Water-turbine generator set startup and shutdown intelligent decision-making method, system and equipment based on multi-constraint coupling and medium

Abstract

The invention discloses an intelligent decision-making method, system and equipment for startup and shutdown of a water-turbine generator set based on multi-constraint coupling and a medium, and relates to the technical field of intelligent decision-making. Establishing a model based on the operation data set, and generating operation state prediction results of the unit in different startup and shutdown and load distribution scenes; constructing a multi-dimensional coupling constraint system, and forming a feasible decision space of a unit startup and shutdown combination; executing multi-target comprehensive optimization, a unit startup and shutdown scheme and a load distribution result in the feasible decision space; based on a preset early warning rule, carrying out rationality check on the unit startup and shutdown scheme and the load distribution result; and outputting the checked unit startup and shutdown scheme and the load distribution result as decisions. According to the method, an optimization result can be ensured to be accurate through a closed-loop execution and feedback control mechanism, and the automation level, the response speed, the scheduling adaptability and the operation safety margin of startup and shutdown decisions of the hydraulic power plant are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent decision-making, and in particular to a method, system, device and medium for intelligent decision-making of start-stop of a hydro-generator unit based on multi-constraint coupling. BACKGROUND

[0002] As the core equipment of a hydropower station, the start-stop scheduling of a hydro-generator unit is directly related to the safety and stability of a power grid and the efficiency of water energy utilization. With the development of a power system in the direction of increasing proportion of new energy, the power grid continues to improve the requirements for the speed of regulation and response, the peak regulation capacity and the operation reliability of the hydro-generator unit, and the start-stop of the unit has gradually changed from the traditional low-frequency operation to high-frequency, fast and complex operation scenarios. However, the current start-stop decision of the hydropower station mainly relies on the experience of the operating personnel, is easily affected by the differences in manual operation, the limitations of information processing capacity and the complexity of the on-site working conditions, and is difficult to achieve efficient and stable decision-making in scenarios such as multi-unit cooperation, rapid load change, and coexistence of maintenance and planning conflicts. At the same time, the constraint conditions involved in the existing start-stop process are scattered in multiple dimensions such as grid scheduling, equipment health, station topology structure and economic operation, and there is a lack of unified expression method between various constraints, which often needs to be checked manually, and is easy to cause delay in judgment or deviation in understanding. In addition, the traditional system only provides basic reference curves in terms of economic operation, and cannot ensure that the unit is running in the high-efficiency interval for a long time according to the real-time state, nor can it avoid the real-time unfavorable working conditions such as vibration area and limit area. The existing start-stop safety warning function mainly focuses on "over-limit alarm", and cannot grade the risks of different severity and different types, which makes it difficult for the operating personnel to accurately identify potential problems in complex scenarios.

[0003] On the other hand, the information such as maintenance plan, load demand and temporary work ticket of the hydropower station often conflicts, and the existing system lacks the ability of automatic identification and comprehensive processing of plan conflicts, which needs to be compared manually, and the response efficiency is low. In summary, the existing technology has deficiencies in intelligentization, automated decision-making, multi-constraint collaborative processing, economic operation control and early warning grading management, and is difficult to meet the complex needs of the start-stop scheduling of the hydro-generator unit under the background of the new power system. SUMMARY

[0004] In view of the above problems, the present application provides a method, system, device and medium for intelligent decision-making of start-stop of a hydro-generator unit based on multi-constraint coupling.

[0005] Therefore, the problem to be solved by the present application is that the existing technology has deficiencies in intelligentization, automated decision-making, multi-constraint collaborative processing, economic operation control and early warning grading management, and is difficult to meet the complex needs of the start-stop scheduling of the hydro-generator unit under the background of the new power system.

[0006] To solve the above technical problems, the application provides the following technical scheme: a water-turbine-generator-unit start-stop intelligent decision-making method based on multi-constraint coupling, which comprises the following steps: obtaining the structure and operation data of the water-turbine-generator-unit, and performing time alignment and structured processing to construct a unified operation data set; establishing a dynamic representation model of the unit operation state based on the operation data set to generate operation state prediction results of the unit under different start-stop and load distribution scenarios; constructing a multi-dimensional coupling constraint system, and forming a feasible decision space of the unit start-stop combination based on the coupling constraint system; performing multi-objective comprehensive optimization in the feasible decision space to obtain a unit start-stop scheme and load distribution result that meet the coupling constraint system; performing rationality checking on the unit start-stop scheme and load distribution result based on preset warning rules, and generating corresponding warning information according to the checking result; and taking the checked unit start-stop scheme and load distribution result as decision output for the water-turbine-generator-unit to perform control.

[0007] As a preferred scheme of the water-turbine-generator-unit start-stop intelligent decision-making method based on multi-constraint coupling, the construction of the unified operation data set comprises the following steps: obtaining structure and operation original data related to the start-stop of the water-turbine-generator-unit, and classifying and arranging the original data according to the unit identifier, the measuring point type and the time stamp; preprocessing the classified and arranged original data, and performing time alignment on the original data from different data sources based on the time stamp, while associating and mapping the original data with the power station topology structure to obtain an intermediate data set with a unified time reference and spatial topology identifier; based on the intermediate data set, performing structure normalization processing on various data according to a preset data organization rule, normalizing the data records after the normalization processing, and integrating the data with the completed identification into the same operation data resource to form the unified operation data set.

