Single-loop working condition classification method and device and computer equipment
Through the multi-round segment identification and similarity analysis methods, the problem of low accuracy in single-loop operating conditions classification is solved, and better capture of nonlinear characteristics of complex systems and improved accuracy in operating conditions classification is achieved.
Patent Information
- Application Number
- CN202411954652.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-27
- Publication Date
- 2025-04-18
AI Technical Summary
In the prior art, the accuracy of single-loop operating conditions classification is low, making it difficult to effectively capture nonlinear characteristics in complex systems.
By collecting historical operation data of a single loop, using multiple rounds of segmented identification methods, data is segmented and identified according to multiple data windows of different lengths, multiple identification models are established, and the target data segment is determined through relative deviation values, and the similar data segments are finally divided into the same operating conditions and merged.
The accuracy of single-loop operating condition classification is improved, and the ability to capture nonlinear characteristics of complex systems is enhanced through multiple rounds of identification.
Smart Images
Figure CN119848696A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of automation control technology, and in particular to a single-loop operating condition classification method, device and computer equipment. Background Art
[0002] For general industrial systems, traditional control strategies (such as PI control or PID control) can achieve the control purpose. However, with the increasing complexity and intelligence of process equipment, the system presents strong nonlinear characteristics, which not only brings many difficulties to process modeling but also puts higher requirements on the controller. It is often necessary to continuously update and adjust the controller parameters according to the actual operating conditions to achieve satisfactory control effects. From the perspective of theoretical implementation and practical application, the application of decomposition and synthesis multi-model modeling methods in nonlinear systems has been increasingly valued by the industry. This method first establishes a sub-model of the system under each local working condition to achieve the purpose of decomposing the complex process covering all working conditions, and then uses switching or weighting methods to achieve the description of the global model using multiple sub-models. Therefore, it is particularly important to first find a practical and feasible dynamic working condition division method.
[0003] In the field of industrial process automation control, with the increasing complexity and intelligence of process equipment, the system often exhibits obvious nonlinear characteristics. This nonlinearity poses a challenge to traditional linear control strategies (such as PI or PID control) because they may not provide satisfactory results under all working conditions. In order to cope with this complexity, multi-model control strategies have gradually gained attention. It achieves the purpose of global control by establishing multiple sub-models under different local working conditions and then synthesizing these sub-models through appropriate switching or weighting methods. However, some methods rely on complex data analysis, such as entropy clustering and fuzzy models, which may increase the computational cost, especially when there is no obvious interference in the system. In addition, model identification techniques, such as the combination of clustering algorithms and linear identification, can automatically divide the working conditions and determine the number of models, but when the model order is uncertain, the reliability of the identification results will be affected because the distance in the feature space may be destroyed by irrelevant information. At the same time, the methods in related technologies cannot fully capture the nonlinear characteristics of the system. Alternatively, the operating conditions can be simply divided by calculating the statistics of the parameters (such as mean, variance) and related functions. This is effective in simple systems but may not be accurate enough in complex systems. Alternatively, the operating conditions can be classified through principal component analysis and clustering methods. This method can adaptively determine the number of operating conditions, but may result in low accuracy when the model order is uncertain or the amount of data is insufficient. Summary of the invention
[0004] The embodiments of the present application provide a single-circuit operating condition classification method, device and computer equipment to at least solve the technical problem of low accuracy in single-circuit operating condition classification in the related art.
[0005] According to one aspect of an embodiment of the present application, a single-loop operating condition classification method is provided, including: collecting historical operating data of a single loop; performing multiple rounds of segmented identification on the historical operating data according to multiple data windows of different lengths, to obtain multiple identification results corresponding to each round of identification, the identification results including model parameters for establishing an identification model, and establishing an identification model based on the identification results to obtain multiple identification models; respectively obtaining first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments, and determining a data segment of the identification model whose first relative deviation value is less than a first preset threshold as a target data segment; dividing mutually similar target data segments among the multiple target data segments into data segments under the same type of operating conditions, and merging the data segments under the same type of operating conditions to obtain data subsets under different operating conditions.
[0006] Optionally, an identification model is established based on the identification result to obtain multiple identification models, including: obtaining the number of rounds of segmented identification, and determining the length of the data window used in each round of segmented identification and the step length of the data window rolling forward during each round of segmented identification, wherein each time the data window rolls forward one step, a data segment corresponding to the data window is obtained; in each segmented identification round, the data window is used to divide the historical operation data into multiple data segments according to the step length, and an identification model of each data segment is constructed based on the data in each data segment.
