Construction method and system for intelligent agent of thermal power plant

Abstract

The invention provides a construction method and system for an intelligent agent of a thermal power plant, and belongs to the technical field of intelligent agents, and the method specifically comprises the steps: determining the adjustment data of unit equipment in different equipment adjustment processes, and combining the composition data of the unit equipment provided with the intelligent agent in the different equipment adjustment processes, determining a potential setting target of an intelligent agent in the unit equipment, and when composition data of the potential setting target in different equipment adjustment processes meet requirements, determining the setting target of the intelligent agent in the unit equipment according to the potential setting target data in combination with association conditions of the potential setting target in the different equipment adjustment processes. And determining the construction method of the agent of the potential setting target. The construction of the agent from the angle of the unit adjustment demand of the thermal power plant is realized.

Description

Technical Field

[0001] This invention belongs to the field of intelligent agent technology, and in particular relates to a method and system for constructing intelligent agents for thermal power plants. Background Technology

[0002] To improve the operational reliability of thermal power plants, constructing intelligent agents within the unit equipment can significantly enhance the reliability of equipment regulation and processing during operation, thereby improving the operational stability of the thermal power plant. However, the following drawbacks exist: In the process of building intelligent agents in thermal power plants, existing technical solutions neglect the interrelationships between different unit equipment. Specifically, when adjusting equipment, it is often necessary to adjust multiple unit equipment simultaneously. Therefore, how to identify potential setting targets in equipment that has not yet been built into intelligent agents based on the interrelationships of unit equipment during different equipment adjustment processes in thermal power plants, and to conduct targeted monitoring and analysis of operational data in advance, thereby improving the efficiency of intelligent agent construction and processing, has become an urgent technical problem to be solved.

[0003] To address the aforementioned technical problems, this application provides a method and system for constructing intelligent agents for thermal power plants. Summary of the Invention

[0004] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a method for constructing an intelligent agent for a thermal power plant, which includes: S1 uses the regulation data of the thermal power plant as a basis to determine the regulation data of the thermal power plant in different equipment regulation processes, and combines the deviation of the regulation data in different equipment regulation processes to determine the construction requirement type of the intelligent agent of the thermal power plant. When the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type, proceed to the next step. S2 determines the adjustment data of the unit equipment during different equipment adjustment processes, and combines it with the composition data of the unit equipment with intelligent agents in different equipment adjustment processes to determine the potential setting target of the intelligent agent in the unit equipment. When the composition data of the potential setting target in different equipment adjustment processes meets the requirements, proceed to the next step. S3 uses the potential setting target data and the association between potential setting targets in different device adjustment processes to determine the method for constructing the intelligent agent of the potential setting target.

[0005] The beneficial effects of this invention are as follows: Based on the adjustment data of the unit equipment during different equipment adjustment processes and the composition data of the unit equipment with intelligent agents in different equipment adjustment processes, the potential setting targets of the intelligent agents in the unit equipment are determined. This takes into account both the interaction needs with other unit equipment with intelligent agents and the frequency of adjustment processing of the unit equipment during the equipment adjustment process. Thus, the potential setting targets are determined from the perspective of the setting processing needs of the unit equipment, thereby improving the efficiency of intelligent agent construction and processing.

[0006] The method for constructing the intelligent agent of the potential setting target is determined by considering the potential setting target data and the correlation between potential setting targets in different equipment adjustment processes. This method ensures that, given a large number of potential setting targets, it enables timely data analysis and processing of the operational data of potential setting targets when there are a large number of equipment adjustment processes with the same potential setting targets. Thus, it can guarantee the adjustment stability and reliability of most equipment adjustment processes with limited data resources.

[0007] Furthermore, the regulation data of the thermal power plant includes the number of equipment regulation processes of the thermal power plant and the unit equipment that is regulated in different equipment regulation processes.

[0008] Furthermore, the equipment adjustment process refers to the process by which operators adjust the equipment according to needs during operation.

[0009] Furthermore, the deviation of the adjustment data includes determining the differences in the unit equipment between different equipment adjustment processes.

