Working condition self-adaptive division method and system of thermal power generating unit controlled object

Abstract

This invention provides an adaptive method and system for classifying the operating conditions of a controlled object in a thermal power unit, belonging to the field of power plant technology. The method includes: preprocessing historical correlation data of the target controlled object based on predetermined filtering parameters and improved adaptive filtering rules; wherein the filtering parameters are obtained based on the improved adaptive filtering rules; and classifying the operating conditions of the target controlled object based on the preprocessed historical correlation data and the improved adaptive DBSCN algorithm, obtaining the operating condition classification result of the target controlled object. This achieves adaptive classification of typical operating conditions of thermal power units, effectively solving the problem of inaccurate operating condition classification of controlled objects in thermal power units.

Description

Technical Field

[0001] This invention relates to the field of power plant technology, specifically to an adaptive method for classifying the operating conditions of a controlled object in a thermal power unit, an adaptive system for classifying the operating conditions of a controlled object in a thermal power unit, a machine-readable storage medium, and an electronic device. Background Technology

[0002] Currently, the installed capacity and grid-connected power generation of new energy sources are increasing rapidly. However, due to the volatility, homogeneity, and anti-peak-shaving characteristics of new energy power generation, a large amount of flexible resources are needed to solve the problem of consumption. Thermal power units, due to their unique flexibility advantages, are gradually shifting from a primary energy supply role to a secondary peak-shaving role. This has led to a large number of units undergoing flexibility modifications to adapt to hourly and daily peak-shaving. During the flexibility modifications of thermal power units, improving the stability of the unit's automatic control system requires designing adjustment parameters for different controlled objects under various operating conditions. Currently, the selection of operating conditions for different objects within the unit relies on reviewing long-term historical data or consulting operators, and the classification of operating conditions is not accurate or detailed enough.

[0003] In summary, existing methods for classifying operating conditions mainly rely on consulting operators or independently reviewing historical databases of distributed control systems. The data obtained is inaccurate and fails to accurately determine the different operating conditions corresponding to different objects. Therefore, it is urgent to address the problem of inaccurate operating condition classification for controlled objects in thermal power units. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive method and system for classifying the operating conditions of controlled objects in thermal power units, so as to at least solve the problem of inaccurate classification of operating conditions for different controlled objects of the unit.

[0005] To achieve the above objectives, the first aspect of the present invention provides a method for adaptively partitioning the operating conditions of a controlled object in a thermal power unit, comprising:

[0006] Based on predetermined filtering parameters and improved adaptive filtering rules, historical correlation data of the target controlled object are preprocessed; wherein, the filtering parameters are obtained based on the improved adaptive filtering rules.

[0007] Based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, the working conditions of the target controlled object are divided, and the working condition division results of the target controlled object are obtained.

[0008] Optionally, the filtering parameters mentioned above include the size of the sliding window and the least squares order;

[0009] The improved adaptive filtering rules include:

[0010] Based on the improved SG filtering method, the observed values ​​at time t, the observed values ​​at n sampling times before time t, and the observed values ​​at n sampling times after time t in the historical correlation data of the target controlled object are selected. Curve fitting is performed using a k-1 order polynomial to filter and denoise the historical correlation data of the target controlled object.

[0011] Where n represents the size of the sliding window in the SG filtering method, and k represents the least squares order. The improved SG filtering method is used to adaptively improve the sliding window size n and the least squares order k of the SG filtering method. The rules for adaptive improvement include:

[0012] The size of the sliding window and the least squares order are determined based on the rate of change of the observed data at time t, n0 sampling periods after time t, and n0 sampling periods before time t.

[0013] Optionally, the above preprocessing of the historical correlation data of the target controlled object includes:

[0014] The historical correlation data of the target controlled object are sequentially filtered, denoised, and normalized.

[0015] The rules for normalization include:

[0016] The historical correlation data of the target controlled object after filtering and denoising is normalized to scale the historical correlation data of the target controlled object to a preset range.

