Steady state detection method and system of industrial system, and electronic equipment

Abstract

The invention provides a steady state detection method and system of an industrial system, and electronic equipment. The method comprises the following steps: carrying out steady state detection on an industrial system according to first operation time sequence data, collected by a preset sliding window, of the industrial system; under the condition that a first steady-state detection result obtained according to the first operation time sequence data represents steady-state operation of the industrial system, gradually expanding a right boundary of the preset sliding window to obtain a longest sliding window; and taking the right boundary of the longest sliding window as the left boundary of the sliding window adopted by the next steady-state detection, and carrying out the next steady-state detection. According to the method, the dynamic sliding window is adopted for steady-state detection, the window size and the sliding step length of the dynamic sliding window adopted in each steady-state detection are related to the steady-state duration, and more valuable steady-state data are reserved based on the operation time sequence data collected by the dynamic sliding window, so that the effectiveness of the steady-state detection is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of steady-state detection technology, and in particular to a steady-state detection method, system, and electronic equipment for an industrial system. Background Art

[0002] For industrial systems, taking coal-fired boiler systems as an example, controlling and modeling the combustion process of coal-fired boiler systems is of great practical significance to the operating efficiency of coal-fired boiler systems, reducing coal consumption and reducing harmful gas emissions. However, the overall performance of coal-fired boiler systems is nonlinear, time-varying, uncertain and incomplete, and the multivariables in the units of coal-fired boiler systems have extremely complex coupling relationships, which makes it extremely difficult to control and model the boiler combustion process. Steady-state detection of the operation process of coal-fired boiler systems is a key link in control and modeling optimization.

[0003] When performing steady-state detection, a sliding window is generally used to collect the operating time data of the coal-fired boiler system to detect the stability of the operating time data. However, the duration of different steady states is different. Using an inappropriate sliding window for steady-state detection will lose a lot of valuable steady-state data, affecting the accuracy of steady-state detection. Summary of the invention

[0004] The technical problem to be solved by the present disclosure is to overcome the above-mentioned defects in the prior art and to provide a steady-state detection method, system and electronic equipment for an industrial system.

[0005] The present invention solves the above technical problems through the following technical solutions:

[0006] In a first aspect, a steady-state detection method for an industrial system is provided, comprising:

[0007] Performing steady-state detection on the industrial system according to first operating time series data of the industrial system collected by a preset sliding window;

[0008] In the case where the first steady-state detection result obtained according to the first operating time series data represents the steady-state operation of the industrial system, the right boundary of the preset sliding window is gradually expanded to obtain the longest sliding window; wherein the second steady-state detection result obtained according to the second operating time series data represents the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third operating time series data represents the non-steady-state operation of the industrial system; the second operating time series data is collected according to the longest sliding window, and the third operating time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window;

[0009] The right boundary of the longest sliding window is used as the left boundary of the sliding window used for the next steady-state detection, and the next steady-state detection is performed.

[0010] Optionally, when the first steady-state detection result indicates that the industrial system is operating in a non-steady-state, the left boundary of the preset sliding window is moved by a sliding step and used as the left boundary of the sliding window used for the next steady-state detection.

[0011] Optionally, steady-state detection, including:

[0012] When the running time series data meets the steady-state condition, the steady-state detection result is that the industrial system is running in a steady state; wherein the steady-state condition includes at least one of the following:

[0013] The operating status of the units included in the industrial system has not changed; the operating status is determined according to the operating sequence data;

[0014] The variation range of the parameter value of each parameter included in the running time series data is within the respective preset range;

[0015] The variation trends of the parameter values ​​of the parameters included in the runtime data are consistent with the respective preset trends.

[0016] Optionally, the step of gradually extending the right boundary of the preset sliding window includes:

[0017] When the steady-state condition includes that the parameter values ​​of the parameters contained in the running time series data are within the respective preset ranges and the changing trends of the parameter values ​​of the parameters contained in the running time series data conform to the respective preset trends, a steady-state detection is performed each time the right boundary is extended; wherein, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the changing trend is determined according to the running time series data stored in a tree array.