[0008] As a preferred scheme of the water-turbine-generator-unit start-stop intelligent decision-making method based on multi-constraint coupling, the generation mode of the operation state prediction result comprises the following steps: extracting features from the unified operation data set to construct a feature parameter set of the unit operation state; dynamically modeling the operation state of the water-turbine-generator-unit according to the feature parameter set; inputting different start-stop schemes and load distribution scenarios into the dynamic model to obtain operation state prediction results under each scenario.

[0009] As a preferred scheme of the intelligent decision-making method for starting and stopping of a hydroelectric generating set based on multi-constraint coupling, the feasible decision-making space of the starting and stopping combination of the generating set comprises: obtaining multi-dimensional constraint information related to starting and stopping of the hydroelectric generating set; determining available operating states of each hydroelectric generating set and corresponding allowable output ranges according to the multi-dimensional constraint information; and screening the starting and stopping combination of the generating set meeting the multi-dimensional constraint information based on the available operating states and the allowable output ranges of each hydroelectric generating set, and determining a set of the starting and stopping combination of the generating set as the feasible decision-making space.

[0010] The preferred technical scheme has the beneficial effect of realizing early avoidance of potential conflicts, ensuring that the optimization mechanism only operates in a safe, feasible and multi-constraint condition meeting solution space, thereby improving the executability and stability of the decision-making scheme.

[0011] As a preferred scheme of the intelligent decision-making method for starting and stopping of a hydroelectric generating set based on multi-constraint coupling, the obtaining of the starting and stopping scheme of the generating set meeting the coupling constraint system and the load distribution result comprises: combination evaluation of the start-stop combination and the load distribution mode of each generating set based on the feasible decision-making space, to obtain a candidate scheme set meeting the coupling constraint system; comparison and screening of the candidate scheme set according to a preset comprehensive evaluation rule, to determine the starting and stopping scheme of the generating set and the load distribution result meeting the target requirements; and output of the starting and stopping scheme of the generating set and the load distribution result screened as the result meeting the coupling constraint system.

[0012] As a preferred scheme of the intelligent decision-making method for starting and stopping of a hydroelectric generating set based on multi-constraint coupling, the generation of the corresponding early warning information according to the checking result comprises: determination of a warning level by judging an out-of-limit condition of the starting and stopping scheme of the generating set and the load distribution result screened; and generation of early warning information corresponding to the operating state according to the warning level.

[0013] The preferred technical scheme has the beneficial effect of realizing early avoidance of potential conflicts, ensuring that the optimization mechanism only operates in a safe, feasible and multi-constraint condition meeting solution space, thereby improving the executability and stability of the decision-making scheme.

[0014] As a preferred scheme of the intelligent decision-making method for starting and stopping of a hydroelectric generating set based on multi-constraint coupling, the execution control of the hydroelectric generating set comprises: conversion of the starting and stopping scheme and the load distribution result checked into control instructions executable by the generating set, and delivery of the control instructions to corresponding generating set control devices to perform the starting and stopping process.

[0015] To solve the above technical problems, the application provides the following technical scheme: a water-turbine-generator-unit start-stop intelligent decision system based on multi-constraint coupling, comprising a data acquisition module, a prediction module, a screening module and an execution module; the data acquisition module acquires the structure and operation data of the water-turbine-generator-unit, and performs time alignment and structured processing to construct a unified operation data set; the prediction module establishes a dynamic representation model of the unit operation state based on the operation data set to generate operation state prediction results of the unit under different start-stop and load distribution scenarios; the screening module constructs a multi-dimensional coupling constraint system and forms a feasible decision space of the unit start-stop combination based on the coupling constraint system; multi-objective comprehensive optimization is performed in the feasible decision space to obtain a unit start-stop scheme and load distribution result that satisfy the coupling constraint system; the execution module performs rationality checking on the unit start-stop scheme and load distribution result based on a preset early warning rule, and generates corresponding early warning information according to the checking result; the checked unit start-stop scheme and load distribution result are taken as decision output for the water-turbine-generator-unit to perform control.

[0016] A computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above-mentioned water-turbine-generator-unit start-stop intelligent decision method based on multi-constraint coupling when executing the computer program.

[0017] A computer-readable storage medium stores a computer program, and the computer program implements the steps of the above-mentioned water-turbine-generator-unit start-stop intelligent decision method based on multi-constraint coupling when executed by a processor.

[0018] The application has the following beneficial effects: the application constructs unit operation characteristic parameters, dynamic operation models, coupling constraint systems, comprehensive optimization mechanisms, hierarchical early warning rules and execution control links on the basis of unified data, realizes intelligent, standardized and refined management of the whole process of water-turbine-generator-unit start-stop, makes the start-stop scheme meet the requirements of power grid dispatching while taking into account the equipment health condition, topological physical limitation and economic operation requirements, can automatically generate an optimal start-stop combination and load distribution mode that balances safety, economy and operability, and can prompt the operator in advance through a multi-level early warning mechanism when the scheme deviates or has potential risks, so that the operation risk can be identified and intervened; meanwhile, the closed-loop execution and feedback control mechanism ensures that the optimization result can be accurately implemented, greatly improves the automation level, response speed, dispatching adaptability and operation safety margin of the water-turbine-generator-unit start-stop decision of the hydropower plant, and thus improves the water energy utilization efficiency, equipment operation life and safety and economy of the overall system operation. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0020] Figure 1 A flow chart of an intelligent decision-making method for starting and stopping a hydro-generator unit based on multi-constraint coupling in Example 1. DETAILED DESCRIPTION

[0021] In order to make the above-mentioned objects, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0022] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in other manners different from those described herein without departing from the scope of the present application, and those skilled in the art can make similar generalizations without departing from the present application. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0023] Example 1, refer to Figure 1 As the first embodiment of the present application, the embodiment provides an intelligent decision-making method for starting and stopping a hydro-generator unit based on multi-constraint coupling, which comprises the following steps. S1: Obtain the structure and operation data of the hydro-generator unit, and perform time alignment and structured processing to construct a unified operation data set.