[0017] Optionally, the above-mentioned working condition classification of the target controlled object based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm includes:

[0018] A1: Mark the historical associated data of the target controlled object after preprocessing as unvisited sample points;

[0019] A2: Determine any unvisited sample point as the current core point and initialize the cluster set corresponding to the current core point;

[0020] A3: Traverse and access the historical associated data within the predetermined neighborhood radius corresponding to the current core point. If the number of historical associated data within the predetermined neighborhood radius corresponding to the current core point is not less than the predetermined minimum number of points in the neighborhood, then mark the current core point as the final core point.

[0021] A4: Recursively access the historical association data within the predetermined neighborhood radius corresponding to the final core point, and add the historical association data of each target controlled object recursively accessed to the cluster set corresponding to the final core point; wherein, the neighborhood radius and the minimum number of points in the neighborhood are determined based on the improved adaptive DBSCN algorithm;

[0022] A5: When the cluster set corresponding to the final core point can no longer be expanded, repeat A2 to A4 until all unvisited sample points are traversed; among them, historical associated data within the neighborhood radius that does not belong to any cluster set are marked as noise points.

[0023] Optional rules for determining the neighborhood radius and the minimum number of points within the neighborhood include:

[0024] The adaptive neighborhood radius is calculated based on a computational model. The computational model for the adaptive neighborhood radius is as follows: adeps is the adaptive neighborhood radius, n is the number of sample points, and x i For the i-th sample point, dist k For sample point x i The maximum distance between the k nearest neighbors of a sample point;

[0025] Based on the adaptive neighborhood radius, the range of values ​​for k in the k-nearest neighbors is adjusted. After each adjustment of the range of values ​​for k in the k-nearest neighbors, the historical association data of the preprocessed target controlled object is classified based on the DBSCN algorithm. The coverage rate of the classification result of the DBSCN algorithm on the historical association data of the preprocessed target controlled object is calculated, and the k value corresponding to the maximum coverage rate is determined as the minimum number of points in the neighborhood.

[0026] Optionally, the above-mentioned classification of the historical association data of the preprocessed target controlled object based on the DBSCN algorithm, and the calculation of the coverage of the corresponding DBSCN algorithm classification results on the historical association data of the preprocessed target controlled object, include:

[0027] The preprocessed historical correlation data of the target controlled object is divided into multiple grid cells of the same length;

[0028] The coverage base data is determined based on the amount of historical associated data of the target controlled object within each grid cell;

[0029] Based on the classification results of the historical association data of the target controlled object obtained by the DBSCN algorithm, determine the number of historical association data of each category in each grid cell;

[0030] Based on the amount of historical association data for each category within each grid cell, determine the coverage of the DBSCN algorithm's classification results on the preprocessed historical association data of the target controlled object.

[0031] Based on the coverage of the baseline data and the classification results of the DBSCN algorithm on the historical association data of the preprocessed target controlled object, the coverage rate of the DBSCN algorithm classification results on the historical association data of the preprocessed target controlled object is obtained.

[0032] Optionally, the historical correlation data of the aforementioned target controlled object includes historical data of thermal power unit load and historical operating data of the target controlled object;

[0033] The method also includes:

[0034] Collect historical data on the load of thermal power units and historical operating data of the target controlled objects.

[0035] A second aspect of the present invention provides an adaptive operating condition partitioning system for a controlled object of a thermal power unit, comprising:

[0036] The preprocessing module is used to preprocess the historical correlation data of the target controlled object based on predetermined filtering parameters and improved adaptive filtering rules; wherein, the filtering parameters are obtained based on the improved adaptive filtering rules.

[0037] The working condition partitioning module is used to partition the working conditions of the target controlled object based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, and obtain the working condition partitioning results of the target controlled object.

[0038] In a third aspect, the present invention provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the aforementioned adaptive division method for the operating conditions of a controlled thermal power unit.

[0039] In a fourth aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned adaptive division method for the operating conditions of a controlled thermal power unit.