[0007] Optionally, first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments are obtained respectively, including: dividing the data segment corresponding to each identification model into multiple sub-data segments; determining the first relative deviation value based on the standard deviation of the output results corresponding to each sub-data segment and the historical output results.
[0008] Optionally, respectively obtaining first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments includes:
[0009]
[0010] In the formula, Re represents the first relative deviation value, i represents the number of the sub-segment, m represents the number of sub-segments, and k i Indicates the number of data points in each sub-segment, PV i Represents the historical output results of the ith sub-data segment, ysim iIt represents the output result of the identification model of the i-th sub-data segment, and std represents the standard deviation.
[0011] Optionally, target data segments that are similar to each other among the multiple target data segments are divided into data segments under the same type of working conditions, including: selecting two data segments to be compared from the multiple target data segments, and determining identification models of the two data segments to be compared; respectively using the identification models of the two data segments to be compared to analyze each other's data segments to obtain output results, and comparing the output results with historical output results, and when the comparison results indicate similarity, determining that the two data segments to be compared are data segments under the same type of working conditions.
[0012] Optionally, the identification models of the two data segments to be compared are respectively used to analyze each other's data segments to obtain output results, and the output results are compared with historical output results, including: determining the two data segments to be compared as a first data segment and a second data segment respectively, wherein the identification model of the first data segment is a first identification model, and the identification model of the second data segment is a second identification model; using the first identification model to analyze the second data segment to obtain a first output result, and using the second identification model to analyze the first data segment to obtain a second output result; obtaining a second relative deviation value between the first output result and the historical output result of the second data segment and a third relative deviation value between the second output result and the historical output result of the first data segment; when at least one of the second relative deviation value and the third relative deviation value is less than a second preset threshold, determining that the first identification model and the second identification model are similar, and the first data segment and the second data segment belong to the same type of operating condition.
[0013] Optionally, data segments under the same type of working conditions are merged to obtain data subsets under different working conditions, including: obtaining all data segments under the same type of working conditions; and determining the union of all data segments under the same type of working conditions as the data subset of the corresponding working condition.
[0014] According to another aspect of an embodiment of the present application, a single-loop operating condition classification device is also provided, including: an acquisition module, used to acquire historical operating data of a single loop; an identification module, used to perform multiple rounds of segmented identification on the historical operating data according to multiple data windows of different lengths, to obtain multiple identification results corresponding to each round of identification, the identification results including model parameters for establishing an identification model, and establishing an identification model based on the identification results to obtain multiple identification models; a determination module, used to respectively obtain a first relative deviation value between the output results of the multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment, and determine the data segment of the identification model whose first relative deviation value is less than a first preset threshold as a target data segment; a classification module, used to divide the target data segments that are similar to each other among the multiple target data segments into data segments under the same type of operating conditions, and merge the data segments under the same type of operating conditions to obtain data subsets under different operating conditions.
[0015] According to another aspect of the embodiment of the present application, a computer device is also provided, including: a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above-mentioned single-loop operating condition classification method.
[0016] According to another aspect of the embodiments of the present application, a non-volatile storage medium is provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the above-mentioned single-loop operating condition classification method by running the computer program.
[0017] According to another aspect of the embodiments of the present application, a computer program product is provided, including computer instructions, which implement the above-mentioned single-loop operating condition classification method when executed by a processor.
[0018] In an embodiment of the present application, historical operating data of a single loop is collected; multiple rounds of segmented identification are performed on the historical operating data according to multiple data windows of different lengths to obtain multiple identification results corresponding to each round of identification, wherein the identification results include model parameters for establishing an identification model, and an identification model is established based on the identification results to obtain multiple identification models; first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments are obtained respectively, and a data segment of the identification model whose first relative deviation value is less than a first preset threshold is determined as a target data segment; target data segments similar to each other in the multiple target data segments are divided into data segments under the same type of operating conditions, and the data segments under the same type of operating conditions are merged to obtain data subsets under different operating conditions, thereby achieving the purpose of improving classification accuracy through multiple rounds of identification, thereby achieving the technical effect of improving the classification accuracy of single-loop operating conditions, and further solving the technical problem of low classification accuracy of single-loop operating conditions in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0020] Figure 1 It is a hardware structure block diagram of a computer terminal for implementing a single-circuit working condition classification method according to an embodiment of the present application;
[0021] Figure 2 is a flow chart of a single-circuit operating condition classification method according to an embodiment of the present application;
[0022] Figure 3 is a flow chart of a method for multi-round segmented identification according to an embodiment of the present application;
[0023] Figure 4 is a flowchart of a method for classifying working conditions according to an embodiment of the present application;
[0024] Figure 5 is a flow chart of another single-circuit working condition classification method according to an embodiment of the present application;
[0025] Figure 6 It is a structural diagram of a single-circuit operating condition classification device according to an embodiment of the present application. DETAILED DESCRIPTION
[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0028] The information collected in the embodiments of the present application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or reject automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.
