can be taken at the same time. The present application does not make any limitation.

[0073] S4: According to the preset rule, the size relationship of the adjacent data in the approximate value data set is obtained, and the linear state coding data set is obtained. The adjacent data in the adjacent linear coding data set is used to judge the data trend, and the data trend includes rising trend, falling trend, stable trend and oscillation trend.

[0074] The present step aims to compare the size relationship of the adjacent data in the approximate value data set, obtain the linear state coding data set, and judge the data trend according to the adjacent data in the adjacent linear coding data set.

[0075] When the trend is analyzed, the data in the approximate value data set is subjected to approximate value, which reduces the influence of data noise, makes the step more simple, improves the rate of trend analysis, and reduces the occupation of computing resources. Moreover, the present application can analyze the rising trend, falling trend, stable trend and oscillation trend, which covers all the trend modes of industrial time series data, and is convenient for identifying the abnormal state of working condition according to the trend of industrial time series data, which helps to find and adjust the problems in production process in time.

[0076] In some embodiments of the present application, in step S4, according to the preset rule, the size relationship of the adjacent data in the approximate value data set is obtained, and the linear state coding data set is obtained. Specifically,

[0077] An initial array of the linear state coding data set is established, and an initial value is set. The data amount of the initial array is one less than that of the approximate value data set;

[0078] According to the value size relationship of the adjacent data in the approximate value data set, the initial array is valued, and the linear state coding data set is formed.

[0079] The present step aims to establish the initial array of the linear state coding data set, so as to store the linear state coding data. Each linear state coding data represents the value size relationship of the adjacent data in the approximate value data set, so the data amount of the initial array is one less than that of the approximate value data set. At the same time, it is convenient to analyze the trend according to the relationship of the adjacent data in the linear state coding data set.

[0080] Specifically, the approximate value data set is recorded as Y={y1,…,y t ,…,y m},t∈[1,m], wherein y t is the industrial time series data xt Approximate value data after approximation.

[0081] The linear state encoding data set is recorded as Z={z 1,2 ,…,z t,t+1 ,…,z m-1,m}, t∈[1, m], wherein z t,t+1 is obtained according to the value size relationship of y t and y t+1 .

[0082] In some embodiments of the present application, the initial array is valued according to the value size relationship of adjacent data in the approximate value data set, specifically:

[0083] If the values of adjacent data in the approximate value data set are equal, the corresponding data in the initial array is valued as 0;

[0084] If the value of the latter data of adjacent data in the approximate value data set is relatively large, the corresponding data in the initial array is valued as 1;

[0085] If the value of the former data of adjacent data in the approximate value data set is relatively large, the corresponding data in the initial array is valued as -1.

[0086] The value size of adjacent data in the approximate value data set includes three cases, i.e. the value of the latter data is equal to the value of the former data, the value of the latter data is relatively large, and the value of the former data is relatively large. According to the one-to-one correspondence of the three cases, the corresponding data in the initial array is valued as 0, 1 and -1, so as to facilitate subsequent trend judgment according to adjacent data in the adjacent linear encoding data set.

[0087] Specifically, if y t =y t+1 , z t,t+1 is valued as 0; if y t >y t+1 , z t,t+1 is valued as -1; if y t <y t-1 , z t,t+1 is valued as 1.

[0088] It should be noted that the present application does not limit the one-to-one correspondence of the values of the corresponding data in the initial array according to the three cases, i.e. the value of the latter data is equal to the value of the former data, the value of the latter data is relatively large, and the value of the former data is relatively large. The values of the corresponding data in the initial array can be -1, 0 and 1, or 1, 2 and 3, as long as the subsequent data trend judgment according to adjacent data in the adjacent linear encoding data set is adjusted accordingly.

[0089] In some embodiments of the present application, the trend of data is determined according to the adjacent data in the adjacent linear coding data set, specifically:

[0090] If the product of the adjacent data in the adjacent linear coding data set is -1, it is determined that the corresponding interval in the industrial time series data set is an oscillation trend.