[0040] The above technical solution provides an adaptive partitioning method and system for the operating conditions of controlled objects in thermal power units. This method and system address the constantly changing load demands in power systems, requiring thermal power units to adjust according to real-time load conditions. By partitioning operating conditions, thermal power units can be divided into different operating conditions, enabling them to better meet load adjustment needs. The method and system utilize improved adaptive filtering rules and pre-determined filtering parameters to preprocess historical correlation data of the target controlled object, achieving filtering, gross error processing, and normalization of the historical correlation data. An improved adaptive DBSCN algorithm is used to partition the operating conditions of the target controlled object, transforming a multi-objective optimization problem into a single-objective optimization problem. This allows the computation process to automatically seek the optimal solution, thereby achieving adaptive partitioning of typical operating conditions of thermal power units and effectively solving the problem of inaccurate partitioning of operating conditions for controlled objects in thermal power units.

[0041] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0042] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0043] Figure 1 This is a flowchart of an adaptive division method for the operating conditions of a controlled object in a thermal power unit, provided by one embodiment of the present invention;

[0044] Figure 2 This is a flowchart of an improved adaptive DBSCN algorithm provided in one embodiment of the present invention;

[0045] Figure 3 This is a block diagram of an adaptive operating condition partitioning system for a controlled object of a thermal power unit, provided in one embodiment of the present invention.

[0046] Figure 4 This is a schematic diagram of an electronic device structure provided by a preferred embodiment of the present invention.

[0047] Explanation of reference numerals in the attached figures

[0048] 10 - Electronic device, 100 - Processor, 101 - Memory, 102 - Computer program. Detailed Implementation

[0049] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0050] Figure 1 This is a flowchart of a method for adaptively partitioning the operating conditions of a controlled object in a thermal power unit, provided by one embodiment of the present invention. Figure 1 As shown, this invention provides a method for adaptively partitioning the operating conditions of a controlled object in a thermal power unit, comprising:

[0051] S110: Based on predetermined filtering parameters and improved adaptive filtering rules, the historical correlation data of the target controlled object is preprocessed; wherein, the filtering parameters are obtained based on the improved adaptive filtering rules;

[0052] In some embodiments of this example, the historical associated data of the target controlled object includes historical data of thermal power unit load and historical operating data of the target controlled object; the method further includes: collecting historical data of thermal power unit load and historical operating data of the target controlled object.

[0053] For example, if the target controlled object is a denitrification system, historical data on the load of the thermal power unit and historical operating data on the NOx concentration at the SCR outlet are collected from the DCS.

[0054] Specifically, after collecting the historical correlation data of the target controlled object, the historical correlation data of the target controlled object is preprocessed by filtering and denoising (that is, the historical data of the thermal power unit load and the historical operation data of the target controlled object are filtered and denoised respectively).

[0055] In some embodiments of this example, the filtering parameters include the size of the sliding window and the least squares order. The improved adaptive filtering rule includes: based on the improved SG filtering method, selecting the observation value at time t, the observation value at n sampling times before time t, and the observation value at n sampling times after time t from the historical correlation data of the target controlled object, and using a k-1 order polynomial for curve fitting to filter and denoise the historical correlation data of the target controlled object; where n represents the size of the sliding window of the SG filtering method, and k represents the least squares order. The improved SG filtering method is used to adaptively improve the size n of the sliding window and the least squares order k of the SG filtering method. The adaptive improvement rule includes: determining the size of the sliding window and the least squares order based on the rate of change of the observation data between time t, n0 sampling periods after time t, and n0 sampling periods before time t.