[0010] Furthermore, the method for determining the construction requirement type of the intelligent agent in the thermal power plant is as follows: Based on the adjustment data of the thermal power plant during different equipment adjustment processes, determine the equipment adjustment process data of the thermal power plant within the most recent preset time period; Based on the deviation of adjustment data during different equipment adjustment processes, the adjustment processes of units with consistent equipment are grouped into the same adjustment process combination; Based on the equipment adjustment process data of the thermal power plant in the most recent preset time period, as well as the combined adjustment process data, the construction requirement type of the intelligent agent of the thermal power plant is determined.

[0011] Furthermore, the method for determining the construction method of the potential target intelligent agent is as follows: Based on the potential setting target data, the number of potential setting targets is determined, and the device adjustment process with the potential setting target adjustment is used as the matching adjustment process for the potential setting target. Based on the correlation of potential setting targets in different equipment adjustment processes, the overlap of potential setting targets in the matching adjustment process of the potential setting targets is determined, and matching adjustment processes with consistent potential setting targets are grouped into the same matching process combination; Based on the number of potential target settings and the matching adjustment process data of potential target settings in different matching process combinations, the method for constructing the agent of the potential target settings is determined.

[0012] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for constructing an intelligent agent for a thermal power plant when running the computer program.

[0013] Other features and advantages will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0014] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0015] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart of a method for constructing intelligent agents for thermal power plants; Figure 2 This is a flowchart illustrating the method for determining the construction requirements of intelligent agents in a thermal power plant; Figure 3 This is a flowchart of a method for determining the potential setting targets of intelligent agents in unit equipment. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] Example 1 To solve the above problems, according to one aspect of the present invention, such as Figure 1 As shown, a method for constructing an intelligent agent for a thermal power plant is provided, specifically including: S1 uses the regulation data of the thermal power plant as a basis to determine the regulation data of the thermal power plant in different equipment regulation processes, and combines the deviation of the regulation data in different equipment regulation processes to determine the construction requirement type of the intelligent agent of the thermal power plant. When the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type, proceed to the next step. Furthermore, the regulation data of the thermal power plant includes the number of equipment regulation processes of the thermal power plant and the unit equipment that is regulated in different equipment regulation processes.

[0019] Furthermore, the equipment adjustment process refers to the process by which operators adjust the equipment according to needs during operation.

[0020] Furthermore, the deviation of the adjustment data includes determining the differences in the unit equipment between different equipment adjustment processes.

[0021] Specifically, such as Figure 2 As shown, the method for determining the construction requirement type of the intelligent agent in the thermal power plant is as follows: The purpose of this approach is to automatically determine whether a thermal power plant needs and is worthwhile to build an "intelligent agent." Here, "intelligent agent" can be understood as an AI system capable of autonomously or semi-autonomously optimizing complex operations. The core basis for this judgment is analyzing the "complexity" and "diversity" of the power plant's historical regulatory operations. If the power plant's operating mode is simple and repetitive, the need for building an intelligent agent is low; if the operating mode is complex and variable, then building an intelligent agent is highly valuable.

[0022] Based on the adjustment data of the thermal power plant during different equipment adjustment processes, determine the equipment adjustment process data of the thermal power plant within the most recent preset time period; Based on the deviation of adjustment data during different equipment adjustment processes, the adjustment processes of units with consistent equipment are grouped into the same adjustment process combination; Based on the equipment adjustment process data of the thermal power plant in the most recent preset time period, as well as the combined adjustment process data, the construction requirement type of the intelligent agent of the thermal power plant is determined.

[0023] It is understood that, based on the equipment adjustment process data of the thermal power plant within the most recent preset time period, and the combined adjustment process data, the construction requirement type of the intelligent agent of the thermal power plant is determined, specifically including: Determine whether the number of equipment adjustment processes in the thermal power plant within the most recent preset time period is greater than the preset adjustment process number threshold. If yes, determine that the construction requirement type of the intelligent agent of the thermal power plant is the target requirement type. If no, proceed to the next step. Step 1: Operation Frequency Check (High Frequency). Judgment criteria: If the total number of equipment adjustment processes within the preset cycle > the preset threshold for the number of adjustment processes, then: Yes: This indicates that the power plant's adjustments are very frequent and the operating conditions are complex. Immediately determine it as a "target demand type". No: This indicates that the adjustments are not particularly frequent, and proceed to the next step.

