Power generation plan assessment method and system based on space-time consistency regularity

Abstract

The invention discloses a power generation plan assessment method and system based on space-time consistency regularity, and belongs to the technical field of data processing. The method comprises the following steps: carrying out space-time consistency normalization on real-time operation data and power generation plan data of a target power plant to obtain a standardized data sequence; performing correlation entropy blind source separation on the deviation data to obtain a plan error source component and an execution error source component; deducing a dynamic assessment reference interval based on the plan error source component; comparing the standardized data sequence with the reference interval moment by moment to obtain out-of-limit representation; analyzing the dynamic characteristics of the operation state by combining out-of-limit characterization and execution error source components; and performing fusion quantitative scoring on the dynamic features to obtain an assessment result. According to the method, the problems of insufficient data preprocessing, extensive error attribution, fixed reference and single evaluation in the prior art are solved, the scientificity, fairness and comprehensiveness of power generation plan assessment are improved, and a precise direction is provided for operation optimization.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, specifically relating to a method and system for evaluating power generation plans based on spatiotemporal consistency regularization. Background Technology

[0002] In the power generation industry's power generation plan assessment, accurate and fair assessment results are crucial for ensuring the stable operation of the power system and promoting optimized management of power plants. However, existing technologies face numerous technical challenges in practical applications that urgently need to be addressed. First, the problem of insufficient accuracy in data preprocessing is prominent. Existing technologies generally lack spatiotemporal consistency standardization between real-time operational data and power generation plan data. The two types of data often suffer from problems such as asynchronous timestamps, inconsistent data formats (e.g., inconsistent units, differences in precision, and chaotic field identification), and are also prone to outlier interference. These problems directly lead to distorted deviation analysis results, creating hidden dangers for subsequent assessment work and failing to provide a reliable data foundation for assessment.

[0003] Secondly, the attribution analysis of deviation data is rather crude. Existing technologies mostly employ a single attribution method, which can only determine the overall magnitude of deviation data and cannot effectively separate the error sources at the planning and execution levels. Errors at the planning level (such as deviations caused by unreasonable planning) are lumped together with errors at the execution level (such as deviations caused by equipment malfunctions or improper operation), leading to vague definitions of assessment responsibilities and making it difficult to achieve precise accountability. This is not conducive to standardizing the planning process, nor can it effectively urge power plants to improve their execution capabilities.

[0004] Secondly, the assessment benchmarks lack flexibility. Existing assessment methods often use fixed threshold models, failing to consider the dynamic fluctuations in planning errors. In actual operation, the rationality of the plan fluctuates with changes in factors such as grid load and environmental conditions. Fixed thresholds cannot dynamically adjust the assessment boundaries according to these changes, easily leading to a mismatch between the assessment standards and the actual operating scenario—some reasonable execution deviations are over-assessed, while some serious unreasonable deviations are not adequately investigated, resulting in unfair assessment results that are difficult for power plants to accept.

[0005] Finally, the evaluation system lacks comprehensiveness. Existing technologies rely solely on a single limit-exceeding indicator for assessment and scoring, focusing only on whether data exceeds a threshold while ignoring key factors such as the duration of limit exceedances, the degree of cumulative deviation, and operational stability. This fails to comprehensively correlate execution errors with operational status. Consequently, the assessment results fail to accurately reflect the actual operational level of the power plant and cannot provide effective support for subsequent operational optimization, thus hindering the overall improvement of power generation plan execution efficiency.

[0006] These intertwined technical problems have seriously affected the scientific rigor, fairness, and practicality of power generation plan assessments, becoming a key bottleneck restricting the refined management and high-quality development of the power generation industry. Summary of the Invention

[0007] The technical problem to be solved by this invention is to address the shortcomings of the prior art by providing a power generation plan assessment method and system based on spatiotemporal consistency regularization. This method and system solve the technical problems of insufficient data preprocessing accuracy, coarse attribution analysis of deviation data, fixed assessment benchmarks, and a single evaluation system in the existing power generation plan assessment. It achieves reliable power generation plan assessment data, accurate error attribution, dynamic assessment benchmarks, and a comprehensive evaluation system, thereby improving the scientificity, fairness, and comprehensiveness of power generation plan assessment and providing effective support for the definition of assessment responsibilities and subsequent operation optimization.

[0008] The present invention adopts the following technical solution: The power generation plan assessment method based on spatiotemporal consistency regularization includes the following steps: S1. Perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain the standardized data sequence of the target power plant; S2. Perform correlation entropy blind source separation on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data; S3. Based on the planned error source components, perform dynamic confidence interval extrapolation on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval of the target power plant; S4. Compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence; S5. Based on the over-limit characterization and the execution error source components, perform multi-dimensional feature deep analysis on the operating state of the target power plant to obtain the dynamic feature representation of the operating state; S6. Perform fusion quantification scoring on the dynamic feature representation to obtain the assessment results of the target power plant.