[0056] Specifically, the improved SG filtering method is adopted, and the observations at time t, the n sampling times before time t, and the n sampling times after time t, totaling 2n+1 sampling times, are selected from the historical correlation data of the target controlled object. Curve fitting is performed using a k-1 order polynomial, as shown in formulas (1) to (5):

[0057]

[0058] X (2n+1)×1 =T (2n+1)×k ·A k×1 +E (2n+1)×1 (2);

[0059]

[0060] P = T·A = T·(T) trans ·T) -1 ·Ttrans ·X(4);

[0061]

[0062] Where t is the t-th sampling time, x t For the historical running data at the t-th sampling time, a i Let X be the parameter to be solved, in formula (2) (2n+1)×1 That is, in formula (3), X is the output value to be solved, and A is the output value to be solved. For the least squares solution of A, T in formula (2) (2n+1)×k That is, in formula (3), T is the observed value, P is the filtered output value, and k is the order of the polynomial. The improved SG filtering method can be used to adaptively improve the sliding window size n and the least squares order k of the SG filtering method. The adaptive improvement strategy is as follows: determine the sliding window size and the least squares order k according to the rate of change of the observed value at time t corresponding to the last n0 sampling periods and the first n0 sampling periods, as shown in formulas (6) and (7):

[0063]

[0064] Where n0 is the initial number of cycles, which defaults to 60, and x t For the historical running data at the t-th sampling time, This represents the historical running data after n0 sampling periods. This represents the historical running data from n0 sampling periods ago.

[0065] In the above implementation process, the method uses an improved least squares curve fitting method with a divided time window (i.e., an improved SG filtering method) to filter the historical correlation data of the target controlled object and eliminate gross errors. The method also adaptively adjusts the size of the sliding window and the least squares order according to the rate of change of the observation data before and after the observation at time t.

[0066] In some embodiments of this example, the above-mentioned preprocessing of the historical association data of the target controlled object includes: sequentially performing filtering, denoising and normalization processing on the historical association data of the target controlled object; wherein, the normalization processing rule includes: performing normalization processing on the historical association data of the target controlled object after filtering and denoising processing, so as to scale the historical association data of the target controlled object to a preset range.

[0067] Specifically, the historical correlation data of the target controlled object after filtering and denoising (e.g., historical operating data of unit load and SCR outlet NOx concentration after filtering and denoising) are normalized and scaled to the (0, 1) range to speed up the calculation. The normalization calculation process is as follows: Where x′ is the normalized input variable, x is the input variable, xmin is the minimum value of the variable, and x max This represents the maximum value of the input variable.

[0068] S120: Based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, the working condition of the target controlled object is divided to obtain the working condition division result of the target controlled object.

[0069] Specifically, the improved adaptive DBSCN algorithm is used to divide the working conditions of the target controlled object, so as to transform the multi-objective optimization problem into a single-objective optimization problem, enabling the calculation process to automatically find the optimal solution.

[0070] In the above implementation process, this method addresses the constantly changing load demand in power systems, requiring thermal power units to adjust according to real-time load conditions. By dividing operating conditions, thermal power units can be categorized into different operating states, enabling them to better meet load adjustment needs. This method utilizes improved adaptive filtering rules and pre-determined filtering parameters to preprocess the historical correlation data of the target controlled object, achieving filtering, gross error handling, and normalization of the historical correlation data. The DBSCN algorithm is improved by using an enhanced adaptive DBSCN algorithm to divide the operating conditions of the target controlled object, transforming the multi-objective optimization problem into a single-objective optimization problem. This allows the computation process to automatically seek the optimal solution, thereby achieving adaptive division of typical operating conditions for thermal power units and effectively solving the problem of inaccurate operating condition division of controlled objects in thermal power units.

[0071] In some implementations of this embodiment, the above-mentioned division of the working conditions of the target controlled object based on the preprocessed historical association data of the target controlled object and the improved adaptive DBSCN algorithm includes: A1: marking each historical association data of the preprocessed target controlled object as an unvisited sample point; A2: determining any unvisited sample point as the current core point and initializing the cluster set corresponding to the current core point; A3: traversing and accessing the historical association data within the predetermined neighborhood radius corresponding to the current core point. If the number of historical association data within the predetermined neighborhood radius corresponding to the current core point is not less than the predetermined... If the minimum number of points in the neighborhood is determined, the current core point is marked as the final core point; A4: Recursively access the historical association data within the predetermined neighborhood radius corresponding to the final core point, and add the historical association data of each recursively accessed target controlled object to the cluster set corresponding to the final core point; where the neighborhood radius and the minimum number of points in the neighborhood are determined based on the improved adaptive DBSCN algorithm; A5: When the cluster set corresponding to the final core point can no longer be expanded, repeat A2 to A4 until all unvisited sample points are traversed; where historical association data within the neighborhood radius that does not belong to any cluster set are marked as noise points.