[0091] If the product of the adjacent data in the adjacent linear coding data set is not -1, it is determined that the previous data in the corresponding two intervals in the industrial time series data set is a linear trend, and the linear trend includes an upward trend, a downward trend and a stable trend.

[0092] It should be noted that in the corresponding approximate value data set, the next data is equal to the previous data, the value of the next data is relatively large, and the value of the previous data is relatively large. According to the one-to-one correspondence of the three conditions, the corresponding data in the initial array is assigned as 0, 1 and -1.

[0093] In this step, the oscillation trend and the linear trend are distinguished according to the product of the adjacent data in the adjacent linear coding data set. When the product of the adjacent data in the adjacent linear coding data set is -1, it represents that the trend of the previous industrial time series adjacent data is opposite to that of the next industrial time series adjacent data. Specifically, the previous industrial time series adjacent data is an upward trend, and the next industrial time series adjacent data is a downward trend, or the previous industrial time series adjacent data is a downward trend, and the next industrial time series adjacent data is an upward trend. Therefore, when the product of the adjacent data in the adjacent linear coding data set is -1, it is determined to be an oscillation trend.

[0094] Specifically, when z t,t+1 = -1, z t+1,t+2 = 1, it represents that x t,t+1 ~ x t+1,t+2 in the industrial time series data set is an upward trend, x t ~ x t+1 is a downward trend, and x t+1 ~ x t+2 is an oscillation trend. Or when z t = -1, z t+2 = 1, it represents that x t,t+1 ~ x t+1,t+2 in the industrial time series data set is a downward trend, x t ~ x t+1 is an upward trend, and x t+1 ~ x t+2 is an oscillation trend. t t+2

[0095] ​​It should be noted that the adjacent linear coding data set is traversed to make a trend analysis on the industrial time series data set. When it is determined that x t ~~x t+2 The interval is an oscillation trend, and the next step is to perform a t+2 operation on t, that is, to make a trend analysis by z t+2,t+3 *z t+3,t+4 to avoid repeated analysis of the same interval.

[0096] When the initial array is assigned by other rules, the corresponding adjustment of the calculation principle needs to be made, which is not limited in the present application.

[0097] Further, if the product of the adjacent data in the adjacent linear coding data set is not-1, it is determined that the first interval in the corresponding two intervals in the industrial time series data set is a linear trend, and according to the preset rule, the rising trend, the falling trend and the stable trend are analyzed, which are as follows:

[0098] If the previous data of the adjacent data in the adjacent linear coding data set is 1, it is determined that the first interval in the corresponding two intervals in the industrial time series data set is a rising trend;

[0099] If the previous data of the adjacent data in the adjacent linear coding data set is-1, it is determined that the first interval in the corresponding two intervals in the industrial time series data set is a falling trend;

[0100] If the previous data of the adjacent data in the adjacent linear coding data set is 0, it is determined that the first interval in the corresponding two intervals in the industrial time series data set is a stable trend.

[0101] Specifically, when z t,t+1 *z t+1,t+2 ≠-1, z t,t+1 =1 indicates that the interval x t ~~x t+1 in the industrial time series data set is a rising trend; z t,t+1 =0 indicates that the interval x t ~~x t+1 in the industrial time series data set is a stable trend; and z t,t+1 =-1 indicates that the interval x t ~x t+1 in the industrial time series data set is a falling trend.

[0102] It should be noted that when z t,t+1 *z t+1,t+2 ≠-1, only the first interval (x t ~~x t+1 interval) in the corresponding two intervals in the industrial time series data set is determined. The next step is to perform a t+1 operation on t, that is, to make a trend analysis by z t+1,t+2 *z t+2,t+3Carrying out trend analysis.

[0103] In some embodiments of the present application, as shown in FIG. 1, the industrial time series data trend analysis method further comprises:

[0104] S5: According to the judged data trend, the time interval statistics corresponding to the rising trend, the falling trend, the stable trend and the oscillation trend are carried out, and data statistical analysis is carried out, including the time range length corresponding to each data trend and the proportion in the corresponding time period of industrial data.