[0009] Preferably, in step S1, the real-time operating data and power generation plan data of the target power plant are subjected to spatiotemporal consistency normalization to obtain a standardized data sequence of the target power plant, including: S101. Perform precise timestamp correction on the real-time operation data and power generation plan data of the target power plant to obtain the time synchronization data pair of the target power plant; S102. Standardize the format of the time synchronization data pair to obtain a unified data sequence of the time synchronization data pair; S103. Remove outliers from the unified data sequence to obtain the standardized data sequence of the target power plant.

[0010] Preferably, in step S2, correlation entropy blind source separation is performed on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data, including: S201. Perform time-frequency distribution mapping on the deviation data of the standardized data sequence to obtain the time-frequency energy distribution spectrum of the deviation data; S202. Based on the power generation plan data, perform periodic pattern feature matching on the time-frequency energy distribution map to obtain the dominant mode component of the deviation data; S203. Track the sudden abnormal process of the deviation data to obtain the transient impact event sequence of the deviation data; S204. Based on the dominant mode component and the transient impact event sequence, perform correlation entropy constraint projection separation on the deviation data to obtain the planning error source component and the execution error source component of the deviation data.

[0011] Preferably, step S204 specifically includes: Time-frequency feature analysis is performed on the dominant mode component and the transient impact event sequence to obtain the time-frequency distribution pair of the deviation data; Based on the time-frequency distribution pairs, calculate the projection separation weight of the time-frequency energy distribution map; Based on the projection separation weight, a weighted mapping matching is performed on the time-frequency energy distribution map to obtain the planning error projection component and the execution error projection component of the time-frequency energy distribution map. The planning error projection component and the execution error projection component are subjected to inverse time-frequency transformation to obtain the planning error source component and the execution error source component of the deviation data.

[0012] Preferably, in step S3, based on the planned error source components, a dynamic confidence interval is extrapolated for the assessment benchmark value of the target power plant to obtain the assessment benchmark interval of the target power plant, including: S301. Reconstruct the kernel density distribution of the planning error source components to obtain the joint probability density surface of the planning error source components; S302. Based on the joint probability density surface, perform contour cluster adaptive extraction on the planning error source components to obtain the probability contour cluster set of the planning error source components; S303. Map the probability equal-height clusters to a preset assessment tolerance risk level to obtain the selected probability equal-height clusters of the planning error source components; S304. Based on the envelope boundary of the selected probability equal-high cluster, the assessment benchmark value of the target power plant is dynamically divided to obtain the assessment benchmark interval of the target power plant.

[0013] Preferably, step S302 specifically includes: A topological invariant analysis is performed on the joint probability density surface to obtain the Betti number sequence of the joint probability density surface; Based on the Betti number sequence, a topological constraint field is constructed on the joint probability density surface to obtain the dynamic decision potential field of the joint probability density surface. The potential energy function of the joint probability density surface is obtained by performing a negative logarithmic transformation on the probability density values ​​on the joint probability density surface. The variance and kurtosis statistical characteristics of the planning error source components are normalized to obtain the equivalent quality parameters of the planning error source components. Based on the dynamic decision potential field, the barrier penetration capability of the probability distribution of the planning error source components is evaluated to obtain the tunneling intensity between two points in the joint probability density surface. Based on the tunneling intensity, the probability mass of the joint probability density surface is redistributed to obtain the enhanced probability density distribution of the joint probability density surface. By contour closure of the enhanced probability density distribution, the probability equal-height clusters of the planning error source components are obtained.

[0014] Preferably, in step S4, the standardized data sequence is compared with the benchmark interval at each time step to obtain the limit-crossing characteristic of the standardized data sequence, including: S401. Based on the assessment benchmark interval, the standardized data sequence is subjected to position difference quantization to obtain the boundary distance vector of the standardized data sequence; S402. Based on the boundary distance vector, perform a preliminary determination of the limit-crossing state of the standardized data sequence to obtain the basic limit-crossing identifier of the standardized data sequence; S403. Evaluate the out-of-limit evolution trend of the standardized data sequence to obtain the out-of-limit persistence index sequence of the standardized data sequence; S404. Based on the basic limit violation identifier and the limit violation persistence index sequence, perform joint limit violation pattern identification on the standardized data sequence to obtain the limit violation characterization of the standardized data sequence.