[0024] Determine whether the number of equipment adjustment processes in the thermal power plant within the most recent preset time period is less than the preset process number threshold (less than the preset adjustment process number threshold). If yes, determine that the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type. If no, proceed to the next step. Step 2: Operation sparsity check (low frequency). Judgment criteria: Within a preset period, the total number of equipment adjustment processes < the preset process number threshold (a lower threshold). If yes: it indicates that the power plant is operating extremely stably with very few adjustments. Immediately determine it as a "non-target demand type" (no need to build an agent). If no: it indicates that the adjustment frequency is at a medium level, and further analysis of its pattern is needed before proceeding to the next step.

[0025] Based on the combined data of the adjustment process, determine the proportion of the number of equipment adjustment processes in different combinations of adjustment processes within the most recent preset time period to the total number of equipment adjustment processes, and use this as the adjustment matching factor of the equipment adjustment process combination. Determine whether there are any adjustment process combinations with an adjustment matching factor greater than a preset matching factor threshold. If so, determine that the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type. If not, proceed to the next step. Step 3: Operation Mode Repeatability Check (Single Mode). Judgment criteria: Does a certain combination of control processes (i.e., operations regulating the same group of units) have a control matching factor (the proportion of this combination of operations to the total number of operations) > the preset matching factor threshold? If yes: This indicates that the power plant's control operations are highly repetitive, mainly concentrated in one or a few fixed modes. This fixed mode can be handled using traditional control logic or simple rules, and is judged as a "non-target demand type." If no: This indicates that the operation modes are diverse, with no obvious single dominant mode, proceed to the next step.

[0026] Determine whether the number of equipment adjustment process combinations is greater than a preset process combination number threshold. If yes, determine that the construction requirement type of the intelligent agent of the thermal power plant belongs to the target requirement type. If no, proceed to the next step. Step 4: Operation Mode Diversity Check (Diverse Modes). Judgment Criteria: If the number of different control process combinations > the preset process combination number threshold, it indicates that the power plant needs to cope with a variety of different equipment combination operating conditions, resulting in high complexity. It is determined to be a "target demand type". If not, it indicates that there are not many types of operation modes, and proceed to the final step.

[0027] Based on the number of unit equipment adjusted during different equipment adjustment processes, it is determined whether the average number of unit equipment adjusted during different equipment adjustment processes is less than a preset number threshold. If not, it is determined that the construction requirement type of the intelligent agent of the thermal power plant belongs to the target requirement type. If yes, it is determined that the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type.

[0028] Step 5: Operational Complexity Check (How many devices are involved in a single operation?). Judgment Criteria: Calculate the average number of units involved in each adjustment process and determine if it is less than the preset threshold. If not, it indicates that multiple devices need to be linked in each adjustment, resulting in high coordination complexity. This is classified as a "target demand type." If yes, it indicates that each adjustment mainly targets one or a few devices, resulting in low complexity. This is classified as a "non-target demand type."

[0029] Equipment adjustment process: A complete operational adjustment made to adapt to load changes, such as "increasing fuel in boiler #1 while simultaneously reducing the inlet valve opening of turbine #2". Adjustment process combination: Multiple equipment adjustment processes that are identical to the set of equipment being adjusted are grouped together. For example, all operations that simultaneously adjust "boiler #1 and turbine #2" belong to the same combination. Adjustment matching factor: The proportion of times a certain adjustment process combination occurs out of the total number of adjustments. The higher this factor, the more singular and repetitive the operation mode.

[0030] Scenario settings: Preset time period: the last 30 days, preset threshold: preset adjustment process number threshold (high): 150 times, preset process number threshold (low): 50 times, preset matching factor threshold: 40%, preset process combination number threshold: 8 types, preset number threshold (average number of adjustment devices): 2.5 units.

[0031] Data analysis results: Total number of equipment adjustment processes: 120 adjustments occurred within 30 days.

[0032] Combined analysis of the regulation process: These 120 adjustments were categorized into 10 different combinations of adjustment processes, depending on the different combinations of equipment used.