[0105] This step aims to count the time range corresponding to the rising trend, the falling trend, the stable trend and the oscillation trend, and calculate the time range length corresponding to the four data trends and the proportion in the corresponding time period of industrial data, which is more convenient for the target user to comprehensively analyze the trend analysis data, and helps to find and adjust the problems in the production process.

[0106] In some embodiments of the present application, the industrial time series data trend analysis method further comprises: in step S3, the approximate value is the upper limit value or the lower limit value of the data interval, to form an approximate value data set, specifically:

[0107] The approximate value is the lower limit value of the data interval;

[0108] Or, the approximate value is the upper limit value of the interval;

[0109] Or, the closeness between the data in the industrial time series data set and the upper limit value and the lower limit value of the corresponding data interval is determined, if the data is more close to the upper limit value, the approximate value is the upper limit value, otherwise the approximate value is the lower limit value.

[0110] This step aims to approximate the industrial time series data, and uniformly select the upper limit value of the interval or the lower limit value of the interval, which is not limited in the present application, and a uniform approximation rule can be adopted.

[0111] Or, the closeness between the data in the industrial time series data set and the upper limit value and the lower limit value of the corresponding data interval is determined, if the data is more close to the upper limit value, the approximate value is the upper limit value, otherwise the approximate value is the lower limit value. By determining the closeness between the industrial time series data and the upper limit value and the lower limit value of the corresponding data interval, the error caused by increasing or decreasing when approximating the industrial time series data is avoided, which leads to the failure of trend analysis.

[0112] In some embodiments of the present application, in step S2, the target user pre-set interval segment number is an integer and greater than or equal to 2.

[0113] When the target user presets the interval segment number as 1, the accuracy of the trend judgment is reduced. That is, the industrial time series data is approximated as a maximum value or a minimum value, the approximation error is increased, and the trend judgment result is affected. The preset interval segment number is an integer, which is more convenient for dividing the industrial time series data set into an integer number of data intervals according to the numerical value, and approximating all industrial time series data with similar approximation accuracy. Avoiding the increase of approximation error of industrial time series data in a certain data interval, and affecting the trend judgment result.

[0114] The target user can control the accuracy of the trend judgment through the interval segment number. For example, increasing the interval segment number, the higher the accuracy of the trend judgment. The interval segment number is less affected by different industrial time series data. By corresponding the interval segment number and the accuracy of the trend judgment, the same trend judgment accuracy can be achieved by using the same interval segment number for different industrial time series data. The user's identification process of the industrial time series data is reduced, and only needs to set the interval segment number according to the requirement of the accuracy of the trend judgment, which improves the accuracy of the method used in the present application. And increase the stability of the trend judgment of the method. For the same industrial time series data, different target users can get the same trend judgment result, reducing the dependence on the knowledge and experience of the target user.

[0115] In some embodiments of the present application, in step S1, the industrial data is verified, specifically:

[0116] The integrity of the industrial data is verified to verify the continuity of the industrial data;

[0117] And / or, the volatility of the industrial data is verified to verify the validity of the industrial data.

[0118] Through the integrity verification, it is verified that the industrial data is continuous without missing values or intervals. The volatility of the industrial data is detected. In order to ensure that the industrial data is obvious and effective, the industrial data should show a certain volatility or variability, that is, to avoid the situation that the data points are static or constant on the time axis (meaning that the industrial data is invalid, and the industrial data acquisition sensor fails). Volatility can be evaluated by calculating the statistical quantity (such as variance, standard deviation) of the industrial data. Of course, volatility can also be calculated in other ways, which is not limited.

[0119] In some embodiments of the present application, in step S1, the collected industrial data is preprocessed, specifically:

[0120] If it is verified that the industrial data has missing data, the industrial data is interpolated;

[0121] If it is verified that the industrial data has abnormal values, the industrial data is smoothed and / or the abnormal values are removed.

[0122] The step is to process abnormal industrial data, so as to facilitate subsequent data trend research. Avoid the influence of data missing or abnormal value on the trend item research result.