[0015] Preferably, in step S5, based on the limit violation characterization and the execution error source components, a multi-dimensional feature depth analysis is performed on the operating state of the target power plant to obtain a dynamic feature representation of the operating state, including: S501. The spatiotemporal causal relationship between the out-of-limit representation and the execution error source component is quantified to obtain the causal correlation matrix of the standardized data sequence. S502. Based on the causal correlation matrix, the core fluctuation pattern of the operating state of the target power plant is mined to obtain the dominant oscillation mode of the operating state. S503. Based on the time-varying trajectory of the dominant oscillation mode, the stability margin of the operating state is dynamically tracked to obtain the stability margin curve of the operating state. S504. Perform multi-source feature fusion on the causal correlation matrix, the dominant oscillation mode parameters, and the stability margin curve to obtain a dynamic feature representation of the operating state.

[0016] Preferably, in step S6, the dynamic feature representation is fused and quantified to obtain the assessment result of the target power plant, including: S601. Based on the historical data of the target power plant, perform multi-dimensional importance determination on the dynamic feature representation to obtain the weight allocation of the dynamic feature representation; S602. Based on the weight allocation, perform feature weighting and aggregation on the dynamic feature representation to obtain the comprehensive performance index of the target power plant; S603. Based on the comprehensive performance indicators, the target power plant is scored in a standardized manner to obtain the assessment results of the target power plant.

[0017] Secondly, embodiments of the present invention provide a power generation plan assessment system based on spatiotemporal consistency regularization, comprising: The data module is used to perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain a standardized data sequence of the target power plant. The analysis module is used to receive the standardized data sequence output by the data module, perform correlation entropy blind source separation on the deviation data of the standardized data sequence, and obtain the planning error source component and the execution error source component of the deviation data. The deduction module is used to receive the planning error source component output by the analysis module, and perform dynamic confidence interval deduction on the assessment benchmark value of the target power plant based on the planning error source component to obtain the assessment benchmark interval of the target power plant. The identification module is used to receive the standardized data sequence output by the data module and the assessment benchmark interval output by the inference module, respectively, and compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence. The parsing module is used to receive the limit violation characterization output by the identification module and the execution error source component output by the analysis module, respectively, and perform multi-dimensional feature deep analysis on the operating state of the target power plant based on the limit violation characterization and the execution error source component to obtain the dynamic feature representation of the operating state. The scoring module is used to receive the dynamic feature representation output by the parsing module, perform fusion quantification scoring on the dynamic feature representation, and obtain the assessment result of the target power plant.

[0018] Thirdly, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described power generation plan assessment method based on spatiotemporal consistency regularization.

[0019] Fourthly, embodiments of the present invention provide a computer-readable storage medium including a computer program, which, when executed by a processor, implements the steps of the above-described power generation plan assessment method based on spatiotemporal consistency regularization.

[0020] Fifthly, a chip includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described power generation plan assessment method based on spatiotemporal consistency regularization.

[0021] In a sixth aspect, embodiments of the present invention provide an electronic device, including a computer program, which, when executed by the electronic device, implements the steps of the above-described power generation plan assessment method based on spatiotemporal consistency regularization.

[0022] Compared with the prior art, the present invention has at least the following beneficial effects: The power generation plan assessment method based on spatiotemporal consistency regularization addresses input quality issues through data regularization, enables accountability through error separation, ensures the rationality of assessment standards through "benchmark deduction," conducts in-depth evaluation through limit violation identification and state analysis, and finally integrates the scoring output results. This closed-loop logic connects key technical links, with the output of each step directly serving as the input for subsequent steps, ensuring the coherence and integrity of the technical solution. By integrating core technologies such as spatiotemporal consistency regularization and relevant entropy blind source separation, the method improves overall assessment efficiency and accuracy, providing a basic framework for subsequent refinement of technical features and guaranteeing the rationality and completeness of the invention's scope of protection.

[0023] Furthermore, precise timestamp correction achieves accurate time matching between real-time operational data and power generation plan data, avoiding distortion in analysis caused by time asynchrony; format standardization eliminates differences in units, precision, field identifiers, etc., ensuring data format uniformity; outlier removal is based on historical data to define a reasonable range and remove interfering data that exceeds the range. These synergistic effects ensure that the final output standardized data sequence possesses the characteristics of time alignment, uniform format, and data reliability, completely solving the problem of insufficient precision in existing data preprocessing technologies, and providing high-quality data support for subsequent error separation, assessment, and scoring processes.