[0033] Among them, the two most important combinations are: Combination A (regulating #1 boiler & #1 turbine): occurred 30 times, Combination B (regulating #2 boiler & #2 turbine): occurred 25 times, and the remaining 8 combinations occurred a total of 65 times.

[0034] Average number of units involved in a single regulation: Calculations show that an average of 2.8 generating units are involved in each regulation.

[0035] Decision-making process: Step 1 (Frequency Check): Total number of adjustments (120 times) < High threshold (150 times), Conclusion: Condition not met, proceed to Step 2.

[0036] Step 2 (Sparseness check): Total number of adjustments (120 times) > Low threshold (50 times), Conclusion: Condition not met, proceed to step 3.

[0037] Step 3 (Repeatability Check): Calculate the moderating matching factor of the main combinations: the matching factor of combination A = 30 / 120 = 25%, the matching factor of combination B = 25 / 120 ≈ 20.8%. The matching factors of all combinations do not exceed the threshold (40%). Conclusion: There is no highly repetitive pattern. Proceed to Step 4.

[0038] Step 4 (Diversity Check): The number of adjustment process combinations (10) > the threshold (8), conclusion: the condition is met! Therefore, the intelligent agent construction requirement type of this thermal power plant is determined to be the "target requirement type".

[0039] Final decision and follow-up actions Since the thermal power plant has been identified as belonging to the "target demand type," the system will activate advanced monitoring and data acquisition mode: Action: During all subsequent equipment adjustments, real-time acquisition and analysis of monitoring data from the participating units (such as boilers #1 / #2 and steam turbines #1 / #2) will be conducted. This high-quality, high-value data will be recorded as the core data foundation for later construction and training of a dedicated intelligent agent for thermal power plants. This agent will learn how to make optimal coordinated control decisions in this complex and variable operating environment.

[0040] It should be noted that when the construction requirement type of the intelligent agent of the thermal power plant belongs to the target requirement type, the monitoring data of the unit equipment that is adjusted during the equipment adjustment process will be acquired, analyzed and processed in real time, so as to lay a targeted foundation for the construction of the intelligent agent in the later stage.

[0041] S2 determines the adjustment data of the unit equipment during different equipment adjustment processes, and combines it with the composition data of the unit equipment with intelligent agents in different equipment adjustment processes to determine the potential setting target of the intelligent agent in the unit equipment. When the composition data of the potential setting target in different equipment adjustment processes meets the requirements, proceed to the next step. Specifically, such as Figure 3 As shown, the method for determining the potential setting targets of the intelligent agents in the unit equipment is as follows: Based on the adjustment data of the unit equipment during different equipment adjustment processes, the equipment adjustment process in which the unit equipment adjustment processing exists is determined, and it is used as the matching adjustment process; Based on the composition data of the unit equipment with intelligent agents in different matching and adjustment processes, the proportion of the number of unit equipment with intelligent agents in different matching and adjustment processes is determined and used as the proportion of the number of intelligent agents. Based on the matching and adjustment process data and the proportion of settings in different matching and adjustment processes, the potential setting targets of the intelligent agents in the unit equipment are determined.

[0042] If we are evaluating whether "#1 turbine" needs to be set up with an agent, the preset time period is the last 30 days.

[0043] Preset thresholds: Preset percentage threshold: 30%, Preset setting quantity percentage threshold: 60%, Preset demand factor threshold: 0.4.

[0044] Data analysis results: Total number of equipment adjustment processes: 100 times in 30 days, matching adjustment processes (including adjustment of #1 steam turbine): 35 times, matching adjustment ratio = 35 / 100 = 35%.

[0045] Analysis of the percentage of settings involved in the matching and adjustment process: Of the 35 operations, we sampled three representative cases: Process A: Involving 4 devices, 2 of which had agents. Percentage of settings = 2 / 4 = 50%; Process B: Involving 3 devices, 2 of which had agents. Percentage of settings ≈ 66.7%; Process C: Involving 5 devices, 4 of which had agents. Percentage of settings = 80%. We calculated the average percentage of settings across these 35 operations, assuming it to be 65%.