[0029] In order to solve the problems existing in the related art, the present application embodiment provides a single-circuit working condition classification method, which can be run on Figure 1 In the computer terminal shown, the computer terminal is explained below.
[0030] The single-circuit operating condition classification method embodiment provided in the embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal for implementing a single-circuit operating condition classification method. Figure 1 As shown, the computer terminal 10 may include one or more (102a, 102b, ..., 102n are used to illustrate) processors (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions connected via a wired and / or wireless network. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.
[0031] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).
[0032] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the single-circuit working condition classification method in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned single-circuit working condition classification method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0033] The transmission module 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission module 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.
[0034] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 .
[0035] It should be noted that, in some optional embodiments, the above Figure 1 The computer terminal shown may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of hardware elements and software elements. It should be noted that Figure 1 This is merely one example of a particular embodiment and is intended to illustrate the types of components that may be present in the computer terminal described above.
[0036] Under the above-mentioned operating environment, an embodiment of the present application provides an embodiment of a single-loop operating condition classification method. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0037] Figure 2 is a flow chart of a single-circuit operating condition classification method according to an embodiment of the present application, such as Figure 2As shown, the method comprises the following steps:
[0038] Step S202, collecting historical operation data of a single loop;
[0039] In step S202, the historical operation data includes at least: a set value (SV), a control variable (MV) and a manipulated variable (PV) of the loop, and prior knowledge about the process, such as a process response time (ts).
[0040] Step S204, performing multiple rounds of segmented identification on the historical operation data according to multiple data windows of different lengths, obtaining multiple identification results corresponding to each round of identification, wherein the identification results include model parameters for establishing an identification model, and establishing an identification model based on the identification results to obtain multiple identification models;
[0041] In step S204, segment identification can be divided into multiple rounds. In each round of segment identification, the required data segment length is fixed. After the identification of the segment data is completed, the position of the data segment is gradually rolled forward according to a certain step length, and identification is performed in sequence until the rolled data segment traverses all historical data. This series of operations is called "one round of segment identification". After that, the length of the rolled data segment in each round can be gradually increased to perform subsequent rounds of segment identification work. The number of rounds and the length of the segmented data segment in each round are predetermined.
[0042] Step S206, respectively obtaining first relative deviation values between the output results of the multiple recognition models corresponding to each round of recognition and the historical output results of the corresponding data segment, and determining the data segment of the recognition model whose first relative deviation value is less than a first preset threshold as the target data segment;
[0043] In step S206, after all the segmented identification work is completed, the identification models that meet the accuracy requirements and the data segment positions corresponding to the model when the model is established are recorded one by one.
[0044] Step S208: divide the target data segments that are similar to each other among the multiple target data segments into data segments under the same type of working conditions, and merge the data segments under the same type of working conditions to obtain data subsets under different working conditions.
[0045] In step S208, the segmented identification results are recorded to classify the working conditions of the historical data. The basis for the working condition classification is the "similarity" between the models. This similarity can be analyzed through cross-validation between the models. If the two models are "similar" after analysis, then it can be considered that the data segments used when establishing the two models belong to the same working condition. The final result of this step is to divide the historical data into several data subsets. In each subset, it can be considered that the system is in a relatively stable working condition at this time, and the data in the subset can be used to establish the system model under this working condition.