[0024] Furthermore, time-frequency distribution mapping transforms the deviation data into a two-dimensional time-frequency energy distribution map, intuitively presenting the time and frequency characteristics of the data; periodic pattern feature matching extracts the dominant pattern component that matches the periodic pattern of the power generation plan, corresponding to the error characteristics at the planning level; sudden anomaly process tracking captures transient impact events in the deviation data, corresponding to sudden errors at the execution level; finally, preliminary separation of the two types of components is achieved through relevant entropy constraint projection separation. Accurate extraction of key feature components from the deviation data provides a clear basis for subsequent accurate separation of planning error sources and execution error sources, solving the problems of existing technologies' coarse deviation attribution and inability to distinguish between planning and execution errors.

[0025] Furthermore, time-frequency distribution pairs are obtained through time-frequency feature analysis. Based on these distribution pairs, projection separation weights are calculated to quantify the correlation between the time-frequency energy distribution map and the dominant mode components and transient impact event sequences. Then, through weighted mapping matching and inverse time-frequency transformation, the precise separation of planning error source components and execution error source components is achieved. By replacing empirical judgment with quantitative calculations and dynamically adapting weight allocation to data characteristics, the accuracy and reliability of error source separation are ensured. This provides precise technical support for defining assessment responsibilities and solves the problems of ambiguous error attribution and difficulty in accurately assigning responsibility in existing technologies.

[0026] Furthermore, the kernel density distribution reconstruction yields the joint probability density surface of the planning error source components, intuitively presenting the distribution characteristics of the planning error; contour clusters adaptively extract and capture the core region of the probability distribution; by mapping a preset allowable risk level, selected probability contour clusters that meet the actual assessment requirements are selected; finally, the assessment benchmark interval is divided based on its envelope boundary. This makes the assessment benchmark no longer a fixed threshold, but dynamically adjusted according to the fluctuation characteristics of the planning error, significantly improving the adaptability of the assessment benchmark to the actual operating scenario and solving the problems of mismatch between existing technical assessment standards and actual scenarios, and unfair assessment results.

[0027] Furthermore, topological invariant analysis and constraint field construction clarified the structural characteristics of the joint probability density surface; the potential energy function and equivalent mass parameters provided a basis for assessing the barrier penetration capability; the tunneling intensity calculation formula quantified the probabilistic penetration capability between two points on the surface, and based on this intensity, probability mass redistribution was performed to make the probability distribution more closely match actual characteristics; finally, probabilistic equal-height clusters were obtained through contour closure. Supported by topological analysis and related theories of quantum mechanics, the adaptive capability and accuracy of equal-height cluster extraction were improved, ensuring the scientific validity of subsequent risk level mapping and assessment benchmark interval division, and providing core technical support for the rationality of dynamic benchmarks.

[0028] Furthermore, the location difference quantification yields a boundary distance vector, accurately reflecting the distance relationship between the data and the assessment benchmark interval; the basic over-limit identifier clarifies the over-limit type at each moment; the over-limit persistence index sequence integrates the over-limit duration and cumulative deviation degree; and joint identification combines both to identify specific patterns such as short-term single over-limit and long-term continuous over-limit. This comprehensive presentation of over-limit types, durations, deviation degrees, and pattern characteristics provides detailed over-limit information for subsequent operational status analysis, ensuring a more comprehensive and accurate assessment of the power plant's operational status.

[0029] Furthermore, the causal correlation matrix obtained by spatiotemporal causal correlation quantification clarifies the degree of correlation between the out-of-limit representation and the source components of execution error; the dominant oscillation mode is extracted to reveal the core fluctuation patterns of the operating state; the stability margin curve dynamically tracks the stability of the operating state; and the multi-source feature fusion integrates three types of key features and eliminates redundant information. By comprehensively analyzing the operating state from three dimensions—correlation features, fluctuation patterns, and stability—the actual operating level of the power plant is accurately characterized, providing comprehensive and in-depth operating state information for subsequent integrated quantitative scoring, ensuring that the assessment results truly reflect the power plant's operating capacity.

[0030] Furthermore, the multi-dimensional importance determination based on historical data ensures that the weight allocation aligns with the actual operation of power plants, strengthening key features. Feature weighting and aggregation integrates key information from dynamic feature representations to generate comprehensive performance indicators reflecting overall operational levels. Standardized scoring defines intervals based on historical best and worst values, ensuring the comparability of assessment results. By relying on historical data, the scoring process is made objective and standardized, avoiding subjective interference and addressing the issues of incomparability and insufficient fairness in existing technologies. The final assessment results accurately evaluate the execution of power generation plans and provide a basis for comparing assessment results from different power plants and at different times.

[0031] It is understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0032] In summary, the method of this invention ensures data quality through precise data preprocessing; clarifies responsibility boundaries through precise error attribution; and ensures fair and comprehensive assessment through dynamic assessment benchmarks and a multi-dimensional evaluation system. This solves the problems of inaccurate data, unclear attribution, rigid standards, and one-sided evaluation in existing power generation plan assessments.