[0046] Furthermore, based on the matching and adjustment process data and the proportion of settings in different matching and adjustment processes, the potential setting targets of the intelligent agents in the unit equipment are determined, specifically including: Obtain the percentage of the number of matching adjustment processes in all device adjustment processes within the most recent preset time period, and use it as the matching adjustment percentage. Determine whether the matching adjustment percentage is greater than a preset percentage threshold. If yes, determine that the unit device is a potential setting target of the intelligent agent. If no, proceed to the next step. Step 1 (Participation Check): Matching adjustment percentage (35%) > preset percentage threshold (30%), Conclusion: Condition met! Therefore, immediately identify "#1 steam turbine" as the potential target for the agent.

[0047] Based on the proportion of the number of settings in different matching and adjustment processes, determine whether there is a matching and adjustment process where the proportion of the number of settings is less than the preset threshold. If not, determine that the unit equipment is not a potential setting target of the intelligent agent. If so, proceed to the next step. Assuming a scenario variation: The matching adjustment percentage is 25%. Step 1 (engagement check): 25% < 30%, the condition is not met, proceed to step 2.

[0048] Step 2 (Intelligent Coverage Check): It is necessary to check whether there is a matching adjustment process where the percentage of settings is less than 60%. According to the data, the percentage of settings in process A (50%) is less than the threshold (60%). Conclusion: The condition is met (there is an adjustment process with insufficient intelligence). Therefore, proceed to the next step.

[0049] Based on the proportion of the number of settings in different matching and adjustment processes and the proportion of matching and adjustment, a setting demand factor is determined. When the setting demand factor is greater than a preset demand factor threshold, the unit equipment is determined to be a potential setting target of the intelligent agent.

[0050] Step 3 (Comprehensive Demand Assessment): Calculate the setting demand factor: Basic Target 1 (Participation) = Matching Adjustment Ratio = 0.25, Basic Target 2 (Non-Intelligent Level) = 1 - Average Setting Quantity Ratio = 1 - 0.5 = 0.5, Setting Demand Factor = (0.25 + 0.5) / 2 = 0.375, Judgment: Setting demand factor (0.375) < Preset demand factor threshold (0.40), Conclusion: It is determined that "#1 Steam Turbine" does not belong to the potential setting target.

[0051] It should be noted that the setting demand factor is determined based on the difference between the matching adjustment ratio and the average of the setting quantity ratio in different matching adjustment processes as the basic target, and the average of the two basic targets is used as the setting demand factor.

[0052] Furthermore, it is determined whether the constituent data of the potential setting targets in different equipment adjustment processes meet the requirements, specifically including: Based on the composition data of potential set targets in different equipment adjustment processes, determine the number of potential set targets in different equipment adjustment processes; Based on the number of potential setting targets in different equipment adjustment processes within the most recent preset time period, determine whether the composition data of potential setting targets in different equipment adjustment processes meet the requirements.

[0053] It is understandable that if, within the most recent preset time period, the average number of potential setting targets in different equipment adjustment processes exceeds a preset number threshold, then it is determined that the composition data of potential setting targets in different equipment adjustment processes does not meet the requirements.

[0054] Preset time period: the last 30 days; preset quantity threshold: 2.5 (this threshold can be set according to the team's development capabilities).

[0055] Data analysis results: We analyzed 100 device adjustment processes over 30 days. In each adjustment process, we identified the "potential setup targets" (i.e. devices that were determined to require an intelligent agent based on the previous method).

[0056] We tallied the number of these targets in each adjustment, and some data are as follows: Adjustment 1: Involved 3 potential targets. Adjustment 2: Involved 2 potential targets. Adjustment 3: Involved 4 potential targets.

[0057] Calculations showed that the average number of "potential setting targets" in these 100 adjustments was 3.1. Compared with the threshold, 3.1 > 2.5. Therefore, it was determined that the composition data of potential setting targets in different equipment adjustment processes did not meet the requirements.

[0058] Final decision and follow-up actions Since the judgment result is "not meeting the requirements", the system will activate the global backup plan: Action: The method for constructing intelligent agents with potential target settings is to acquire, analyze and process monitoring data in real time during equipment adjustment, thereby laying a targeted foundation for the construction of intelligent agents in the later stage.