[0018] Optionally, the running state is determined according to the encoding result of the running timing data;

[0019] and / or, performing median filtering on the running time series data before steady-state detection;

[0020] And / or, a Mann-Kendall algorithm is used to determine a change trend of a parameter value of each parameter.

[0021] In a second aspect, a steady-state detection system for an industrial system is provided, comprising:

[0022] A collection module, used for performing steady-state detection on the industrial system according to first running time series data of the industrial system collected by a preset sliding window;

[0023] A dynamic sliding window module, used for gradually expanding the right boundary of the preset sliding window to obtain a longest sliding window when the first steady-state detection result obtained according to the first running time series data represents the steady-state operation of the industrial system, wherein the second steady-state detection result obtained according to the second running time series data represents the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third running time series data represents the non-steady-state operation of the industrial system; the second running time series data is collected according to the longest sliding window, and the third running time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window;

[0024] The steady-state detection module is used to use the right boundary of the longest sliding window as the left boundary of the sliding window used for the next steady-state detection to perform the next steady-state detection.

[0025] Optionally, the dynamic sliding window module is further used to move the left boundary of the preset sliding window by a sliding step size as the left boundary of the sliding window used for the next steady-state detection when the first steady-state detection result indicates that the industrial system is operating in a non-steady-state.

[0026] Optionally, the steady-state detection module is specifically configured to determine that the steady-state detection result is steady-state operation of the industrial system when the running time series data satisfies a steady-state condition; wherein the steady-state condition includes at least one of the following:

[0027] The operating status of the units included in the industrial system has not changed; the operating status is determined according to the operating sequence data;

[0028] The variation range of the parameter value of each parameter included in the running time series data is within the respective preset range;

[0029] The variation trends of the parameter values ​​of the parameters included in the runtime data are consistent with the respective preset trends.

[0030] Optionally, the step of gradually extending the right boundary of the preset sliding window includes:

[0031] When the steady-state condition includes that the parameter values ​​of the parameters contained in the running time series data are within the respective preset ranges and the changing trends of the parameter values ​​of the parameters contained in the running time series data conform to the respective preset trends, a steady-state detection is performed each time the right boundary is extended; wherein, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the changing trend is determined according to the running time series data stored in a tree array.

[0032] In a third aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and used to run on the processor, wherein when the processor executes the computer program, the steady-state detection method for an industrial system described in any one of the first aspects is implemented.

[0033] On the basis of being in accordance with the common sense in the art, the above-mentioned preferred conditions can be arbitrarily combined to obtain the preferred embodiments of the present disclosure.

[0034] The positive progressive effect of the present disclosure lies in that the present disclosure adopts a dynamic sliding window for steady-state detection. The window size and sliding step length of the dynamic sliding window adopted in each steady-state detection are related to the duration of the steady state. The running time data collected based on the dynamic sliding window retains more valuable steady-state data, thereby improving the effectiveness of steady-state detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flowchart of a steady-state detection method for an industrial system provided by an exemplary embodiment of the present disclosure;

[0036] Figure 2 A schematic diagram of determining the longest sliding window in a steady-state detection process of an industrial system provided by an exemplary embodiment of the present disclosure;

[0037] Figure 3 A flow chart of another steady-state detection method for an industrial system provided by an exemplary embodiment of the present disclosure;

[0038] Figure 4 A module schematic diagram of a steady-state detection system for an industrial system provided by an exemplary embodiment of the present disclosure;

[0039] Figure 5 The present invention provides a schematic structural diagram of an electronic device according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0040] The present disclosure is further described below by way of examples, but the present disclosure is not limited to the scope of the examples.

[0041] Prefixes such as "first" and "second" are used in the embodiments of the present disclosure only to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. The use of prefixes such as ordinal numbers to distinguish description objects in the embodiments of the present disclosure does not constitute a limitation on the described objects. For the statement of the described objects, please refer to the description in the context of the claims or embodiments, and no unnecessary limitation should be constituted due to the use of such prefixes. In addition, in the description of the present embodiment, unless otherwise specified, the meaning of "plurality" is two or more.