[0033] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0034] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the system of the present invention; Figure 3 A schematic diagram of a computer device provided in an embodiment of the present invention; Figure 4 This is a block diagram of a chip provided according to an embodiment of the present invention.

[0035] Among them, 60. Computer equipment; 61. Processor; 62. Memory; 63. Computer program; 600. Electronic device; 610. Processing unit; 620. Storage unit; 6201. Random access memory unit; 6202. Cache memory unit; 6203. Read-only memory unit; 6204. Program / utility; 6205. Program module; 630. Bus; 640. Display unit; 650. Input / output interface; 660. Network adapter; 700. External device. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0038] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0039] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0040] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0041] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0042] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0043] This invention provides a power generation plan assessment method based on spatiotemporal consistency regularization. It involves performing spatiotemporal consistency regularization on the real-time operating data and power generation plan data of the target power plant to obtain a standardized data sequence; performing correlation entropy blind source separation on the deviation data of the standardized data sequence to obtain the planning error source component and execution error source component of the deviation data; performing dynamic confidence interval extrapolation on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval of the target power plant; comparing the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence; performing multi-dimensional feature deep analysis on the operating status of the target power plant to obtain the dynamic feature representation of the operating status; and performing fusion quantitative scoring on the dynamic feature representation to obtain the assessment result of the target power plant. This invention can improve the efficiency of power generation plan assessment based on spatiotemporal consistency regularization.

[0044] Please see Figure 1 This invention provides a power generation plan assessment method based on spatiotemporal consistency regularization, comprising the following steps: S1. Perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain the standardized data sequence of the target power plant; In this embodiment of the invention, the step of performing spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain a standardized data sequence of the target power plant includes: The real-time operation data and power generation plan data of the target power plant are precisely corrected using timestamps to obtain the time synchronization data pair of the target power plant; The time synchronization data pairs are format-standardized to obtain a unified data sequence for the time synchronization data pairs; Outliers in the unified data sequence are removed to obtain a standardized data sequence for the target power plant.

[0045] When performing precise timestamp correction on the real-time operation data and power generation plan data of the target power plant, the original timestamp information carried by each type of data is first extracted. Then, the industry-standard time is selected as the time reference. The original timestamps of the real-time operation data and the power generation plan data are calibrated and adjusted against the standard time so that each data point in the real-time operation data can be accurately matched with the corresponding time point in the power generation plan data, and finally a one-to-one time synchronization data pair is formed.

[0046] When standardizing the format of time synchronization data pairs, first define the pre-defined unified format specification, which covers core elements such as data units, data precision, and field identifiers. Then, process each set of data in the time synchronization data pair according to the specification, convert power data in different units to a unified unit, adjust data with different precision to a consistent precision, and replace field identifiers with different expressions with standard identifiers to ensure that the format of all data in the time synchronization data pair is completely unified, thereby obtaining a unified data sequence.

[0047] When removing outliers from the unified data sequence, the reasonable value range of each type of data is first determined based on the historical data of the target power plant under normal operating conditions. Then, each data in the unified data sequence is checked one by one to determine whether each data is within the corresponding reasonable value range. If the data exceeds the reasonable value range, it is determined to be an outlier. All data determined to be outliers are removed from the unified data sequence. The remaining data that meet the reasonable value range constitute the standardized data sequence of the target power plant.

[0048] The beneficial effects are that precise time matching between real-time operation data and power generation plan data is achieved through timestamp-based accurate correction, forming time-synchronized data pairs with clear correspondence. Then, the differences in units, precision, and field identifiers are eliminated through format standardization, resulting in a unified data sequence with a uniform format. Finally, by removing abnormal data that exceeds the reasonable value range, the final standardized data sequence is guaranteed to have the characteristics of time alignment, uniform format, and data reliability, providing high-quality data support for subsequent deviation analysis and assessment.

[0049] S2. Perform correlation entropy blind source separation on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data; In this embodiment of the invention, the step of performing correlation entropy blind source separation on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data includes: The time-frequency distribution mapping of the deviation data of the standardized data sequence is performed to obtain the time-frequency energy distribution spectrum of the deviation data; Based on the power generation plan data, periodic pattern feature matching is performed on the time-frequency energy distribution map to obtain the dominant mode component of the deviation data; The sudden anomaly process of the deviation data is tracked to obtain the transient impact event sequence of the deviation data; Based on the dominant mode component and the transient impact event sequence, the deviation data is subjected to correlation entropy constraint projection separation to obtain the planning error source component and the execution error source component of the deviation data.