[0046] Through the above steps S202 to S208, by collecting the historical operation data of a single loop; performing multiple rounds of segmented identification on the historical operation data according to multiple data windows of different lengths, obtaining multiple identification results corresponding to each round of identification, the identification results include model parameters for establishing an identification model, and establishing an identification model based on the identification results to obtain multiple identification models; respectively obtaining the first relative deviation values of the output results of the multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment, and determining the data segment of the identification model whose first relative deviation value is less than the first preset threshold as the target data segment; dividing the target data segments similar to each other in the multiple target data segments into data segments under the same type of working conditions, and merging the data segments under the same type of working conditions to obtain data subsets under different working conditions, thereby achieving the purpose of improving the classification accuracy through multiple rounds of identification, thereby achieving the technical effect of improving the classification accuracy of the single-loop working condition, and then solving the technical problem of low classification accuracy of the single-loop working condition in the related technology. The following is a detailed description.
[0047] In some embodiments of the present application, an identification model is established based on the identification results, and specific steps for obtaining multiple identification models include: obtaining the number of rounds of segmented identification, and determining the length of the data window used in each round of segmented identification and the step length of the data window rolling forward during each round of segmented identification, wherein each time the data window rolls forward one step, a data segment corresponding to the data window is obtained; in each segmented identification round, the data window is used to divide the historical operation data into multiple data segments according to the step length, and an identification model of each data segment is constructed based on the data in each data segment.
[0048] like Figure 3 As shown in the figure, the specific process of multi-round segmentation recognition is mainly divided into three steps, including:
[0049] Step 21, perform segmented identification on the entire historical data. Specifically: divide the entire segmented identification process into multiple rounds. In each round of identification, fix the length of the data window used to identify the model (the data in a data window is continuous on the time scale, including continuous SV, MV, and PV data). After completing the model identification of this segment of data, the data window rolls forward on the time scale with a certain step length, and uses the data segments in the window to perform model identification in turn until these data windows traverse all historical data on the time scale. This series of operations is called "one round of segmented identification." In one round of segmented identification, several identification models can be obtained, and each identification model has its corresponding data segment (that is, the identification model is established using the data in this data segment). After completing one round of segmented identification, increase the length of the data window, repeat the above operations, and finally complete several rounds of segmented identification. The number of segmented identification rounds, the length of the data window in each round of identification, and the rolling step of the data window can be determined based on experience. For example, in each round of segmented identification, the data window length is 1, 3, 5, ..., 21 times the process response time (ts), and 11 rounds of segmented identification are performed in sequence. In each round of segmented identification, the rolling step of the data window is k times the data window length, and k can be determined based on the accuracy requirements, for example: k≤1 (the specific value of k can be set according to actual needs, and the embodiments of the present application are not limited to this). In each identification process, the data segment in the current data window is used to perform identification based on the least squares identification algorithm to obtain the model parameters of the identification model.
[0050] It can be understood that the process of model identification is to determine the model parameters of the identification model based on the data in the corresponding data segment, and the identification model can be an ARX (AutoRegressive model with eXogenous inputs) model.
[0051] In addition, the process of respectively obtaining the first relative deviation value of the output results of multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment includes: dividing the data segment corresponding to each identification model into multiple sub-data segments; determining the first relative deviation value based on the standard deviation of the output results corresponding to each sub-data segment and the historical output results.
[0052] Step 22: After each identification, the obtained ARX model is tested by calculating the relative deviation Re (first relative deviation value) between the model output ysim and the historical output PV in the current data window.
[0053] The specific calculation method is shown in the following formula:
[0054]
[0055] In the formula, Re represents the first relative deviation value, i represents the number of the sub-segment, m represents the number of sub-segments, and k i Indicates the number of data points in each sub-segment, PV i Represents the historical output results of the ith sub-data segment, ysim i It represents the output result of the identification model of the i-th sub-data segment, and std represents the standard deviation.
[0056] It is understandable that the historical output result can be obtained from the database, which is the data vector of the historical actual output, and the output result of the identification model can be the data vector output by analyzing the data in the corresponding data segment using the identification model. Optionally, the data segment corresponding to the data window can be divided into m continuous non-overlapping sub-data segments in time scale, for example: the data window time length corresponding to each sub-data segment is 1 times the process response time length.
[0057] It should be noted that the first preset threshold can be determined based on prior experience, for example: 0.45.
[0058] Step 3, when the relative deviation value of a model obtained by the test is less than the first preset threshold, the model and the data window position used in the modeling process are recorded. After completing all the identification work, a series of models that have passed the test and their corresponding target data segments will be recorded, such as model Gi and its corresponding target data segment windowi, model Gj and its corresponding target data segment windowj.