[0072] Please refer to Figure 2 , Figure 2 This is a flowchart of an improved adaptive DBSCN algorithm provided by one embodiment of the present invention. Specifically, the DBSCN algorithm is designed. The DBSCN algorithm searches for regions with a density reaching a given threshold and expands within these regions to form a high-density core region and a low-density noise region. The specific processing flow of the DBSCN algorithm is as follows: First, select the neighborhood radius R and the minimum number of samples M (i.e., the minimum number of points in the neighborhood), and mark the historical association data of the preprocessed target controlled object as unvisited sample points, and set the category number. Select an unvisited sample point as the current core point, and initialize the cluster set corresponding to the current core point. At this time, the cluster set corresponding to the current core point is an empty cluster set. For the current core point, find all sample points (i.e., historical association data) within the eps neighborhood (i.e., neighborhood radius R) of the current core point. If the number of sample points in the neighborhood is not less than P... minIf the minimum number of points in the neighborhood is reached, the current core point is marked as the final core point. For the final core point, sample points within the eps neighborhood of the final core point are recursively visited. All sample points within the radius of the final core point and its neighborhood are marked as visited. The historical association data of the recursively visited target controlled object are added to the cluster set corresponding to the final core point, thereby completing the cluster set expansion operation. If the cluster set corresponding to the final core point cannot be expanded further, a new unvisited sample point is selected as the current core point, and the cluster set expansion operation is repeated until all unvisited sample points are marked as visited. Among them, for unvisited sample points, if they are not within the eps neighborhood of any cluster, they are marked as noise points.

[0073] In some implementations of this embodiment, the rules for determining the neighborhood radius and the minimum number of points within the neighborhood include: calculating an adaptive neighborhood radius based on an adaptive neighborhood radius calculation model; wherein, the adaptive neighborhood radius calculation model is as follows: adeps is the adaptive neighborhood radius, n is the number of sample points, and x i For the i-th sample point, dist k For sample point x i The maximum distance between the k nearest neighbors is calculated. Based on the adaptive neighborhood radius, the range of values ​​for k in the k nearest neighbors is adjusted. After each adjustment, the historical association data of the preprocessed target controlled object is classified according to the DBSCN algorithm. The coverage rate of the classification result of the DBSCN algorithm on the historical association data of the preprocessed target controlled object is calculated. The k value corresponding to the maximum coverage rate is determined as the minimum number of points in the neighborhood.

[0074] In some embodiments of this example, the above-mentioned classification of the historical association data of the preprocessed target controlled object based on the DBSCN algorithm, and the calculation of the coverage rate of the DBSCN algorithm classification result to the historical association data of the preprocessed target controlled object, includes: dividing the historical association data of the preprocessed target controlled object into multiple grid cells of the same length; determining the coverage base data according to the number of historical association data of the target controlled object in each grid cell; determining the number of historical association data of each category in each grid cell according to the classification result of the historical association data of the target controlled object obtained by the DBSCN algorithm; determining the coverage range of the DBSCN algorithm classification result to the historical association data of the preprocessed target controlled object according to the number of historical association data of each category in each grid cell; and obtaining the coverage rate of the DBSCN algorithm classification result to the historical association data of the preprocessed target controlled object according to the coverage base data and the coverage range of the DBSCN algorithm classification result to the historical association data of the preprocessed target controlled object.