[0059] This means: pausing the plan to immediately develop intelligent agents for all potential targets. Instead, launching a large-scale, high-frequency data collection campaign where operational data (such as temperature, pressure, flow rate, vibration, valve opening, etc.) of all units and equipment marked as "potential setting targets" will be recorded, stored, and preliminarily analyzed in real time during each adjustment process.

[0060] It should be noted that if the constituent data of potential setting targets in different equipment adjustment processes do not meet the requirements, then the method for constructing the intelligent agent for determining potential setting targets is to acquire, analyze and process the monitoring data in real time during the equipment adjustment process, thereby laying a targeted foundation for the construction of intelligent agents in the later stage.

[0061] S3 uses the potential setting target data and the association between potential setting targets in different device adjustment processes to determine the method for constructing the intelligent agent of the potential setting target.

[0062] Furthermore, the method for determining the construction method of the potential target intelligent agent is as follows: This method intelligently allocates limited data acquisition resources when the system determines that there are few potential setup targets (meaning the data meets the requirements). Its core idea is to prioritize high-quality data acquisition for device groups that work together in the "most common and typical operating scenarios," as the data from these scenarios has the highest value.

[0063] Based on the potential setting target data, the number of potential setting targets is determined, and the device adjustment process with the potential setting target adjustment is used as the matching adjustment process for the potential setting target. Based on the correlation of potential setting targets in different equipment adjustment processes, the overlap of potential setting targets in the matching adjustment process of the potential setting targets is determined, and matching adjustment processes with consistent potential setting targets are grouped into the same matching process combination; Based on the number of potential target settings and the matching adjustment process data of potential target settings in different matching process combinations, the method for constructing the agent of the potential target settings is determined.

[0064] It is understood that the method for constructing an agent for determining the potential target based on the number of potential target settings and the matching adjustment process data of the potential target settings in different matching process combinations specifically includes: If the number of potential target objects is less than a preset threshold, then the method for constructing the intelligent agents of all potential target objects is to acquire, analyze and process the monitoring data during the equipment adjustment process in real time, thereby laying a foundation for the construction of intelligent agents in the later stage. If not, proceed to the next step. Step 1: Scale Check (Are there too many targets?). Judgment criteria: If the number of potential targets is less than the preset threshold, then: the number of target devices is not large, and deployment can be widespread. Therefore, comprehensive real-time data collection should be performed on all potential targets. If not, then the number of target devices is large, resources are limited, and priority must be given to other options. Proceed to the next step.

[0065] Based on the matching adjustment process data of the potential setting targets in different matching process combinations, determine the number of matching adjustment processes in different matching process combinations, and determine whether there is a matching process combination with a number of matching adjustment processes greater than a preset adjustment process number threshold. If not, determine that the construction method of the intelligent agent of all potential setting targets is to acquire, analyze and process the monitoring data during the equipment adjustment process in real time, so as to lay a foundation for the construction of intelligent agents in the later stage. If yes, proceed to the next step. Step 2: High-frequency scene identification (Does a very common device combination exist?). Judgment criteria: Does the number of matching adjustment processes for a certain matching process combination (i.e., the same group of potential setting targets) exceed the preset adjustment process number threshold? If yes, it means that there are one or more "high-frequency device combinations". If so, then a high-frequency combination exists, proceed to the next step.

[0066] If there are no high-frequency combinations, it indicates that the coordination between different combinations is low. Therefore, the composition of different equipment changes significantly during the adjustment process. In order to improve the reliability of the adjustment, the construction method of the intelligent agent for all potential setting targets is determined to acquire, analyze and process the monitoring data in real time during the equipment adjustment process, thereby laying a foundation for the construction of intelligent agents in the later stage.

[0067] Based on the similarity of the adjusted unit equipment in different matching process combinations and different matching adjustment processes, matching adjustment processes with identical adjusted unit equipment are identified. The overlap factor is determined based on the proportion of matching adjustment processes with identical adjusted unit equipment in the matching process combinations. It is then determined whether the overlap factor is less than a preset overlap factor threshold. If so, the method for constructing the intelligent agent of the potential setting target is to perform real-time acquisition, analysis, and processing of monitoring data during equipment adjustment processes in all potential setting targets in matching process combinations where the number of matching adjustment processes exceeds the preset threshold, thereby laying a targeted foundation for the subsequent construction of the intelligent agent. If not, the method for constructing the intelligent agent of the potential setting target is that real-time acquisition, analysis, and processing of monitoring data during equipment adjustment processes is only required when the number of non-response times of the potential setting target does not meet the requirements, thereby laying a targeted foundation for the subsequent construction of the intelligent agent.