[0059] In some embodiments of the present application, the method of dividing mutually similar target data segments among multiple target data segments into data segments under the same type of working conditions is as follows: two data segments to be compared are selected from the multiple target data segments, and identification models of the two data segments to be compared are determined; the identification models of the two data segments to be compared are respectively used to analyze each other's data segments to obtain output results, and the output results are compared with historical output results, and when the comparison results indicate similarity, it is determined that the two data segments to be compared are data segments under the same type of working conditions.
[0060] In order to better explain the classification process, the specific classification process is as follows: the two data segments to be compared are respectively determined as the first data segment and the second data segment, wherein the identification model of the first data segment is the first identification model, and the identification model of the second data segment is the second identification model; the first identification model is used to analyze the second data segment to obtain a first output result, and the second identification model is used to analyze the first data segment to obtain a second output result; a second relative deviation value between the first output result and the historical output result of the second data segment and a third relative deviation value between the second output result and the historical output result of the first data segment are obtained; when at least one of the second relative deviation value and the third relative deviation value is less than a second preset threshold, it is determined that the first identification model and the second identification model are similar, and the first data segment and the second data segment belong to the same type of working condition.
[0061] Figure 4 A flow chart of operating condition classification is shown, such as Figure 4 As shown, including:
[0062] Step 31, classify the identification models and target data segments corresponding to all target data segments. The basic idea of classification is to compare the "similarity" between the identification models. If two models are considered "similar", it can be considered that the data segments corresponding to them are in the same working condition. For example, for each model corresponding to the target data segment, two models Gi, Gj and their corresponding data segments windowi, windowj are taken. In order to analyze the "similarity" of the model, cross-validation is performed. In the data segment windowj, the relative deviation Reij between the output of model Gi and the historical output PV of the current data segment is calculated. In the data segment windowi, the relative deviation Reji between the output of model Gj and the historical output PV of the current data segment is calculated. When any one of Reij and Reji is less than the second preset threshold, it is considered that model Gi and model Gj are "similar", and the data segments windowi and windowj are classified into the same working condition. According to the above principles, the recorded models are cross-validated in pairs. If data segments windowi and windowj belong to the same type of working condition, and data segments windowj and data segments windowz belong to the same type of working condition, then data segments windowi and windowz are considered to belong to the same type of working condition. After the classification is finally completed, the following conditions should be met: (1) All data windows in the same type of working condition belong to the same working condition; (2) Any two data segments in different classes do not belong to the same working condition.
[0063] Optionally, data segments under the same type of working conditions are merged to obtain data subsets under different working conditions, including: obtaining all data segments under the same type of working conditions; determining the union of all data segments under the same type of working conditions as the data subset for the corresponding working condition, in each subset, it can be considered that the system is in a relatively stable working condition at this time, and the data in the subset can be used to establish a system model under this working condition.
[0064] Figure 5 Another single-circuit operating condition classification method is shown, such as Figure 5 As shown, including:
[0065] 1. Collect single-loop historical data that needs to be classified for working conditions; 2. Perform segmented system identification based on historical operating data; 3. Perform working condition classification of historical data based on the segmented identification results obtained from the records.
[0066] Figure 6 A single-circuit operating condition classification device according to an embodiment of the present application comprises:
[0067] A collection module 60 is used to collect historical operation data of a single circuit;
[0068] an identification module 62, configured to perform multiple rounds of segmented identification on the historical operation data according to multiple data windows of different lengths, obtain multiple identification results corresponding to each round of identification, wherein the identification results include model parameters for establishing an identification model, and establish an identification model based on the identification results to obtain multiple identification models;
[0069] A determination module 64 is used to respectively obtain first relative deviation values of output results of multiple recognition models corresponding to each round of recognition and historical output results of corresponding data segments, and determine the data segment of the recognition model whose first relative deviation value is less than a first preset threshold as the target data segment;
[0070] The classification module 66 is used to classify the target data segments that are similar to each other among the multiple target data segments into data segments under the same type of working conditions, and merge the data segments under the same type of working conditions to obtain data subsets under different working conditions.