[0050] The step of performing correlation entropy-constrained projection separation on the deviation data based on the dominant mode component and the transient impact event sequence to obtain the planning error source component and execution error source component of the deviation data includes: Time-frequency feature analysis is performed on the dominant mode component and the transient impact event sequence to obtain the time-frequency distribution pair of the deviation data; Based on the time-frequency distribution pairs, the projection separation weight of the time-frequency energy distribution map is calculated, wherein the calculation formula for the projection separation weight is: ; In the formula, The time-frequency energy distribution spectrum in time and frequency Projection separation weights at the location, It is a natural exponential function. The time-frequency energy distribution spectrum in time and frequency The energy amplitude at that location, For the dominant mode component in the time-frequency distribution pair and frequency Time-frequency distribution at location, For the time-frequency distribution pair of transient impact event sequences in time and frequency Time-frequency distribution at location, The kernel width parameter is a preset value for the fluctuation stability based on the dominant mode component. The kernel width parameter is preset based on the sudden intensity of the transient impact event sequence. This is the Euclidean distance operator; Based on the projection separation weight, a weighted mapping matching is performed on the time-frequency energy distribution map to obtain the planning error projection component and the execution error projection component of the time-frequency energy distribution map. The planning error projection component and the execution error projection component are subjected to inverse time-frequency transformation to obtain the planning error source component and the execution error source component of the deviation data.

[0051] When performing time-frequency distribution mapping on the deviation data of the standardized data sequence, the characteristics of the deviation data changing with time and the fluctuation characteristics at different frequencies are extracted. The time-varying characteristics are correlated with the frequency fluctuation characteristics to construct a two-dimensional distribution pattern of the deviation data. This two-dimensional distribution pattern is the time-frequency energy distribution spectrum of the deviation data.

[0052] When performing periodic pattern feature matching on the time-frequency energy distribution map based on the power generation plan data, the inherent periodic variation pattern in the power generation plan data is extracted. This periodic variation pattern is then compared one by one with each feature region in the time-frequency energy distribution map. The feature part in the map that has the highest degree of agreement with the periodic variation pattern is selected, and this feature part is the dominant mode component of the deviation data.

[0053] When tracking sudden abnormal processes in the deviation data, the real-time change trend of the deviation data is continuously monitored, and sudden fluctuation nodes that deviate from the normal fluctuation range are identified. The occurrence time and fluctuation amplitude information of these sudden fluctuation nodes are sorted out in chronological order to form a transient impact event sequence of the deviation data.

[0054] When performing time-frequency feature analysis on the dominant mode component and the transient impact event sequence, the time variation features and frequency distribution features of the dominant mode component are extracted respectively, and the time variation features and frequency distribution features of the transient impact event sequence are extracted at the same time. The time-frequency features of the dominant mode component and the time-frequency features of the transient impact event sequence are paired and combined to obtain the time-frequency distribution pair of the deviation data.

[0055] When determining the projection separation weight of the time-frequency energy distribution map based on the time-frequency distribution pair, the characteristic differences between the two components in the time-frequency distribution pair are referenced to clarify the proportion of planning error and execution error corresponding to different regions in the time-frequency energy distribution map. This proportion is the projection separation weight.

[0056] When performing weighted mapping matching on the time-frequency energy distribution map based on the projection separation weight, the time-frequency energy distribution map is divided into two independent feature regions according to the proportion relationship determined by the projection separation weight. The feature region corresponding to the proportion of planning error is the planning error projection component of the time-frequency energy distribution map, and the feature region corresponding to the proportion of execution error is the execution error projection component of the time-frequency energy distribution map.

[0057] When performing inverse time-frequency transformation on the projected components of the planning error and the projected components of the execution error, the projected components of the planning error and the projected components of the execution error are restored from the two-dimensional time-frequency distribution to the deviation data form in the time dimension. The deviation data form after restoration of the projected components of the planning error is the projected error source component of the deviation data, and the deviation data form after restoration of the projected components of the execution error is the execution error source component of the deviation data.

[0058] The energy amplitude at the corresponding time and frequency is obtained by mapping the time and frequency distribution of the deviation data to obtain the time and frequency energy distribution spectrum, and then directly extracting the energy amplitude at the corresponding time and frequency position of the spectrum. The time-frequency distribution at the corresponding time and frequency is obtained by matching the periodic mode features of the time-frequency energy distribution map based on the power generation plan data, obtaining the dominant mode component, and then performing time-frequency feature analysis on the dominant mode component to extract the time-frequency distribution at the corresponding time and frequency position. The time-frequency distribution at corresponding times and frequencies is obtained by tracking transient impact event sequences from sudden anomaly processes in deviation data, and then extracting the time-frequency distribution at corresponding time and frequency positions after analyzing the time-frequency features of the sequence. The kernel width parameter preset based on the fluctuation stability of the dominant mode component is obtained by statistically analyzing the fluctuation range and frequency of the dominant mode component in historical data, and directly setting the corresponding kernel width parameter based on these fluctuation stability characteristics. The kernel width parameter preset based on the suddenness intensity of the transient impact event sequence is obtained by statistically analyzing the fluctuation amplitude and duration of each event in the transient impact event sequence, and directly setting the corresponding kernel width parameter based on these suddenness intensity characteristics.