[0068] Step 3: Pattern Consistency Check (Are the operating patterns fixed in common scenarios?). Judgment criteria: Calculate the overlap factor (the proportion of devices with completely identical operating patterns within the same matching process combination) and determine if it is less than the preset overlap factor threshold. If yes (low overlap factor): This indicates that even if the same group of devices is frequently adjusted, the specific objects and methods of adjustment differ significantly each time. This represents a highly complex and variable operating condition, requiring significant data to understand its patterns. Therefore, an active strategy is adopted to collect data for the targets in these combinations. If no (high overlap factor): This indicates that the operating patterns of common device combinations are very fixed and repetitive. Such fixed patterns can be handled well using traditional methods, and the urgency of data collection is relatively low. Therefore, a conservative strategy is adopted, triggering data collection only when the device's "number of unresponsive events" exceeds the limit (i.e., poor performance).

[0069] Furthermore, if the number of unresponsive events exceeds a preset threshold, then it is determined that the number of unresponsive events for the potential target does not meet the requirements.

[0070] Matching process grouping: Grouping multiple equipment regulation processes with the same set of potential setting targets into one category. For example, all regulation processes that simultaneously include "#1 boiler intelligent agent and #2 steam turbine intelligent agent" as potential targets belong to the same group.

[0071] Matching adjustment processes with identical unit equipment: This refers to two adjustments where the specific unit equipment lists (including non-potential targets) actually operated on are completely identical within the same matching process combination. This indicates that the operational context of the two adjustments is exactly the same.

[0072] Overlap factor: Within a given combination of matching processes, the proportion of regulation processes with identical operating modes out of the total number of regulation processes in that combination. The higher this factor, the more singular and repetitive the operations are within that combination.

[0073] Number of unresponsive events: This refers to the number of times a potential target device fails to achieve the expected control effect after an adjustment command is issued. This is an indicator for measuring equipment performance or the effectiveness of an existing control system.

[0074] Assume the potential setup target: Five devices that require intelligent agents have been identified, let's say: {A, B, C, D, E}.

[0075] Preset thresholds: Preset threshold for the number of potential target settings: 3, Preset threshold for the number of adjustment processes: 15, Preset threshold for overlap factor: 60%, Preset threshold for the number of non-response times: 5 times / month.

[0076] Data analysis results: Total number of potential targets = 5. Matching process combination analysis: We found that in historical data, potential targets often appear in fixed combinations, forming two main matching process combinations: Combination 1: Potential targets are {A, B, C}. This combination appears 20 times (number of matching adjustments). Combination 2: Potential targets are {D, E}. This combination appears 8 times.

[0077] Pattern consistency analysis of combination 1: In the 20 adjustments of combination 1, the specific equipment list operated in 16 adjustments is exactly the same (for example, all of them operate equipment A, B, C, F, and G). The overlap factor of combination 1 = 16 / 20 = 80%.

[0078] Decision-making process: Step 1 (scale check): Potential target number (5) > threshold (3), conclusion: condition not met, proceed to step 2.

[0079] Step 2 (High-frequency scene recognition): Check the occurrence frequency of each matching process combination: Combination 1 occurs 20 times > threshold (15 times), Combination 2 occurs 8 times < threshold (15 times).

[0080] The condition is met (a high-frequency combination, combination 1, exists). Therefore, proceed to the next step: calculate the overlap factor of combination 1 as 80%. 80% > the preset overlap factor threshold (60%), so the condition is not met (high overlap factor, fixed pattern). Therefore, the final decision is: the method for constructing the agent of the potential setting target is—data collection is only required if and only if the number of non-response times of the potential setting target does not meet the requirements.

[0081] In practice: The system checks the "number of unresponsive events" for devices A, B, C, D, and E monthly. For example, if device C has 7 unresponsive events, exceeding the threshold of 5, then real-time data collection will only be initiated for device C. Data collection will not be initiated for devices A, B, D, and E.