[0042] Figure 1 A flowchart of a steady-state detection method for an industrial system provided by an exemplary embodiment of the present disclosure, the steady-state detection method comprising the following steps:

[0043] Step 101: Perform a steady-state detection on the industrial system according to first operating time series data of the industrial system collected by a preset sliding window.

[0044] During the operation of the industrial system, the operating sequence data of the industrial system can be traversed by moving the sliding window. The window size of the preset sliding window can be set according to actual conditions, for example, the minimum sliding window size that can achieve steady-state detection.

[0045] Industrial systems may include, but are not limited to, boiler combustion systems, nuclear reaction systems, power systems, etc. Operational timing data refers to data generated during the operation of each system. Taking the industrial system as a boiler combustion system as an example, its operational timing data may include, but are not limited to, at least one of the following data: coal grinding amount of each coal mill, power generation of the unit, air door switch angle, steam temperature, main steam pressure, and feed water flow, etc.

[0046] Taking the boiler combustion system as an example, the operation sequence data can be obtained from the supervisory information system (SIS) and the distributed control system (DCS). The DCS and SIS store the operation status parameters of the corresponding equipment in the boiler combustion process, such as the coal grinding amount of the pulverizer, the power generation of the unit, the air door switch angle, the steam temperature, the main steam pressure and the feed water flow rate.

[0047] Step 102: When the first steady-state detection result obtained according to the first operating time series data represents the steady-state operation of the industrial system, gradually expand the right boundary of the preset sliding window to obtain the longest sliding window.

[0048] Among them, the second steady-state detection result obtained according to the second operating time series data characterizes the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third operating time series data characterizes the non-steady-state operation of the industrial system; the second operating time series data is collected according to the longest sliding window, and the third operating time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window.

[0049] Expand the right edge of the sliding window, that is, move the right edge to the right by the expansion step size. The expansion step size can be set according to actual needs.

[0050] In step 102, when the steady-state operation of the industrial system is determined based on the first running time series data collected based on the preset sliding window, the size of the sliding window can be appropriately expanded to collect running time series data with a longer time series as the data basis for steady-state detection, so as to avoid losing many valuable features due to the short running time series data, thereby affecting the accuracy of steady-state detection.

[0051] In step 102, each time the right boundary of the preset sliding window is extended, a steady-state detection is performed. If the steady-state result represents the steady-state operation of the industrial system, the right boundary is continued to be extended. If the steady-state result represents the non-steady-state operation of the industrial system, the right boundary is stopped from being extended. For example, the window size of the preset sliding window is T, the left boundary is i, and the right boundary is i+T. The steady-state operation of the industrial system is determined based on the first running time series data ts[i:i+T] collected by the preset sliding window, and the right boundary is extended to i+T+1*j; the steady-state operation of the industrial system is determined based on the first running time series data ts[i: i+T+1*j] collected by the sliding window after the right boundary is extended, and the right boundary is continued to be extended to i+T+2*j; ...; the non-steady-state operation of the industrial system is determined based on the first running time series data ts[i: i+T+n*j] collected by the sliding window after the right boundary is extended, and the right boundary is stopped from being extended, and i+T+(n-1)*j is determined as the right boundary of the longest sliding window. That is, the left boundary of the longest sliding window is i, the right boundary is i+T+(n-1) *j, and the window size is T+n*j - 1. Wherein, j represents the expansion step of the right boundary, and the size of j can be set according to the actual situation. For example, j can be 1, 2, 3, etc., and this is not particularly limited in the embodiments of the present disclosure.

[0052] The first running time series data collected by the longest sliding window is used as the running time series data collected by this sliding window detection and stored for subsequent data tracing or as the data basis for industrial system control and optimization. The running time series data ts[i:i+T], ts[i:i+T+1*j], ts[i:i+T+2*j], ..., ts[i:i+T+(n-2) *j] do not need to be retained.