[0059] This content is used to determine the projection separation weight of the time-frequency energy distribution spectrum at the corresponding time and frequency positions. Specifically, it is used to allocate the weight ratio of the planning error and execution error corresponding to that position by comparing the correlation between the energy amplitude of the time-frequency energy distribution spectrum and the time-frequency distribution of the dominant mode component and the transient impact event sequence.

[0060] when and The higher the correlation, the larger this weight will be, meaning that the energy at that time and frequency location corresponds more to the projected components of the planning error; when and The higher the correlation, the smaller the value of this weight, which means that the energy at that time and frequency position corresponds more to the projected component of the execution error.

[0061] The beneficial effects are that by performing a series of operations on the deviation data, such as time-frequency distribution mapping, periodic pattern feature matching, and sudden abnormal process tracking, the dominant pattern components and transient impact event sequences of the deviation data are extracted. Then, the projection separation weights are determined through time-frequency feature analysis, and the weighted mapping matching and inverse time-frequency transformation are completed. This achieves accurate separation of the planning error source components and the execution error source components in the deviation data, providing clear and reliable error attribution support for subsequent benchmark interval extrapolation and in-depth analysis of multi-dimensional features of the operating status.

[0062] By extracting parameters from the products of the corresponding steps or setting relevant content based on features, the basis for calculating the projection separation weight is ensured to be clear and in line with the actual scenario. This content can accurately determine the projection separation weight of the corresponding position in the time-frequency energy distribution map and reasonably allocate the weight ratio of the planning error and execution error corresponding to that position. At the same time, the weight size is dynamically adjusted according to the different degrees of correlation, so that the energy of the time-frequency energy distribution map can accurately correspond to the corresponding error projection component, providing reliable weight support for the subsequent accurate separation of the planning error source component and the execution error source component of the deviation data.

[0063] S3. Based on the planned error source components, perform dynamic confidence interval extrapolation on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval of the target power plant; In this embodiment of the invention, the step of dynamically extrapolating the assessment benchmark value of the target power plant based on the planned error source components to obtain the assessment benchmark interval of the target power plant includes: The kernel density distribution of the planning error source components is reconstructed to obtain the joint probability density surface of the planning error source components; Based on the joint probability density surface, the planning error source components are subjected to contour cluster adaptive extraction to obtain the probability contour cluster set of the planning error source components; The probability equal-height clusters are mapped to a preset assessment tolerance risk level to obtain the selected probability equal-height clusters of the planning error source components; Based on the envelope boundary of the selected probability equal-high cluster, the assessment benchmark value of the target power plant is dynamically boundary-divided to obtain the assessment benchmark interval of the target power plant.

[0064] The step of adaptively extracting the planning error source components based on the joint probability density surface to obtain the probability equal-height cluster set of the planning error source components includes: A topological invariant analysis is performed on the joint probability density surface to obtain the Betti number sequence of the joint probability density surface; Based on the Betti number sequence, a topological constraint field is constructed on the joint probability density surface to obtain the dynamic decision potential field of the joint probability density surface. The potential energy function of the joint probability density surface is obtained by performing a negative logarithmic transformation on the probability density values ​​on the joint probability density surface. The variance and kurtosis statistical characteristics of the planning error source components are normalized to obtain the equivalent quality parameters of the planning error source components. Based on the dynamic decision potential field, the barrier penetration capability is evaluated on the probability distribution of the planning error source components to obtain the tunneling strength between two points in the joint probability density surface. The formula for calculating the tunneling strength is: ; In the formula, The tunneling strength is... It is a natural exponential function. To reduce Planck's constant, The tunneling initiation point on the joint probability density surface, The tunneling termination point on the joint probability density surface. The equivalent quality parameter is... Let be the potential energy function. The reference energy level is preset based on the energy distribution in the dynamic decision potential field. For the joint probability density surface at point The change in probability density at a given location. Operators for Euclidean norms; Based on the tunneling intensity, the probability mass of the joint probability density surface is redistributed to obtain the enhanced probability density distribution of the joint probability density surface. By contour closure of the enhanced probability density distribution, the probability equal-height clusters of the planning error source components are obtained.