[0075] Specifically, this method improves the DBSCN algorithm. The performance of the DBSCAN algorithm largely depends on two key parameters: the neighborhood radius eps and the minimum number of points P in the neighborhood. min The traditional DBSCAN algorithm requires manual setting of these parameters, which is tedious and prone to unstable clustering results. Therefore, this method designs an improved adaptive parameter selection algorithm. The specific processing procedure of the improved DBSCN algorithm is as follows:

[0076] First, the categories generated by the DBSCN algorithm need to cover all sample points as completely as possible. Therefore, coverage rate is proposed as an evaluation metric for the DBSCN algorithm. The coverage rate calculation process is as follows:

[0077] Step 1: Define grid cells, dividing the data space of each dimension into regions of equal length, thereby dividing the historical correlation data of the target controlled object into several grid cells with a side length of ε;

[0078] Step 2: Determine the coverage baseline data based on the number of original sample points falling within each grid cell;

[0079] Step 3: Determine the coverage area by obtaining the number of sample points of all categories in each grid cell using the DBSCN algorithm;

[0080] Step 4: Calculate the coverage rate based on the grid overlap rate between the base data and the coverage area.

[0081] Secondly, the adaptive neighborhood radius adeps is designed, and the calculation formula is as follows:

[0082]

[0083] Where adeps is the adaptive neighborhood radius, n is the number of samples, and x i For the i-th sample point, dist k For sample point x i The maximum distance between the k nearest neighbors is given, where k takes values ​​in the range [2, 20].

[0084] After clarifying the calculation process of coverage and adaptive neighborhood radius, the two-dimensional objective optimization of the DBSCN algorithm is transformed into single-objective optimization. That is, by adjusting the range of values ​​of k in the k-nearest neighbors, the coverage of the DBSCN algorithm's classification result on the original sample set is calculated, and the k value corresponding to the result with the largest coverage is taken as the optimal Pmin (i.e., the minimum number of points in the neighborhood). The optimal k value and adeps are used as DBSCN algorithm parameters for the division of working conditions.

[0085] In the above implementation process, this method addresses the constantly changing load demand in the power system, requiring thermal power units to adjust according to real-time load conditions. By dividing operating conditions, thermal power units can be categorized into different operating states, enabling them to better meet load adjustment needs. First, an improved adaptive filtering method is used, selecting an appropriate sliding window and least squares order to filter the data and handle coarse errors. Further, a max-min normalization method is employed to scale data from different value ranges to the range [0,1]. The DBSCN algorithm is improved by designing coverage as the optimization objective and an adaptive strategy to transform the multi-objective optimization of the DBSCN algorithm into a single-objective optimization problem, thereby achieving adaptive division of typical operating conditions for thermal power units.

[0086] Please refer to Figure 3 , Figure 3 This is a block diagram of an adaptive operating condition partitioning system for a controlled object of a thermal power unit, provided by one embodiment of the present invention. The embodiment of the present invention provides an adaptive operating condition partitioning system for a controlled object of a thermal power unit, comprising:

[0087] The preprocessing module is used to preprocess the historical correlation data of the target controlled object based on predetermined filtering parameters and improved adaptive filtering rules; wherein, the filtering parameters are obtained based on the improved adaptive filtering rules.

[0088] The working condition partitioning module is used to partition the working conditions of the target controlled object based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, and obtain the working condition partitioning results of the target controlled object.

[0089] Specifically, this system addresses the constantly changing load demands within the power system, requiring thermal power units to adjust according to real-time load conditions. By classifying operating conditions, the system categorizes thermal power units into different operating states, enabling them to better meet load regulation needs. The system utilizes improved adaptive filtering rules and pre-determined filtering parameters to preprocess historical correlation data of the controlled object, achieving filtering, gross error handling, and normalization of this data. An improved adaptive DBSCN algorithm is employed to classify the operating conditions of the controlled object, transforming a multi-objective optimization problem into a single-objective optimization problem. This allows for automatic optimization during the computation process, achieving adaptive classification of typical operating conditions for thermal power units and effectively solving the problem of inaccurate operating condition classification for controlled thermal power units.

[0090] The present invention also provides a machine-readable storage medium storing instructions that, when executed by a processor 100, configure the processor 100 to execute the aforementioned adaptive division method for the operating conditions of the controlled thermal power unit.