[0123] As a specific embodiment of the present application, the amplitude data of the drive shaft of a certain coal mine ventilator is taken as a test case. The industrial data is sampled at a time interval of about 5 seconds, and the entire data set covers 1947 sampling points. The approximate value curve and the trend research result are shown in FIGS. 2 to 4. The method proposed in the present application can accurately identify the duration of each trend in the industrial time series data and the relative proportion in the entire time period, so as to more objectively understand and predict the dynamic changes in the industrial system, facilitate timely discovery of problems in the production process and adjustment.

[0124] In addition, some embodiments of the present application also provide an electronic device, as shown in FIG. 5, which includes a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502 and the memory 503 complete mutual communication through the communication bus 504,

[0125] The memory 503 is used to store computer programs.

[0126] The processor 501 is used to execute the programs stored in the memory 503, and realize the method in the above-mentioned embodiments.

[0127] The communication bus mentioned in the above terminal can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0128] The communication interface is used for communication between the above-mentioned terminal and other devices.

[0129] The memory can include a random access memory (RAM), and can also include a non-volatile memory, for example, at least one disk memory. In some embodiments, the memory can also be at least one storage device located away from the aforementioned processor.

[0130] The processor described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0131] In a further embodiment provided in the present application, a computer nonvolatile readable storage medium is also provided, which stores instructions, and when the instructions run on a computer, the computer executes the method in the above embodiments.

[0132] The places not described in the present application can be implemented or referred to the existing technology.

[0133] Each of the embodiments in the present specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each of the embodiments mainly describes the difference from other embodiments.

[0134] The above is only an embodiment of the present application, and is not used to limit the present application. The present application can have various changes and modifications for those skilled in the art. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of claims of the present application.

Claims

1. An industrial time series data trend analysis method, characterized in that, according to the set start time and end time, the industrial data in this time period is collected, and the collected industrial data is verified and preprocessed to obtain an industrial time series data set; feature extraction is performed on the industrial time series data set to obtain feature data of the industrial time series data set, the feature data including maximum and minimum values in the industrial time series data set; a plurality of data intervals are formed in combination with a target user preset interval number; processing the industrial time series data set, according to the data interval where the data in the industrial time series data set is located, the industrial time series data is sequentially approximated, and the approximated value is the upper limit value or the lower limit value of the data interval, to form an approximated value data set; according to a preset rule, in combination with the size relationship of adjacent data in the approximated value data set, a linear state coding data set is obtained; judging data trend according to adjacent data in the adjacent linear coding data set, the data trend including upward trend, downward trend, stable trend, and oscillation trend.

2. The method of claim 1, wherein, According to a preset rule, in combination with the size relationship of adjacent data in the approximated value data set, a linear state coding data set is obtained, including: According to a preset rule, in combination with the size relationship of adjacent data in the approximated value data set, according to the three cases of the latter data equal to the former data, the latter data value being relatively large, and the former data value being relatively large, the linear state coding data set is obtained.

3. The method of claim 1, wherein: The method further includes: if the trend of the former industrial time series adjacent data is opposite to that of the latter industrial time series adjacent data, it is determined as oscillation trend, and if the trend of the former industrial time series adjacent data is the same as that of the latter industrial time series adjacent data, it is determined as linear trend, the linear trend including upward trend, downward trend, and stable trend.

4. The method of claim 1, wherein, According to a preset rule, in combination with the size relationship of adjacent data in the approximated value data set, a linear state coding data set is obtained, specifically: an initial array of the linear state coding data set is established, and an initial value is set, the data amount of the initial array being one less than that of the approximated value data set; according to the numerical size relationship of adjacent data in the approximated value data set, the initial array is valued to form the linear state coding data set.

5. The method of claim 4, wherein, According to the numerical size relationship of adjacent data in the approximated value data set, the initial array is valued, specifically: if the numerical size of adjacent data in the approximated value data set is equal, the corresponding data in the initial array is valued as 0; if the latter data value of adjacent data in the approximated value data set is larger, the corresponding data in the initial array is valued as 1; if the former data value of adjacent data in the approximated value data set is larger, the corresponding data in the initial array is valued as -1.