[0071] Through the above-mentioned single-loop operating condition classification device, historical operating data of a single loop is collected; multiple rounds of segmented identification are performed on the historical operating data according to multiple data windows of different lengths to obtain multiple identification results corresponding to each round of identification, and the identification results include model parameters for establishing an identification model, and an identification model is established based on the identification results to obtain multiple identification models; the first relative deviation values of the output results of the multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment are respectively obtained, and the data segment of the identification model whose first relative deviation value is less than a first preset threshold is determined as the target data segment; the target data segments similar to each other in the multiple target data segments are divided into data segments under the same type of operating conditions, and the data segments under the same type of operating conditions are merged to obtain data subsets under different operating conditions, thereby achieving the purpose of improving the classification accuracy through multiple rounds of identification, thereby achieving the technical effect of improving the classification accuracy of the single-loop operating condition, and further solving the technical problem of low accuracy in the classification of single-loop operating conditions in the related technology.
[0072] The identification module 62 includes: a model submodule, which is used to establish an identification model based on the identification results to obtain multiple identification models, including: obtaining the number of rounds of segmented identification, and determining the length of the data window used in each round of segmented identification and the step length of the data window rolling forward during each round of segmented identification, wherein the data segment corresponding to the data window is obtained each time the data window rolls forward one step; in each segmented identification round, the data window is used to divide the historical operation data into multiple data segments according to the step length, and an identification model of each data segment is constructed based on the data in each data segment.
[0073] The model submodule includes: a deviation unit, which is used to obtain the first relative deviation value of the output results of multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment, including: dividing the data segment corresponding to each identification model into multiple sub-data segments; determining the first relative deviation value based on the standard deviation of the output results corresponding to each sub-data segment and the historical output results.
[0074] The deviation unit includes an acquisition subunit, which is used to respectively obtain the first relative deviation values of the output results of the multiple identification models corresponding to each round of identification and the historical output results of the corresponding data segment, including:
[0075]
[0076] In the formula, Re represents the first relative deviation value, i represents the number of the sub-segment, m represents the number of sub-segments, and k i Indicates the number of data points in each sub-segment, PV i Represents the historical output results of the ith sub-data segment, ysim iIt represents the output result of the identification model of the i-th sub-data segment, and std represents the standard deviation.
[0077] The classification module 66 includes: a classification submodule, which is used to classify target data segments that are similar to each other among the multiple target data segments into data segments under the same type of working conditions, including: selecting two data segments to be compared from the multiple target data segments, and determining the identification models of the two data segments to be compared; using the identification models of the two data segments to be compared to analyze each other's data segments, respectively, to obtain output results, and compare the output results with historical output results, and when the comparison results indicate similarity, determine that the two data segments to be compared are data segments under the same type of working conditions.
[0078] The classification submodule includes: a classification unit, which is used to respectively use the identification models of the two data segments to be compared to analyze the other party's data segment to obtain an output result, and compare the output result with the historical output result, including: determining the two data segments to be compared as a first data segment and a second data segment, respectively, wherein the identification model of the first data segment is a first identification model, and the identification model of the second data segment is a second identification model; using the first identification model to analyze the second data segment to obtain a first output result, and using the second identification model to analyze the first data segment to obtain a second output result; obtaining a second relative deviation value between the first output result and the historical output result of the second data segment and a third relative deviation value between the second output result and the historical output result of the first data segment; when at least one of the second relative deviation value and the third relative deviation value is less than a second preset threshold, it is determined that the first identification model and the second identification model are similar, and the first data segment and the second data segment belong to the same type of working condition.
[0079] The classification unit includes: a determination subunit, which is used to merge data segments under the same type of working conditions to obtain data subsets under different working conditions, including: obtaining all data segments under the same type of working conditions; determining the union of all data segments under the same type of working conditions as the data subset of the corresponding working condition.
[0080] It should be noted that Figure 6 The single circuit classification device shown is used to perform Figure 2 The single-circuit operating condition classification method shown in the figure, therefore the relevant explanations in the above single-circuit operating condition classification method are also applicable to the single-circuit operating condition classification device, and will not be repeated here.
[0081] An embodiment of the present application also provides a computer device, including: a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above-mentioned single-loop operating condition classification method.
[0082] An embodiment of the present application further provides a non-volatile storage medium, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the above-mentioned single-loop operating condition classification method by running the computer program.
[0083] An embodiment of the present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the single-loop operating condition classification method in the present application.
[0084] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0085] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0086] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.
[0087] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0088] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0089] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.