[0091] Machine-readable storage media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0092] The present invention also provides an electronic device 10, which includes a memory 101, a processor 100, and a computer program 102 stored in the memory 101 and executable on the processor 100. When the processor 100 executes the computer program 102, it implements the above-mentioned adaptive division method of operating conditions of the controlled object of the thermal power unit.

[0093] like Figure 4 The diagram shown is a schematic representation of an electronic device according to an embodiment of the present invention. Figure 4 As shown, the electronic device 10 of this embodiment includes a processor 100, a memory 101, and a computer program 102 stored in the memory 101 and executable on the processor 100. When the processor 100 executes the computer program 102, it implements the steps in the method embodiment described above. Alternatively, when the processor 100 executes the computer program 102, it implements the functions of each module / unit in the device embodiment described above.

[0094] For example, computer program 102 can be divided into one or more modules / units, one or more of which are stored in memory 101 and executed by processor 100 to complete the present invention. One or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of computer program 102 in electronic device 10. For example, computer program 102 can be divided into a preprocessing module and a working condition partitioning module.

[0095] Electronic device 10 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Electronic device 10 may include, but is not limited to, processor 100 and memory 101. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 10 and does not constitute a limitation on electronic device 10. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device may also include input / output devices, network access devices, buses, etc.

[0096] The processor 100 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0097] The memory 101 can be an internal storage unit of the electronic device 10, such as a hard disk or RAM of the electronic device 10. The memory 101 can also be an external storage device of the electronic device 10, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the electronic device 10. Furthermore, the memory 101 can include both internal and external storage units of the electronic device 10. The memory 101 is used to store computer programs and other programs and data required by the electronic device 10. The memory 101 can also be used to temporarily store data that has been output or will be output.

[0098] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0099] Those skilled in the art will understand that embodiments of this application can be provided as a method, system, or computer program 102 product. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program 102 product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0100] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program 102 products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program 102 instructions. These computer program 102 instructions can be provided to a processor 100 of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor 100 of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0101] These computer program 102 instructions may also be stored in a computer-readable storage medium 101 that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium 101 produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0102] These computer program 102 instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0103] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0104] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for adaptively partitioning the operating conditions of a controlled object in a thermal power unit, characterized in that, include: Based on predetermined filtering parameters and improved adaptive filtering rules, historical correlation data of the target controlled object are preprocessed; wherein, the filtering parameters are obtained based on the improved adaptive filtering rules. Based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, the working condition of the target controlled object is divided, and the working condition division result of the target controlled object is obtained. The step of classifying the operating conditions of the target controlled object based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm includes: A1: Mark the historical associated data of the target controlled object after preprocessing as unvisited sample points; A2: Determine any unvisited sample point as the current core point and initialize the cluster set corresponding to the current core point; A3: Traverse and access the historical associated data within the predetermined neighborhood radius corresponding to the current core point. If the number of historical associated data within the predetermined neighborhood radius corresponding to the current core point is not less than the predetermined minimum number of points in the neighborhood, then mark the current core point as the final core point. A4: Recursively access the historical association data within the predetermined neighborhood radius corresponding to the final core point, and add the historical association data of each target controlled object recursively accessed to the cluster set corresponding to the final core point; wherein, the neighborhood radius and the minimum number of points in the neighborhood are determined based on the improved adaptive DBSCN algorithm; A5: When the cluster set corresponding to the final core point can no longer be expanded, repeat A2 to A4 until all unvisited sample points are traversed; among them, historical associated data within the neighborhood radius that does not belong to any cluster set are marked as noise points; The rules for determining the neighborhood radius and the minimum number of points within the neighborhood include: The adaptive neighborhood radius is calculated based on an adaptive neighborhood radius calculation model; wherein, the adaptive neighborhood radius calculation model is as follows: , adeps For adaptive neighborhood radius, n Number of sample points ，x i For the first i One sample point, dist k For sample points x i of k The maximum distance between the nearest neighbor sample points; Based on the adaptive neighborhood radius, the range of values ​​for k in the k-nearest neighbors is adjusted. After each adjustment of the range of values ​​for k in the k-nearest neighbors, the historical association data of the preprocessed target controlled object is classified based on the DBSCN algorithm. The coverage rate of the classification result of the DBSCN algorithm on the historical association data of the preprocessed target controlled object is calculated, and the k value corresponding to the maximum coverage rate is determined as the minimum number of points in the neighborhood.