6. The method of claim 5, wherein: According to adjacent data in the adjacent linear coding data set, the data trend is judged, specifically: the product of adjacent data in the adjacent linear coding data set is used to distinguish oscillation trend and linear trend, the linear trend including upward trend, downward trend, and stable trend.

7. The method of claim 6, wherein: The discrimination of the oscillation trend and the linear trend is performed according to the product of the adjacent data in the adjacent linear coding data set, specifically: If the product of the adjacent data in the adjacent linear coding data set is -1, it is determined that the corresponding two intervals in the industrial time series data set are the oscillation trend; If the product of the adjacent data in the adjacent linear coding data set is not -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is the linear trend, and the linear trend includes the upward trend, the downward trend and the stable trend.

8. The method of claim 7, wherein: If the product of the adjacent data in the adjacent linear coding data set is not -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is the linear trend, and according to a preset rule, the upward trend, the downward trend and the stable trend are judged, specifically: If the previous data of the adjacent data in the adjacent linear coding data set is 1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is the upward trend; If the previous data of the adjacent data in the adjacent linear coding data set is -1, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is the downward trend; If the previous data of the adjacent data in the adjacent linear coding data set is 0, it is determined that the previous interval in the corresponding two intervals in the industrial time series data set is the stable trend.

9. The method of claim 1, wherein: Further comprising: According to the judged data trend, the time interval statistics corresponding to the upward trend, the downward trend, the stable trend and the oscillation trend are performed, and data statistical analysis is performed, including the time range length corresponding to each data trend and the proportion in the corresponding time period of the industrial data.

10. The method of claim 1, wherein: The approximate value is the upper limit value or the lower limit value of the data interval to form an approximate value data set, specifically: The approximate value is the lower limit value of the data interval; Or, the approximate value is the upper limit value of the interval; Or, the closeness between the data in the industrial time series data set and the upper limit value and the lower limit value of the corresponding data interval is determined, if the data is closer to the upper limit value, the approximate value is the upper limit value, otherwise the approximate value is the lower limit value.

11. The industrial time series data trend judgment method according to claim 1, characterized in that: The target user preset interval segment number is an integer and is greater than or equal to 2.

12. The method of claim 1, wherein: The industrial data is verified, specifically: The completeness of the industrial data is verified to verify the continuity of the industrial data; And / or, the volatility of the industrial data is verified to verify the effectiveness of the industrial data.

13. The method of claim 12, wherein: The completeness of the industrial data is verified to verify the continuity of the industrial data, specifically: If there is no missing value or interval in the industrial data, it is verified that the industrial data is continuous.

14. The method of claim 12, wherein: The volatility of the industrial data is verified to verify the effectiveness of the industrial data, specifically: The volatility of the industrial data is evaluated according to the statistical quantity of the industrial data, and the statistical quantity is variance or standard deviation.

15. The method of claim 1, wherein: The collected industrial data is preprocessed, specifically: If it is verified that the industrial data has data missing, the industrial data is interpolated; If it is verified that the industrial data has abnormal values, the industrial data is smoothed and / or the abnormal values are removed.

16. The method of claim 1, wherein: The interval segment number is used to control the accuracy of the trend analysis.

17. The method of claim 1, wherein: The data trend is determined according to adjacent data in the adjacent linear coding data set, and specifically: The adjacent linear coding data set is traversed to analyze the trend of the industrial time series data set.

18. The method of claim 1, wherein: The industrial data at least includes temperature, pressure, flow, speed, power, current, voltage and content.

19. An electronic device, comprising: The processor, the communication interface, the memory and the communication bus are included, wherein the processor, the communication interface and the memory complete the communication among each other through the communication bus; The memory is configured to store a computer program; The processor is configured to execute the program stored on the memory, and realize the method in any one of claims 1-18. 20.A computer non-volatile readable storage medium, having instructions stored thereon, which, when executed by one or more processors, cause the processors to perform the method in any one of claims 1-18.

Images