[0024] S2: Establish a dynamic representation model of the unit operation state based on the operation data set to generate operation state prediction results of the unit under different starting and stopping and load distribution scenarios.

[0025] S3: Construct a multi-dimensional coupling constraint system, and form a feasible decision space of the unit starting and stopping combination based on the coupling constraint system.

[0026] S4: Perform multi-objective comprehensive optimization in the feasible decision space to obtain a unit starting and stopping scheme and load distribution result that satisfy the coupling constraint system.

[0027] S5: Check the rationality of the unit starting and stopping scheme and load distribution result based on preset warning rules, and generate corresponding warning information according to the checking result.

[0028] S6: Take the checked unit starting and stopping scheme and load distribution result as the decision output for the hydro-generator unit to execute control.

[0029] It should be noted that the existing water turbine generator set start-stop decision mode generally relies on the experience judgment of the operation personnel, lacks systematic and data-based analysis capability, and leads to low response efficiency of start-stop, which is difficult to meet the demand of power grid for rapid peak shaving and high-frequency start-stop; the existing method usually only judges a small number of single-dimensional conditions, cannot cooperatively process multiple constraints such as power grid load constraint, equipment health state, physical topology relationship and economic operation interval, and is easy to cause constraint conflicts to be ignored, thereby bringing safety hazards.

[0030] Therefore, in view of the above problems, as shown in Figure 1 the steps of S1-S6, the multi-source data of the structure information, real-time detection data, power grid dispatching demand and operation plan of the water turbine generator set are centrally integrated, the extraction of the unit operation characteristics and the dynamic state modeling are completed first, and then based on the multi-dimensional coupling system composed of the operation constraints, the power system demand, the equipment health state and the topology limitation, the different start-stop modes and the load distribution schemes are screened one by one to generate an acceptable candidate decision set, on the basis of which the candidate schemes are comprehensively optimized and the operation state is deduced to determine a more optimal start-stop combination and load distribution scheme, then the optimization result is checked combined with the early warning rules and the operation risk level is judged, and finally the checked recommended scheme is issued to the execution system to complete the corresponding unit start-stop and load adjustment, realizing the automation and intelligentization of the operation decision of the hydropower station.

[0031] Embodiment 2, which is different from the first embodiment, is a kind of water turbine generator set start-stop intelligent decision method based on multi-constraint coupling, which further comprises the following steps A1-A3 in step S1 of constructing a unified operation data set: A1: obtaining the structure and operation original data related to the start-stop of the water turbine generator set, and classifying and arranging the original data according to the unit identification, the measuring point type and the time stamp.

[0032] A2: preprocessing the original data after classification and arrangement, and time aligning the original data from different data sources based on the time stamp, and associating and mapping the original data with the power station topology structure to obtain an intermediate data set with unified time reference and spatial topology identification.

[0033] A3: based on the intermediate data set, the data of various types is subjected to structure normalization processing according to the preset data organization rule, the data records after normalization processing are subjected to standardized identification, and the data after identification is integrated and stored in the same operation data resource to form a unified operation data set.

[0034] Specifically, the preprocessing in A2 includes completing the missing data, identifying and removing the abnormal data.

[0035] In the embodiments of the present application, in step A3, the data organization rule adopts a data organization mode based on fieldization rules, including the following steps A311-A313: A311: Identifying the common attribute fields of various types of operation data in the intermediate data set, and mapping the structure parameters, operation monitoring quantities, power grid dispatch quantities and planning information into the preset field set according to their attributes.

[0036] A312: Formatting and converting the data of different sources according to the unified field set, so that the time, unit identification, measurement point type and constraint association information have consistent field structures.

[0037] A313: Standardizing the identification of the formatted data according to the fieldization rules, and uniformly storing various types of records as data structures described by fixed fields, so as to form a unified operation data set.

[0038] In an optional embodiment, the data organization rule can also adopt a data organization mode based on a tag system, including the following steps A321-A323: A321: Establishing a tag system according to the data characteristics in the intermediate data set, and corresponding the operation data, structure data, power grid data and planning data to different tag categories.

[0039] A322: Adding corresponding tags to each data record according to its category, source and association relationship, so that different categories of data can be searched and clustered through the tags to realize unified structural expression.

[0040] A323: Organizing and managing the data records according to the tag system, and integrating the data set with tagged structure into a unified data resource indexed by tags, so as to form a unified operation data set.

[0041] In another optional embodiment, the data organization rule can also adopt a data organization mode based on hierarchical structure rules, including the following steps A331-A333: A331: Structurally dividing the intermediate data set according to the hierarchy of unit-subsystem-measurement point, and determining the belonging position of each type of data in the hierarchical system.