[0053] From the above example, we can see that the window size of the sliding window used in this steady-state detection is related to the steady-state duration of the industrial system, and is not a fixed value. Therefore, steady-state segments (operational series data) of any time length during the operation of the industrial system can be detected.

[0054] Step 103: Use the right boundary of the longest sliding window as the left boundary of the sliding window used for the next steady-state detection to perform the next steady-state detection.

[0055] See also Figure 2 , the left boundary of the sliding window used for the next steady-state detection is i+T+(n-1)*j, that is, the sliding step of the sliding window is T+n-1. It can be seen that the sliding step of the sliding window used for the next steady-state detection is related to the steady-state duration of the industrial system and is not a fixed value.

[0056] It should be noted that the window size of the sliding window used for the next steady-state detection can be set according to actual conditions. The window size of the sliding window used for the next steady-state detection can be T or T+(n-1)*j, and the embodiments of the present disclosure do not specifically limit this.

[0057] In this embodiment, a dynamic sliding window is used for steady-state detection. The window size and sliding step of the dynamic sliding window used in each steady-state detection are related to the duration of the steady state. The running time data collected based on the dynamic sliding window retains more valuable steady-state data, thereby improving the effectiveness of steady-state detection.

[0058] In one embodiment, see Figure 3 , when the first steady-state detection result indicates that the industrial system is operating in a non-steady-state, step 102' is performed.

[0059] Step 102': move the left edge of the preset sliding window by the sliding step length and use it as the left edge of the sliding window used for the next steady-state detection.

[0060] In step 102', the sliding step size only needs to be less than or equal to the window size of the preset sliding window. The specific value of the sliding step size can be set according to the actual situation. The sliding step size can be 1, 2, 3, etc., and the embodiment of the present disclosure does not specifically limit this. The window size of the sliding window used for the next steady-state detection in step 102' can be set according to the actual situation, for example, set to T.

[0061] In one embodiment, median filtering is performed on the running time series data before steady-state detection, and steady-state detection is performed on the running time series data after median filtering during steady-state detection.

[0062] During the collection and transmission of time series data, some abnormal values ​​are often caused by unstable sensors or aging transmission lines. Median filtering can effectively filter out these abnormal values ​​in time series data without causing much impact on the collected signal itself. The calculation method is as follows:

[0063] ;

[0064] in, It represents the sampling value of the parameters contained in the runtime series data at time t, and s is the window size of the median filter.

[0065] In one embodiment, the window size of the median filter is 3, and the parameters involved in the median filter can be all the parameters included in the running time series data, or can be part of the parameters, which is not particularly limited in the embodiment of the present disclosure.

[0066] The process of steady-state detection is described below.

[0067] When the operating time series data meets the steady-state conditions, the steady-state detection result is the steady-state operation of the industrial system; wherein the steady-state conditions include at least one of the following: the operating status of the units included in the industrial system has not changed; the operating status is determined based on the operating time series data; the change amplitude of the parameter values ​​of each parameter included in the operating time series data is within the respective preset range; the change trend of the parameter values ​​of each parameter included in the operating time series data conforms to the respective preset trend.

[0068] If the operating time series data is determined to be steady-state based on each steady-state condition, the industrial system is determined to be operating in a steady state; if the operating time series data is determined to be non-steady-state based on at least one steady-state condition, the industrial system is determined to be operating in a non-steady-state.

[0069] In one embodiment, the operating sequence data of the industrial system may be encoded in combination with the operating characteristics of the industrial system to obtain the operating status.

[0070] Taking the industrial system as an example of a boiler combustion system, the coal mill combination contained in the collected operation sequence data of the boiler combustion system is binary-encoded, and the encoding result is used to represent the operation status. The specific encoding method is as follows: at a certain moment, a binary string of length 5 corresponds to five coal mills. At any position of the binary string, if the corresponding coal mill coal grinding amount exceeds the threshold G, the bit is 1, otherwise it is 0. Finally, the binary string is converted into a decimal value as the state code of the coal mill at this moment. For example, if the state code of a coal mill is 27, its binary representation is 11011. Therefore, the code indicates that except for the third coal mill which is not working, the other five coal mills are in working state. The binary encoding result is a categorical variable. The steady-state conditions of steady-state detection include uniqueness detection of the coal mill combination code corresponding to the operating time series data in the sliding window, and judging the number of different coal mill codes in the operating time series data in the sliding window. If there are two or more different coal mill combinations, it means that during the time period shown in the sliding window, one or more coal mills included in the boiler combustion system have undergone one or more adjustments. At this time, it is determined that the operating time series data in the sliding window is non-steady-state.