[0082] Example 2 In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for constructing an intelligent agent for a thermal power plant when running the computer program.

[0083] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0084] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0085] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for constructing an intelligent agent for a thermal power plant, characterized in that, Specifically, it includes: S1 uses the regulation data of the thermal power plant as a basis to determine the regulation data of the thermal power plant in different equipment regulation processes, and combines the deviation of the regulation data in different equipment regulation processes to determine the construction requirement type of the intelligent agent of the thermal power plant. When the construction requirement type of the intelligent agent of the thermal power plant does not belong to the target requirement type, proceed to the next step. S2 determines the adjustment data of the unit equipment during different equipment adjustment processes, and combines it with the composition data of the unit equipment with intelligent agents in different equipment adjustment processes to determine the potential setting target of the intelligent agent in the unit equipment. When the composition data of the potential setting target in different equipment adjustment processes meets the requirements, proceed to the next step. S3 uses the potential setting target data and the association between potential setting targets in different device adjustment processes to determine the method for constructing the intelligent agent of the potential setting target.

2. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, The regulation data of the thermal power plant includes the number of equipment regulation processes in the thermal power plant and the unit equipment that is regulated in different equipment regulation processes.

3. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, The deviation of the adjustment data includes determining the differences in the unit equipment between different equipment adjustment processes.

4. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, The method for determining the construction requirement type of the intelligent agent in the thermal power plant is as follows: Based on the adjustment data of the thermal power plant during different equipment adjustment processes, determine the equipment adjustment process data of the thermal power plant within the most recent preset time period; Based on the deviation of adjustment data during different equipment adjustment processes, the adjustment processes of units with consistent equipment are grouped into the same adjustment process combination; Based on the equipment adjustment process data of the thermal power plant in the most recent preset time period, as well as the combined adjustment process data, the construction requirement type of the intelligent agent of the thermal power plant is determined.

5. The method for constructing an intelligent agent for a thermal power plant as described in claim 4, characterized in that, Based on the equipment adjustment process data of the thermal power plant within the most recent preset time period, and the combined adjustment process data, the construction requirement type of the intelligent agent of the thermal power plant is determined, specifically including: When the number of equipment adjustment processes in the thermal power plant within the most recent preset time period exceeds the preset threshold for the number of adjustment processes, the construction requirement type of the intelligent agent of the thermal power plant is determined to be the target requirement type.

6. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, When the construction requirement type of the intelligent agent of the thermal power plant belongs to the target requirement type, the monitoring data of the unit equipment that is adjusted during the equipment adjustment process will be acquired, analyzed and processed in real time, thereby laying a targeted foundation for the construction of the intelligent agent in the later stage.

7. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, Determine whether the constituent data of potential setting targets meet the requirements during different equipment adjustment processes, specifically including: Based on the composition data of potential set targets in different equipment adjustment processes, determine the number of potential set targets in different equipment adjustment processes; Based on the number of potential setting targets in different equipment adjustment processes within the most recent preset time period, determine whether the composition data of potential setting targets in different equipment adjustment processes meet the requirements.

8. The method for constructing an intelligent agent for a thermal power plant as described in claim 7, characterized in that, If, within the most recent preset time period, the average number of potential setting targets in different equipment adjustment processes exceeds a preset number threshold, then it is determined that the composition data of potential setting targets in different equipment adjustment processes does not meet the requirements.

9. The method for constructing an intelligent agent for a thermal power plant as described in claim 1, characterized in that, The method for determining the construction method of the potential target intelligent agent is as follows: Based on the potential setting target data, the number of potential setting targets is determined, and the device adjustment process with the potential setting target adjustment is used as the matching adjustment process for the potential setting target. Based on the correlation of potential setting targets in different equipment adjustment processes, the overlap of potential setting targets in the matching adjustment process of the potential setting targets is determined, and matching adjustment processes with consistent potential setting targets are grouped into the same matching process combination; Based on the number of potential target settings and the matching adjustment process data of potential target settings in different matching process combinations, the method for constructing the agent of the potential target settings is determined.

10. A computer system, comprising: A memory and processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes a method for constructing an intelligent agent for a thermal power plant as described in any one of claims 1-9.

Images