[0042] A332: Hierarchically coding the data according to the hierarchical system, so that the unit-level, subsystem-level and measurement point-level data have upper and lower level association relationship in structure.

[0043] A333: Organizing and storing the hierarchically coded data records according to the hierarchical structure rules, so that the unified operation data set presents a structured expression mode with hierarchical association.

[0044] It should be noted that by time aligning, structuring converting and uniformly storing the unit structure information, operation monitoring quantities, power grid dispatching data and maintenance plans and the like from different data sources, the spatio-temporal consistency of the data and the accuracy of the model input can be improved, the subsequent model construction, state prediction and constraint coupling can be processed on a single data basis, the modeling deviation, inaccurate prediction and decision conflict caused by the data fragmentation in the traditional system can be avoided, and thus the reliability and reproducibility of the entire start-stop intelligent decision-making process can be improved.

[0045] Further, in step S2, the generation mode of the operation state prediction result includes the following steps B1-B3: B1: Feature extraction is performed on the unified operation data set to construct a feature parameter set of the unit operation state.

[0046] B2: The operation state of the hydroelectric generating set is dynamically modeled according to the feature parameter set.

[0047] B3: Different start-stop schemes and load distribution scenarios are input into the dynamic model to obtain operation state prediction results under each scenario.

[0048] In the embodiment of the application, in step B2, the dynamic modeling adopts a dynamic modeling mode based on mechanism relationship, including the following steps B211-B213: B211: Structure classes and physical quantity features required for constructing a mechanism model of the unit operation state are extracted from the unified operation data set, including parameters such as guide vane opening, main shaft vibration amplitude, water head height, tail water pressure, output change rate, and the features are aligned in time sequence to reflect the true response of the unit under different load change conditions on the same time axis, so as to construct a feature parameter set capable of describing the hydraulic, mechanical and electromagnetic coupling behaviors.

[0049] B212: According to the constructed feature parameter set, a mechanism model reflecting the dynamic evolution process of the unit operation state is established by analyzing multiple mechanism laws such as the relationship between water head change and output conversion efficiency, the relationship between vibration amplitude and speed change, and the relationship between guide vane opening and water flow impact force, so that the model can describe the response trend of the unit under different operating conditions through the functional relationship between these physical quantities, such as describing that the vibration value increases approximately linearly with the load and jumps in the critical zone, or that the temperature rise increases step by step with the output change, and the like, so as to obtain a mechanism modeling framework that can be used for predicting dynamic behavior.

[0050] B213: Different start-stop schemes and load distribution scenarios are input into the mechanism model as input conditions. The mechanical and hydrodynamic response processes corresponding to factors such as guide vane opening degree change path, load climbing speed, and target output are simulated to predict possible vibration trend, temperature rise trend, and stability change of the unit, and output the operation state prediction results including "maximum vibration trend value and occurrence time", "whether to enter the limit operation area", "whether there is a risk of hydraulic oscillation", etc.

[0051] In an optional embodiment, dynamic modeling can also use a dynamic modeling method based on historical behavior rules, including the following steps B221-B223: B221: Extract the behavior characteristics of the unit under different historical operating scenarios from the unified operating data set, including the operating stability at each load point, the vibration and temperature rise change pattern under different load climbing speeds, the decay law of monitoring quantities during shutdown stage, etc., to construct a historical behavior template library that can be used for subsequent matching.

[0052] B222: According to the historical behavior template library, the most similar historical operating mode is extracted as a reference by identifying the similarity relationship between the target start-stop scenario and the historical scenario, and the feature trends of multiple historical scenarios are weighted and fused to form an experience-based model that can reflect the dynamic behavior under the new scenario.

[0053] B223: The target start-stop scheme and load distribution scenario are input into the experience-based model to predict the vibration trend, temperature rise rate, efficiency change interval, and possible risk points of the unit during the target load change process by matching the most similar historical behavior template and fusing its dynamic trend, and output the operation state prediction results including "predicted vibration peak value", "possibility of temperature rise exceeding threshold", "whether to enter the vibration critical area", etc.

[0054] In another optional embodiment, dynamic modeling can also use a dynamic modeling method based on multi-source feature fusion, including the following steps B231-B233: B231: Extract structural features (such as unit rated parameters, tailrace tunnel geometric information), operating monitoring features (such as vibration spectrum, pressure fluctuation, temperature rise rate), power grid features (such as dispatching instructions, load trend, frequency deviation), and environmental features (such as water head change, seasonal hydrological conditions) from the unified operating data set.

[0055] B232: A fusion dynamic model is constructed according to multi-source fusion features, dynamic weights are set for different types of features, the influence of water-related features is increased in a high water head environment, the influence of vibration-related features is increased when approaching a limit operation area, and a coupling relationship between features (for example, the influence of tail water pressure change on vibration trend, the influence of power grid load climbing speed on temperature rise change, and the influence of guide vane dynamic opening on vibration interval triggering probability) is expressed, so as to form a model capable of deducing the time evolution law of the unit operation state, and the model can generate state prediction trends in multiple future time periods.

[0056] B233: The target start-stop scheme and load distribution scenario are input into the fusion model, the influence of fusion features on unit vibration, temperature rise, efficiency and stability under different operating conditions is calculated, and the operation state prediction results including "vibration prediction curve in multiple time scales", "temperature rise change law", "migration trend between high efficiency area and limit area" and the like are output, providing sufficient and reliable prediction basis for subsequent constraint coupling and optimization.