[0071] Still taking the industrial system as an example of a boiler combustion system, the unit power generation contained in the collected operation time series data of the boiler combustion system is encoded to discretize the variable information, and the encoding result is used to characterize the operation state. The specific encoding method is as follows: divide the power generation value by P, and the result is the encoding result. For example, at a certain moment, the unit power generation is 621. Assuming that the interval size P of the unit power generation group is 20, the corresponding code of the unit power generation is 31. The steady-state condition of the steady-state detection includes a uniqueness detection of the unit power generation code corresponding to the operation time series data in the sliding window, and judging the different unit power generation codes in the operation time series data. If there are two or more different unit power generation codes, it means that the unit power generation has been adjusted once or more times during the time period shown in the sliding window. At this time, it is determined that the operation time series data in the sliding window is non-steady-state.

[0072] It should be noted that the above encoding algorithms are only examples, and in actual application, a suitable encoding algorithm can be selected according to the characteristics of the runtime data of the industrial system.

[0073] The steady-state condition includes that the variation range of the parameter values ​​of each parameter contained in the running time series data is within the respective preset range, that is, the steady-state detection is performed on the parameters contained in the running time series data. Taking the parameters of main steam temperature and main steam pressure as examples, the maximum and minimum values ​​of the main steam temperature and main steam pressure in the running time series data are traversed to find them. Then the variation range of the main steam temperature and main steam pressure is calculated respectively. and , and further judge , If a certain change amplitude exceeds the maximum change amplitude (preset range), the running time series data in the sliding window is determined to be non-steady.

[0074] The changing trend of the parameter values ​​of the parameters included in the steady-state condition conforms to the respective preset trends, that is, the trend detection is performed on the parameters contained in the running time series data. Taking the parameters of main steam temperature and main steam pressure as an example, it is determined whether the main steam temperature and main steam pressure show a monotonically increasing or decreasing trend. If the running time series data in the sliding window is in a monotonically increasing or decreasing trend, it is determined that the running time series data in the sliding window is non-steady-state.

[0075] In this embodiment, the steady-state condition is determined according to the specific application scenario and the mechanism knowledge of the industrial system. By integrating the mechanism knowledge of the specific scenario of steady-state detection into the steady-state detection algorithm, it is undoubtedly more targeted and can have a higher accuracy.

[0076] In one implementation, the Mann-Kendall algorithm is used to calculate the relative changes between data points at different times to determine whether the main steam temperature and the main steam pressure present a monotonically increasing or decreasing trend.

[0077] In step 102, each time the right boundary of the preset sliding window is extended, a steady-state detection is performed. In order to reduce the amount of calculation and time complexity, a monotonic queue (monotonically increasing or monotonically decreasing queue) is used to optimize the calculation of the maximum and minimum values ​​of the parameters, and a tree array is used to optimize the calculation of the number of reverse pairs or positive pairs in the Mann-Kendall algorithm. Specifically, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time data when the industrial system is running in a steady state, and the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the change trend is determined according to the running time data stored in the tree array.

[0078] A monotonic queue is a special queue that allows the maximum (or minimum) element in the queue to be found in constant time. In an embodiment, two monotonic queues are used: one for maintaining the maximum value in the sliding window, and the other for maintaining the minimum value in the sliding window. Each time the right boundary of the preset sliding window is extended, the incremental data (the data between the right boundary before the extension and the right boundary after the extension) is compared with the maximum value and the minimum value in the monotonic queue. If the incremental data has a parameter value of a parameter that is larger than the maximum value in the first monotonic queue, the parameter value is stored in the first monotonic queue; if the incremental data has a parameter value of a parameter that is smaller than the minimum value in the second monotonic queue, the parameter value is stored in the second monotonic queue. When determining the maximum and minimum values, there is no need to traverse all the time series data in the sliding window, only the incremental data needs to be compared, thereby achieving O(1) time complexity to obtain the maximum and minimum values ​​in the window.