2. The adaptive division method for the operating conditions of the controlled object of a thermal power unit according to claim 1, characterized in that, The filtering parameters include the size of the sliding window and the least squares order; The improved adaptive filtering rules include: Based on the improved SG filtering method, the observed values ​​at time t, the observed values ​​at n sampling times before time t, and the observed values ​​at n sampling times after time t in the historical correlation data of the target controlled object are selected. Curve fitting is performed using a k-1 order polynomial to filter and denoise the historical correlation data of the target controlled object. Where n represents the size of the sliding window in the SG filtering method, and k represents the least squares order. The improved SG filtering method is used to adaptively improve the sliding window size n and the least squares order k of the SG filtering method. The rules for adaptive improvement include: The size of the sliding window and the least squares order are determined based on the rate of change of the observed data at time t, n0 sampling periods after time t, and n0 sampling periods before time t.

3. The adaptive division method for the operating conditions of the controlled object of a thermal power unit according to claim 1, characterized in that, The preprocessing of historical correlation data of the target controlled object includes: The historical correlation data of the target controlled object are sequentially filtered, denoised, and normalized. The rules for the normalization process include: The historical correlation data of the target controlled object after filtering and denoising is normalized to scale the historical correlation data of the target controlled object to a preset range.

4. The adaptive division method for the operating conditions of the controlled object of a thermal power unit according to claim 1, characterized in that, The process of classifying the historical association data of the preprocessed target controlled object based on the DBSCN algorithm and calculating the coverage of the corresponding DBSCN algorithm classification results on the historical association data of the preprocessed target controlled object includes: The preprocessed historical correlation data of the target controlled object is divided into multiple grid cells of the same length; The coverage base data is determined based on the amount of historical associated data of the target controlled object within each grid cell; Based on the classification results of the historical association data of the target controlled object obtained by the DBSCN algorithm, determine the number of historical association data of each category in each grid cell; Based on the amount of historical association data for each category within each grid cell, determine the coverage of the DBSCN algorithm's classification results on the preprocessed historical association data of the target controlled object. Based on the coverage of the historical association data of the preprocessed target controlled object by the coverage base data and the classification results of the DBSCN algorithm, the coverage of the historical association data of the preprocessed target controlled object by the classification results of the DBSCN algorithm is obtained.

5. The adaptive division method for the operating conditions of the controlled object of a thermal power unit according to claim 1, characterized in that, Historical data related to the target controlled object includes historical load data of thermal power units and historical operating data of the target controlled object; The method further includes: Collect historical data on the load of thermal power units and historical operating data of the target controlled objects.

6. A condition adaptive partitioning system for a controlled object in a thermal power unit, characterized in that, The method for adaptively dividing the operating conditions of a controlled thermal power unit as described in any one of claims 1 to 5 includes: A preprocessing module is used to preprocess the historical correlation data of the target controlled object based on predetermined filtering parameters and improved adaptive filtering rules; wherein the filtering parameters are obtained based on the improved adaptive filtering rules. The working condition partitioning module is used to partition the working conditions of the target controlled object based on the preprocessed historical correlation data of the target controlled object and the improved adaptive DBSCN algorithm, and obtain the working condition partitioning results of the target controlled object.

7. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by the processor, the instruction causes the processor to be configured to perform the adaptive division method of the operating conditions of the controlled object of the thermal power unit as described in any one of claims 1 to 5.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the adaptive division method for the operating conditions of the controlled object of the thermal power unit as described in any one of claims 1 to 5.

Images