[0057] It should be noted that by extracting features based on a unified operating data set to construct a feature parameter set of the unit operating state, and establishing a dynamic model to simulate the operation trend of the unit under different start-stop modes and load distribution scenarios, the changes of key indicators such as vibration, temperature rise and hydraulic response can be predicted in advance, so that the decision-making process no longer relies on artificial experience, thereby the subsequent constraint coupling and optimization link has accurate working condition pre-judgment ability, which helps to avoid the unit entering the vibration critical area or the efficiency decreasing area, and improves the safety and economy.

[0058] Further, in step S3, forming a feasible decision space of unit start-stop combination includes the following steps C1-C3: C1: Obtain multi-dimensional constraint information related to the start-stop of the hydroelectric generating set.

[0059] C2: Determine the available operating state and the corresponding allowable output range of each hydroelectric generating set according to the multi-dimensional constraint information.

[0060] C3: Filter the unit start-stop combination that meets the multi-dimensional constraint information based on the available operating state and the allowable output range of each hydroelectric generating set, and determine the set of unit start-stop combinations as the feasible decision space.

[0061] Specifically, the multi-dimensional constraint information includes: Rigid constraint: rigid execution of load curve and dynamic out-of-limit early warning.

[0062] Equipment health constraint: defect level quantitative evaluation and health degree dynamic derating.

[0063] Topology physical constraint: try to avoid arranging two units of the same tail water main hole to run; when the unit single switch operation, unit DC maintenance, the same tail water main hole three constraints, the unit start priority from high to low in turn for the same tail water main hole, unit DC maintenance, single switch operation.

[0064] Economic optimization constraint: optimal operation interval control of unit and start-up and shut-down loss optimization; and other constraints.

[0065] In the embodiment of the application, in step C3, the screening adopts a feasibility determination mode based on constraint item-by-item checking, including the following steps C311-C313: C311: After the unit operation constraints, power grid dispatching constraints, equipment health constraints and topology structure constraints are summarized, each type of constraint is disassembled into a constraint item that can be directly checked, for example, the power grid output requirement is disassembled into the minimum climbing rate, the maximum time limit and the minimum start-stop number, the equipment health degree constraint is disassembled into the output upper limit or the prohibited operation condition corresponding to the defect level, and the topology constraint is disassembled into the condition that the same tail water hole cannot run at the same high load or the single switch operation unit cannot undertake rapid climbing, so that all constraints are stored in the constraint table in a comparable form.

[0066] C312: For each candidate start-stop mode and load distribution scheme, the constraint table is checked item by item in a predetermined order, which includes checking whether the unit is limited to start due to defects, checking whether the output combination condition of the same tail water hole is violated, checking whether the operation limit under the single switch or DC maintenance state is violated, checking whether the allocated load breaks the boundary of the high efficiency zone or the limit zone, and finally judging whether the scheme completely meets all rigid constraints, while marking the deviation degree under flexible constraints as a basis for feasibility judgment.

[0067] C313: The scheme that passes all rigid constraint checking and has a flexible constraint deviation within the allowable range is marked as a "feasible scheme", the scheme that violates any rigid constraint is marked as an "infeasible scheme", and the scheme with a large flexible deviation is marked as a "weak feasible scheme", thereby forming a feasible decision space including strictly feasible, weak feasible and infeasible classification, so that the subsequent optimization algorithm can select executable start-stop combinations according to the classification.

[0068] In an optional embodiment, the screening can also adopt a feasibility determination mode based on a multi-level constraint priority matrix, including the following steps C321-C323: C321: Construct a priority matrix for each type of constraint from grid dispatch, equipment health status, topological features and economic operation interval, in which rigid constraints are assigned the highest priority, and flexible constraints are assigned lower priorities according to their impact on operation risk, for example, the tailrace constraint is assigned a higher priority than the equipment defect constraint, and the equipment defect constraint is assigned a higher priority than the economic deviation constraint, so that all constraints form a calculable hierarchical structure.

[0069] C322: For each candidate start-stop combination, its corresponding operation state prediction result is matched with the priority matrix level by level, starting from the highest priority rigid constraint, and if any level rigid constraint is violated, the scheme is directly judged as infeasible, and if all rigid constraints are passed, the flexible level is entered for evaluation, in which the influence degree of the scheme on the economic constraint, the vibration risk constraint or the efficiency deviation constraint is measured by calculating the deviation score, and the deviation scores obtained from each level are weighted according to the matrix weight to form the comprehensive feasibility grade of the scheme.

[0070] C323: According to the comprehensive feasibility grade, the schemes are classified, in which the highest grade is the completely feasible scheme, which satisfies all rigid constraints and has the smallest flexible deviation; the middle grade is the acceptable scheme, which satisfies all rigid constraints but has a slightly larger flexible deviation; and the lowest grade is considered infeasible.

[0071] In another optional embodiment, the screening can also use a feasibility judgment method based on a constraint logic tree, including the following steps C331-C333: C331: Convert all constraint relationships related to unit operation into a tree structure, place grid rigid constraints at the root node of the tree, place equipment health constraints, unit state constraints and topological structure constraints at the trunk nodes of the tree, and place economic operation interval and vibration risk flexible constraints at the branch nodes of the tree, so that the dependency relationship of the constraints is expressed in the form of a logic tree.