[0090] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A single-circuit operating condition classification method, characterized in that: include: Collect historical operation data of a single circuit; Performing multiple rounds of segmented identification on the historical operation data according to multiple data windows of different lengths, respectively, to obtain multiple identification results corresponding to each round of identification, wherein the identification results include model parameters for establishing an identification model, and establishing an identification model based on the identification results to obtain multiple identification models; Respectively obtain first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments, and determine the data segment of the identification model whose first relative deviation value is less than a first preset threshold as the target data segment; The target data segments similar to each other among the multiple target data segments are divided into data segments under the same type of working conditions, and the data segments under the same type of working conditions are merged to obtain data subsets under different working conditions.
2. The method according to claim 1, characterized in that Based on the identification results, an identification model is established to obtain multiple identification models, including: Obtaining the number of rounds of segmented identification, and determining the length of the data window used in each round of segmented identification and the step length of the data window rolling forward in each round of segmented identification, wherein each time the data window rolls forward by one step, a data segment corresponding to the data window is obtained; In each segmented identification round, the historical operation data is divided into a plurality of data segments using the data window according to the step size, and an identification model of each data segment is constructed based on the data in each data segment.
3. The method according to claim 2, characterized in that Respectively obtaining first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments, including: Divide the data segment corresponding to each identification model into multiple sub-data segments; The first relative deviation value is determined based on the standard deviation of the output result corresponding to each sub-data segment and the historical output result.
4. The method according to claim 3, characterized in that Respectively obtaining first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments, including: In the formula, Re represents the first relative deviation value, i represents the number of the sub-segment, m represents the number of sub-segments, and k i Indicates the number of data points in each sub-segment, PV i Represents the historical output results of the ith sub-data segment, ysim i It represents the output result of the identification model of the i-th sub-data segment, and std represents the standard deviation.
5. The method according to claim 1, characterized in that Dividing the target data segments similar to each other among the plurality of target data segments into data segments under the same type of working conditions, comprising: Selecting two data segments to be compared from the plurality of target data segments, and determining recognition models of the two data segments to be compared; The identification models of the two data segments to be compared are respectively used to analyze each other's data segments to obtain output results, and the output results are compared with historical output results. When the comparison results indicate similarity, it is determined that the two data segments to be compared are data segments under the same type of working conditions.
6. The method according to claim 5, characterized in that The identification models of the two data segments to be compared are respectively used to analyze the data segments of the other party to obtain output results, and the output results are compared with historical output results, including: Determine the two data segments to be compared as a first data segment and a second data segment, respectively, wherein the recognition model of the first data segment is a first recognition model, and the recognition model of the second data segment is a second recognition model; Using the first identification model to analyze the second data segment to obtain a first output result, and using the second identification model to analyze the first data segment to obtain a second output result; Obtaining a second relative deviation value between the first output result and a historical output result of the second data segment and a third relative deviation value between the second output result and a historical output result of the first data segment; When at least one of the second relative deviation value and the third relative deviation value is smaller than a second preset threshold, it is determined that the first identification model and the second identification model are similar, and the first data segment and the second data segment belong to the same type of operating condition.
7. The method according to claim 1, characterized in that The data segments under the same type of working conditions are merged to obtain data subsets under different working conditions, including: Get all data segments under the same type of working conditions; The union of all data segments under the same type of working conditions is determined as the data subset of the corresponding working condition.
8. A single-circuit operating condition classification device, characterized in that: include: The acquisition module is used to collect the historical operation data of a single loop; an identification module, configured to perform multiple rounds of segmented identification on the historical operation data according to multiple data windows of different lengths, obtain multiple identification results corresponding to each round of identification, wherein the identification results include model parameters for establishing an identification model, and establish an identification model based on the identification results to obtain multiple identification models; A determination module, used to respectively obtain first relative deviation values of output results of multiple identification models corresponding to each round of identification and historical output results of corresponding data segments, and determine the data segment of the identification model whose first relative deviation value is less than a first preset threshold as the target data segment; The classification module is used to classify the target data segments that are similar to each other among the multiple target data segments into data segments under the same type of working conditions, and merge the data segments under the same type of working conditions to obtain data subsets under different working conditions.
9. A computer device, characterized in that: include: A memory and a processor, wherein the memory is used to store program instructions; The processor is connected to the memory and is used to execute the single-circuit operating condition classification method described in any one of claims 1-7.
10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the single-circuit operating condition classification method described in any one of claims 1-7 is implemented.