[0079] Similarly, each time the right boundary of the sliding window is expanded to the right, the query and update functions of the tree array are used to quickly calculate the number of reverse order pairs and positive order pairs in the current window, so that the trend of the sequence in the window can be quickly detected. Through the above optimization strategy, it is ensured that when a variable-length sliding window is used to perform steady-state detection on historical data of length N, the time complexity can reach O(NlogN).

[0080] The Mann-Kendall algorithm is a nonparametric statistical test method used to analyze whether there is a monotonically increasing or decreasing trend in time series data. Assume that the main steam pressure data in a sliding window is , n is the window size of the current sliding window, calculate the statistic S:

[0081] ;

[0082] in, , are the parameter values ​​corresponding to the jth and kth moments respectively, and , is a symbolic function and is defined as follows:

[0083] ;

[0084] The statistic S roughly follows a normal distribution, and its mean ,variance In this embodiment, since it is a variable-length window (dynamic sliding window), n is an integer greater than or equal to T, and the specific value is the window size. The statistical value Z can be obtained by normalizing the statistic S as follows:

[0085] ;

[0086] The Z value is used to test the statistical significance of the trend. In the bilateral trend test, at a given confidence level In the case of , indicating that the null hypothesis is rejected and the significance level is A bilateral significant trend test is performed to determine whether there is an upward or downward trend. A positive value of Z indicates that the sequence is increasing, and a negative value indicates that the sequence is decreasing. In one implementation, the confidence level Set to 0.05, the critical value .

[0087] Take the time series data of a sliding window with a length of 20 in the sampling data of the main steam pressure of a boiler combustion system in a certain period of time as an example to calculate the statistical test value of the time series data , indicating that at the confidence level The changing trend is not significant.

[0088] It should be noted that, since different parameters have their own preset ranges, the first monotonic queue and the second monotonic queue are respectively set for different parameters. And different parameters are respectively subjected to hypothesis testing, so the tree array is respectively set for different parameters.

[0089] Corresponding to the aforementioned steady-state detection method embodiment of an industrial system, the present disclosure also provides an embodiment of a steady-state detection system for an industrial system.

[0090] Figure 4 A module schematic diagram of a steady-state detection system for an industrial system provided by an exemplary embodiment of the present disclosure, the system comprising:

[0091] A collection module 41, configured to perform steady-state detection on the industrial system according to first operating time series data of the industrial system collected by a preset sliding window;

[0092] A dynamic sliding window module 42 is used to gradually expand the right boundary of the preset sliding window to obtain a longest sliding window when the first steady-state detection result obtained according to the first running time series data represents the steady-state operation of the industrial system, wherein the second steady-state detection result obtained according to the second running time series data represents the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third running time series data represents the non-steady-state operation of the industrial system; the second running time series data is collected according to the longest sliding window, and the third running time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window;

[0093] The steady-state detection module 43 is used to use the right boundary of the longest sliding window as the left boundary of the sliding window used for the next steady-state detection to perform the next steady-state detection.

[0094] Optionally, the dynamic sliding window module is further used to move the left boundary of the preset sliding window by a sliding step size as the left boundary of the sliding window used for the next steady-state detection when the first steady-state detection result indicates that the industrial system is operating in a non-steady-state.

[0095] Optionally, the steady-state detection module is specifically configured to determine that the steady-state detection result is steady-state operation of the industrial system when the running time series data satisfies a steady-state condition; wherein the steady-state condition includes at least one of the following:

[0096] The operating status of the units included in the industrial system has not changed; the operating status is determined according to the operating sequence data;

[0097] The variation range of the parameter value of each parameter included in the running time series data is within the respective preset range;

[0098] The variation trends of the parameter values ​​of the parameters included in the runtime data are consistent with the respective preset trends.