[0072] C332: When evaluating a candidate start-stop scheme, the scheme is input into the logic tree and traversed from the root node downward, and only when all constraint conditions of a node are satisfied can the next level node be entered, if a condition is not satisfied at a certain level node, the traversal is terminated at this node, which is marked as an infeasible scheme, and the corresponding constraint conflict is recorded according to the type of the node that fails; if the traversal is successfully completed to the leaf node, it means that the scheme satisfies all hard conditions, and then the deviation degree is calculated according to the flexible constraints on the leaf node to judge the grade of the scheme.

[0073] C333: Forming a feasible decision space according to the traversal result, marking the scheme capable of reaching the leaf node and the flexible deviation amount within the acceptable range as a valid feasible scheme, marking the scheme that can only pass through the trunk node but deviates greatly at the flexible node as a standby scheme, marking the scheme blocked at the tree root or trunk as a completely infeasible scheme, and uniformly storing the above multi-class schemes.

[0074] It should be noted that by integrating the grid rigidity constraint, the equipment health constraint, the topology physical constraint and the economic operation interval constraint based on the operation state prediction result, a multi-dimensional coupling constraint system covering the combination of unit start-stop and load distribution mode is constructed, and the feasible decision space of unit start-stop is formed accordingly, so that the intelligent decision no longer produces a scheme that violates the grid demand, equipment capacity or physical structure limit, realizes the early avoidance of potential conflicts, ensures that the optimization mechanism only operates in the solution space that is safe, feasible and meets the multi-constraint conditions, thereby significantly improving the executability and stability of the decision scheme.

[0075] Further, in step S4, obtaining the unit start-stop scheme and load distribution result satisfying the coupling constraint system includes the following steps D1-D3: D1: Based on the feasible decision space, the start-stop combination and load distribution mode of each unit are combined and evaluated to obtain a candidate scheme set satisfying the coupling constraint system.

[0076] D2: The candidate scheme set is compared and screened according to the preset comprehensive evaluation rule to determine the unit start-stop scheme and load distribution result satisfying the target requirement.

[0077] D3: The unit start-stop scheme and load distribution result screened are output as the result satisfying the coupling constraint system.

[0078] In the embodiment of the present application, in step D2, the comprehensive evaluation rule adopts a multi-objective optimization mode driven by the priority of mechanism constraints, including the following steps D211-D213: D211: After forming the candidate scheme set, the system first reads each candidate start-stop combination operation state prediction result item by item, and takes the key indicators in the prediction result such as the main shaft vibration amplitude, the tail water pressure change rate, the temperature rise rate, the guide vane dynamic response characteristic and the speed stability as the evaluation objects. These evaluation objects are classified and weighted according to the safety level standard in actual equipment operation, so that the indicators that are more likely to cause over-limit risk have higher weights, so as to reflect the importance of the safety target in the optimization model.

[0079] D212: The system constructs a comprehensive evaluation function based on the above-mentioned weighted indicators and applies the function to all candidate solutions. The comprehensive performance score of each solution in the safety target, economic target, and response speed target is calculated one by one. The safety indicator score is influenced by the over-limit risk quantification, critical area proximity, and predicted trend change rate. The economic score is determined by whether it remains in the high-efficiency zone, avoids long-term operation in the restricted zone, and the predicted unit water energy utilization rate. The response speed score is determined by the time margin of meeting the dispatching instruction and the output change rate. In the comprehensive evaluation process, the system calculates the comprehensive score of each candidate solution according to the relative size of each weight, and selects the solution with the highest comprehensive score as the optimal solution for the current step.

[0080] D213: The system performs a secondary consistency check on the selected optimal solution, including checking whether it meets the device health constraints, topology physical constraints, and grid rigidity constraints. If the solution does not conflict with all mandatory constraints, it is determined as the output result. If a mandatory constraint is not met, the solution with the second highest score that meets all rigid constraints is found in the candidate set as an alternative output, ensuring that the final output optimization solution meets the multi-objective trade-off result and does not violate any key constraint conditions.

[0081] In an optional implementation, the comprehensive evaluation rule can also use a multi-objective optimization method driven by historical operation experience, including the following steps D221-D223: D221: After reading the candidate solutions, the most similar operation scenarios to the current working condition are retrieved from the historical operation database, and the device operation characteristics and operation risk performance recorded in these historical scenarios are extracted as experience reference templates.

[0082] D222: Compare the predicted operation results of the candidate solutions with the historical experience templates. By evaluating the similarity between the solution and the historical excellent scenarios, an experience score is generated. The experience score, safety score, economic score, and response score are included in the comprehensive optimization calculation. On this basis, all solutions are sorted and the solution that performs best in both experience consistency and predicted safety is selected.

[0083] D223: The selected optimal solution is again checked for abnormal scenarios that have occurred in the historical experience, such as whether there are similar scenarios with sudden vibration spikes, tail water fluctuation amplification, or sudden efficiency decline. If there is no similar historical risk, the solution is taken as the final result. If there is a significant similar dangerous historical scenario, the reliability score of the solution is automatically reduced, and the solution with the second highest score but lower risk is selected as the output.

[0084] In another optional embodiment, the comprehensive evaluation rule can also be based on multi-objective optimization of real-time environment and multi-source characteristics, including the following steps D231-D233: D231: After obtaining the candidate scheme, the current water head, seasonal water level change trend, tail water hole flow characteristics, environmental temperature and other external environmental parameters are read at the same time, and these external conditions are dynamically coupled with the operation prediction curve of the candidate scheme, so that the parameters of each candidate scheme can be locally corrected according to the real-time external environment.