[0099] Optionally, the step of gradually extending the right boundary of the preset sliding window includes:

[0100] When the steady-state condition includes that the parameter values ​​of the parameters contained in the running time series data are within the respective preset ranges and the changing trends of the parameter values ​​of the parameters contained in the running time series data conform to the respective preset trends, a steady-state detection is performed each time the right boundary is extended; wherein, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the changing trend is determined according to the running time series data stored in a tree array.

[0101] Optionally, the running state is determined according to the encoding result of the running timing data;

[0102] and / or, performing median filtering on the running time series data before steady-state detection;

[0103] And / or, a Mann-Kendall algorithm is used to determine a change trend of a parameter value of each parameter.

[0104] As for the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The system embodiment described above is only illustrative, wherein the units described as separate components may or may not be physically separated, and the components as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the disclosed solution.

[0105] Figure 5 This is a structural schematic diagram of an electronic device shown in an example embodiment of the present disclosure, the electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor, and when the processor executes the computer program, the steady-state detection method of the industrial system described in any of the above embodiments is implemented. Figure 5 The electronic device 50 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0106] like Figure 5 As shown, the electronic device 50 may be in the form of a general-purpose computing device, for example, it may be a server device. The components of the electronic device 50 may include, but are not limited to: at least one processor 51, at least one memory 52, and a bus 53 connecting different system components (including the memory 52 and the processor 51).

[0107] The bus 53 includes a data bus, an address bus, and a control bus.

[0108] The memory 52 may include a volatile memory, such as a random access memory (RAM) 521 and / or a cache memory 522 , and may further include a read-only memory (ROM) 523 .

[0109] The memory 52 may also include a program tool 525 (or utility) having a set (at least one) of program modules 524, such program modules 524 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0110] The processor 51 executes various functional applications and data processing by running the computer program stored in the memory 52, such as the steady-state detection method for the industrial system provided in any of the above embodiments.

[0111] The electronic device 50 may also communicate with one or more external devices 54 (e.g., keyboards, pointing devices, etc.). Such communication may be performed via an input / output (I / O) interface 55. In addition, the electronic device 50 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 56. As shown, the network adapter 56 communicates with other modules of the electronic device 50 via a bus 53. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 50, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems, etc.

[0112] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided into multiple units / modules to be embodied.

[0113] An embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steady-state detection method for an industrial system provided by any of the above embodiments is implemented.

[0114] The readable storage medium may include but is not limited to: a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device or any suitable combination of the above.

[0115] The embodiment of the present disclosure further provides a computer program product, including a computer program, wherein when the computer program is executed by a processor, the steady-state detection method for an industrial system described in any one of the above items is implemented.

[0116] Among them, the program code for executing the computer program product of the present disclosure can be written in any combination of one or more programming languages, and the program code can be executed completely on the user device, partially on the user device, as an independent software package, partially on the user device and partially on a remote device, or completely on the remote device.

[0117] Although the specific embodiments of the present disclosure are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present disclosure is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present disclosure, but these changes and modifications all fall within the protection scope of the present disclosure.

Claims

1. A steady-state detection method for an industrial system, characterized in that: include: Performing steady-state detection on the industrial system according to first operating time series data of the industrial system collected by a preset sliding window; In the case where the first steady-state detection result obtained according to the first operating time series data represents the steady-state operation of the industrial system, the right boundary of the preset sliding window is gradually expanded to obtain the longest sliding window; wherein the second steady-state detection result obtained according to the second operating time series data represents the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third operating time series data represents the non-steady-state operation of the industrial system; the second operating time series data is collected according to the longest sliding window, and the third operating time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window; The right boundary of the longest sliding window is used as the left boundary of the sliding window used for the next steady-state detection, and the next steady-state detection is performed.

2. The steady-state detection method for an industrial system according to claim 1, characterized in that: In the case where the first steady-state detection result indicates that the industrial system is operating in a non-steady-state, the left boundary of the preset sliding window is moved by a sliding step size and used as the left boundary of the sliding window used for the next steady-state detection.