[0085] D232: Based on the prediction curve corrected according to the real-time environment, the safety, economy and adjustability scores of each scheme are recalculated, wherein the safety score is amplified or reduced according to the influence of environmental conditions on some risk indicators, for example, the influence of vibration amplitude change on the model score is stronger under high water head conditions, and the three scores are weighted and synthesized according to the environmental related weight to obtain the comprehensive adaptability score of each scheme.

[0086] D233: According to the comprehensive adaptability score, the optimal scheme is selected and environmental robustness verification is performed, including verifying whether the scheme can still maintain safe operation under conditions such as short-term fluctuation of water head, sudden load adjustment and seasonal tail water fluctuation, if the verification is passed, it is taken as the final output, if the verification is not passed, the scheme with the second highest comprehensive score but stronger environmental robustness is selected from the candidate scheme list.

[0087] It should be noted that by performing multi-objective comprehensive optimization in the feasible decision space to obtain the start-stop combination and load distribution mode that meets the multi-dimensional coupling constraints, the water energy utilization efficiency, unit operation cost, start-stop loss and other optimization objectives can be considered under the premise that the safety boundary and operation constraints are met, so that the generated scheme can balance the safety, economy and scheduling demand satisfaction degree, which can significantly shorten the scheme generation time and improve the scheme quality compared with artificial experience type judgment, so that the whole decision process is more efficient, intelligent and stable.

[0088] Further, in step S5, the corresponding warning information is generated according to the checking result, including the following steps E1-E2: E1: The over-limit condition of the unit start-stop scheme and load distribution result screened out is determined, and the warning level is determined.

[0089] E2: The warning information corresponding to the operating state is generated according to the warning level.

[0090] Specifically, the unit start-stop scheme and its corresponding load distribution results are checked item by item. First, the key operating indicators in the scheme are compared according to the system set rigid safety threshold, including but not limited to the maximum allowable load of the unit, the minimum stable operating load, the load climbing speed that the speed regulation system can withstand, the transient pressure change allowable range of the hydraulic system, and the upper limit value of the main shaft vibration amplitude. It is judged whether the operating conditions in the scheme exceed the corresponding threshold through item-by-item checking. If a certain indicator exceeds the forced limit range, the situation is marked as an "over-limit event", and the occurrence position, affected unit and related operating parameters are recorded. At the same time, according to the over-limit degree, the pre-set warning level rules in the system are matched. For example, if the load distribution makes a certain unit enter the limited operation area, it is marked as a general level warning. If the predicted vibration value may approach the equipment protection action area, it is marked as a higher level warning. If the predicted result shows that a certain indicator significantly exceeds the safety limit and may lead to operating risk, it is marked as the highest level warning, so as to realize the fine division of the warning level according to the over-limit type, severity and duration, etc.

[0091] Further, in step S6, the water turbine generator unit executes control includes converting the checked start-stop scheme and load distribution results into control instructions executable by the unit, and issuing the control instructions to the corresponding unit control device to perform the start-stop process.

[0092] Specifically, the checked unit start-stop scheme and load distribution results are mapped to an executable control instruction set. The corresponding start or stop control flag is generated for each water turbine generator unit that needs to be started or stopped. Combined with the target output value, the load climbing rate and the allowed execution time window, it is converted into a control instruction parameter set including unit start command, stop command, target active output setting value and staged climbing plan. At the same time, each control instruction is labeled with unit number, execution priority, execution condition and timeout protection parameters to ensure that subsequent control execution can accurately identify the specific operating actions of each unit.

[0093] Through the man-machine interface or dispatching communication interface of the plant station automation system, the above control instruction parameter set is issued to the automatic control system or on-site control device of the power station. The control system issues start or stop instructions in a predetermined order, and automatically adjusts the guide vane opening or excitation setting value according to the target output and climbing plan, so that the unit output gradually approaches the target output interval according to the path given by the decision result.

[0094] Embodiment 3, which is different from the previous two embodiments, is a multi-constraint coupling-based intelligent decision-making system for starting and stopping of a hydroelectric generating set, comprising a data acquisition module, a prediction module, a screening module and an execution module; the data acquisition module acquires the structure and operation data of the hydroelectric generating set, and performs time alignment and structured processing to construct a unified operation data set; the prediction module establishes a dynamic representation model of the operating state of the generating set based on the operation data set to generate operating state prediction results of the generating set under different starting and stopping and load distribution scenarios; the screening module constructs a multi-dimensional coupling constraint system, and forms a feasible decision space of the starting and stopping combination of the generating set based on the coupling constraint system; multi-objective comprehensive optimization is performed in the feasible decision space to obtain a starting and stopping scheme of the generating set and a load distribution result that satisfy the coupling constraint system; the execution module performs rationality checking on the starting and stopping scheme of the generating set and the load distribution result based on preset warning rules, and generates corresponding warning information according to the checking result; the checked starting and stopping scheme of the generating set and the load distribution result are taken as decision outputs for execution control of the hydroelectric generating set.

[0095] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of a software product stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk and various program code storage media.

[0096] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered a list of executable instructions for implementing logical functions, and can be specifically embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute the instructions. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport a program for use by or in connection with an instruction execution system, apparatus or device or in conjunction with these instruction execution systems, apparatus or devices.