3. The steady-state detection method for an industrial system according to claim 1 or 2, characterized in that: Steady-state testing, including: When the running time series data meets the steady-state condition, the steady-state detection result is that the industrial system is running in a steady state; wherein the steady-state condition includes at least one of the following: The operating status of the units included in the industrial system has not changed; the operating status is determined according to the operating sequence data; The variation range of the parameter value of each parameter included in the running time series data is within the respective preset range; The variation trends of the parameter values ​​of the parameters included in the runtime data are consistent with the respective preset trends.

4. The steady-state detection method for an industrial system according to claim 3, characterized in that: The step of gradually extending the right boundary of the preset sliding window includes: When the steady-state condition includes that the parameter values ​​of the parameters contained in the running time series data are within the respective preset ranges and the changing trends of the parameter values ​​of the parameters contained in the running time series data conform to the respective preset trends, a steady-state detection is performed each time the right boundary is extended; wherein, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the changing trend is determined according to the running time series data stored in a tree array.

5. The steady-state detection method for an industrial system according to claim 3, characterized in that: The running state is determined according to the encoding result of the running time sequence data; and / or, performing median filtering on the running time series data before steady-state detection; And / or, a Mann-Kendall algorithm is used to determine a change trend of a parameter value of each parameter.

6. A steady-state detection system for an industrial system, characterized in that: include: A collection module, used for performing steady-state detection on the industrial system according to first running time series data of the industrial system collected by a preset sliding window; A dynamic sliding window module, used for gradually expanding the right boundary of the preset sliding window to obtain a longest sliding window when the first steady-state detection result obtained according to the first running time series data represents the steady-state operation of the industrial system, wherein the second steady-state detection result obtained according to the second running time series data represents the steady-state operation of the industrial system, and the third steady-state detection result obtained according to the third running time series data represents the non-steady-state operation of the industrial system; the second running time series data is collected according to the longest sliding window, and the third running time series data is collected according to a sliding window whose right boundary is greater than the right boundary of the longest sliding window; The steady-state detection module is used to use the right boundary of the longest sliding window as the left boundary of the sliding window used for the next steady-state detection to perform the next steady-state detection.

7. The steady-state detection system for an industrial system according to claim 6, characterized in that: The dynamic sliding window module is also used to move the left boundary of the preset sliding window by the sliding step size as the left boundary of the sliding window used for the next steady-state detection when the first steady-state detection result indicates that the industrial system is operating in a non-steady-state.

8. The steady-state detection system for an industrial system according to claim 6 or 7, characterized in that: The steady-state detection module is specifically used to determine that the steady-state detection result is the steady-state operation of the industrial system when the running time series data meets the steady-state condition; wherein the steady-state condition includes at least one of the following: The operating status of the units included in the industrial system has not changed; the operating status is determined according to the operating sequence data; The variation range of the parameter value of each parameter included in the running time series data is within the respective preset range; The variation trends of the parameter values ​​of the parameters included in the runtime data are consistent with the respective preset trends.

9. The steady-state detection system for an industrial system according to claim 8, characterized in that: The step of gradually extending the right boundary of the preset sliding window includes: When the steady-state condition includes that the parameter values ​​of the parameters contained in the running time series data are within the respective preset ranges and the changing trends of the parameter values ​​of the parameters contained in the running time series data conform to the respective preset trends, a steady-state detection is performed each time the right boundary is extended; wherein, the upper limit value of the preset range of each steady-state detection is determined according to the first monotonic queue, and the lower limit value of the preset range is determined according to the second monotonic queue; the first monotonic queue is used to store the maximum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the second monotonic queue is used to store the minimum parameter value of the parameter contained in the running time series data when the industrial system is running in a steady state, which is represented by the result of each steady-state detection; the changing trend is determined according to the running time series data stored in a tree array.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and used to run on the processor, characterized in that: When the processor executes the computer program, the steady-state detection method for an industrial system according to any one of claims 1 to 5 is implemented.

Images