[0097] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer readable medium can be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example via an optical scanner, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory in a suitable manner.

[0098] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the embodiments described above, various steps or methods can be implemented, for example, in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any forms of hardware, or combinations thereof, of discrete logic circuitry, an application specific integrated circuit (ASIC) with logic gates, a programmable gate array (PGA), a field-programmable gate array (FPGA), and / or the like, can be used.

[0099] It should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the present application. Even though the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalent replaced, without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A method for intelligent decision of starting and stopping of a hydroelectric generating unit based on multi-constraint coupling, characterized in that: The method comprises the following steps: acquiring structural and operating data of the hydroelectric generating set and performing time alignment and structured processing to construct a unified operating data set; establishing a dynamic representation model of the operating state of the generating set based on the operating data set to generate operating state prediction results of the generating set under different start-stop and load distribution scenarios; constructing a multi-dimensional coupling constraint system and forming a feasible decision space of the start-stop combination of the generating set based on the coupling constraint system; performing multi-objective comprehensive optimization in the feasible decision space to obtain a start-stop scheme of the generating set and a load distribution result that satisfy the coupling constraint system; performing rationality checking on the start-stop scheme of the generating set and the load distribution result based on preset warning rules and generating corresponding warning information according to the checking result; outputting the checked start-stop scheme of the generating set and the load distribution result as a decision output for the hydroelectric generating set to perform control.

2. The method of claim 1, wherein the method is characterized in that: The method comprises the following steps: acquiring structural and operating original data related to the start-stop of the hydroelectric generating set, classifying and organizing the original data according to the identification of the generating set, the type of the measuring point and the time stamp; performing preprocessing on the classified and organized original data, time aligning the original data from different data sources based on the time stamp, associating and mapping the original data with the topology structure of the power station to obtain an intermediate data set with a unified time reference and spatial topology identification; based on the intermediate data set, performing structural normalization processing on each type of data according to a preset data organization rule, normalizing the identification of the data records after the normalization processing, and integrating the data with the completed identification into the same operating data resource to form a unified operating data set.

3. The method of claim 2, wherein the method is characterized in that: The method comprises the following steps: performing feature extraction on the unified operating data set to construct a feature parameter set of the operating state of the generating set; dynamically modeling the operating state of the hydroelectric generating set according to the feature parameter set; inputting different start-stop schemes and load distribution scenarios into the dynamic model to obtain operating state prediction results under each scenario.

4. The method of claim 3, wherein the method is characterized in that: The method comprises the following steps: acquiring multi-dimensional constraint information related to the start-stop of the hydroelectric generating set; determining the available operating state and the corresponding allowable output range of each hydroelectric generating set according to the multi-dimensional constraint information; filtering the start-stop combination of the generating set that satisfies the multi-dimensional constraint information based on the available operating state and the allowable output range of each hydroelectric generating set, and determining the set of the start-stop combination of the generating set as the feasible decision space.

5. The method of claim 4, wherein the method is characterized in that: The method comprises the following steps: performing combination evaluation on the start-stop combination and the load distribution mode of each generating set based on the feasible decision space to obtain a candidate scheme set that satisfies the coupling constraint system; comparing and filtering the candidate scheme set according to a preset comprehensive evaluation rule to determine a start-stop scheme of the generating set and a load distribution result that satisfy the target requirements; outputting the start-stop scheme of the generating set and the load distribution result filtered as the result that satisfies the coupling constraint system.

6. The method of claim 5, wherein the method is characterized in that: The method comprises the following steps: performing out-of-limit condition determination on the start-stop scheme of the generating set and the load distribution result filtered to determine the warning level; According to the early warning level, the early warning information corresponding to the running state is generated.

7. The method of claim 6, wherein the method is characterized by: The water-turbine-generator-unit-execution control includes converting the checked start-stop scheme and load distribution result into control instructions executable by the unit, and issuing the control instructions to the corresponding unit control device to perform the start-stop process.

8. A multi-constraint coupling-based intelligent decision-making system for starting and stopping a hydroelectric generator unit, which applies the multi-constraint coupling-based intelligent decision-making method for starting and stopping a hydroelectric generator unit according to any one of claims 1 to 7, characterized in that: The method comprises a data collection module, a prediction module, a screening module and an execution module. The data collection module acquires the structure and operation data of the water-turbine-generator-unit, and performs time alignment and structured processing to construct a unified operation data set. The prediction module establishes a dynamic representation model of the unit operation state based on the operation data set to generate operation state prediction results of the unit under different start-stop and load distribution scenarios. The screening module constructs a multi-dimensional coupling constraint system and forms a feasible decision space of the unit start-stop combination based on the coupling constraint system. In the feasible decision space, multi-objective comprehensive optimization is performed to obtain a unit start-stop scheme and load distribution result that satisfies the coupling constraint system. The execution module checks the rationality of the unit start-stop scheme and load distribution result based on a preset early warning rule, generates corresponding early warning information according to the checking result, and outputs the checked unit start-stop scheme and load distribution result as a decision for water-turbine-generator-unit-execution control. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to implement the steps of the method for intelligent decision-making of water-turbine-generator-unit start-stop based on multi-constraint coupling according to any one of claims 1 to 7.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the method for intelligent decision-making of water-turbine-generator-unit start-stop based on multi-constraint coupling according to any one of claims 1 to 7.

Images