[0065] When reconstructing the kernel density distribution of the planned error source components, all data points of the planned error source components are sorted out. Based on the distribution of these data points, a surface shape that can reflect the overall density of the data distribution is gradually fitted. This surface shape is the joint probability density surface of the planned error source components.

[0066] When performing topological invariant analysis on the joint probability density surface, the number of mutually independent connected regions and the number of holes in the region are identified. These numbers are arranged in a specific order to form a sequence, which is the Betti number sequence of the joint probability density surface.

[0067] When constructing a topological constraint field for the joint probability density surface based on the Betti number sequence, the constraint rules for different regions within the surface are defined based on the connected regions and hole characteristics reflected by the Betti number sequence, forming a field that can guide the direction of subsequent analysis. This field is the dynamic decision potential field of the joint probability density surface.

[0068] When performing a negative logarithmic transformation on the probability density values ​​on the joint probability density surface, the probability density values ​​at each position on the surface are selected, and these values ​​are transformed through negative logarithmic operations. The transformed results are combined to form a function, which is the potential energy function of the joint probability density surface.

[0069] When normalizing the variance and kurtosis statistical characteristics of the planning error source components, the variance and kurtosis values ​​of the planning error source components are calculated, and these values ​​are adjusted to the same dimension range. The adjusted comprehensive value is the equivalent quality parameter of the planning error source components.

[0070] When evaluating the barrier penetration capability of the probability distribution of the planning error source components based on the dynamic decision potential field, the ability of the probability distribution between two points on the joint probability density surface to break through the constraints is analyzed by referring to the constraint rules of the dynamic decision potential field. The quantification result of this capability is the tunneling strength between the two points on the joint probability density surface.

[0071] When redistributing probability mass on the joint probability density surface based on the tunneling intensity, the probability density values ​​of different regions on the surface are adjusted according to the magnitude of the tunneling intensity to make the probability distribution more closely match the actual characteristics. The adjusted probability distribution is the enhanced probability density distribution of the joint probability density surface.

[0072] When closing the contour of the enhanced probability density distribution, points with the same value in the enhanced probability density distribution are identified, and these points are connected to form a closed contour region. Multiple such contour regions are combined to form the probability equal-height cluster of the planning error source components.

[0073] When mapping the probability contour clusters to a preset assessment risk level, the probability range corresponding to the preset assessment risk level is defined, and clusters matching this range are selected from the probability contour clusters. The selected clusters are the selected probability contour clusters of the planning error source components.

[0074] When dynamically dividing the assessment benchmark value of the target power plant based on the envelope boundary of the selected probability equal-high cluster, the numerical range corresponding to the envelope boundary of the selected probability equal-high cluster is determined, and this range is used as the upper and lower boundaries of the assessment benchmark value. The interval defined by the upper and lower boundaries is the assessment benchmark interval of the target power plant.

[0075] The reduced Planck constant is directly adopted as a general fixed physical constant value. The tunneling start point and tunneling end point are the evaluation start and end points selected from the joint probability density surface, respectively. The equivalent quality parameter is the result obtained after dimensional normalization of the variance and kurtosis statistical characteristics of the planning error source components, and this result can be directly extracted. The potential energy function is the function obtained by performing a negative logarithmic transformation on the probability density values ​​on the joint probability density surface, and this function is directly used. The reference energy level is the preset energy level value based on the energy distribution in the dynamic decision potential field. The probability density change is the result obtained by analyzing the probability density change at the corresponding point on the joint probability density surface.

[0076] This content is used to evaluate the barrier penetration capability between two points on the joint probability density surface. Specifically, it combines the relevant characteristics of the dynamic decision potential field with the distribution of the joint probability density surface to quantify the ability of the probability distribution between the two points to break through the constraints. The quantification result is the tunneling strength.

[0077] The smaller the difference between the potential energy function and the reference energy level, the greater the tunneling intensity and the stronger the barrier penetration ability between the two points; the smaller the equivalent mass parameter, the greater the tunneling intensity and the stronger the barrier penetration ability between the two points; the larger the Euclidean norm of the probability density change, the greater the tunneling intensity, the more obvious the change in probability distribution between the two points, and the stronger the barrier penetration ability.

[0078] The beneficial effect is that by carrying out a series of operations such as kernel density distribution reconstruction, topological invariant analysis, and topological constraint field construction on the planning error source components, the probability contour clusters of the planning error source components are extracted. Then, the selected probability contour clusters are determined by mapping the allowable risk level. Based on their envelope boundaries, the dynamic boundary division of the assessment benchmark value is completed. Finally, an assessment benchmark interval that fits the actual distribution characteristics of the planning error is obtained, providing a scientific and reasonable assessment basis for subsequent time-by-time variable comparison and multi-dimensional feature analysis of the operating status.

[0079] By obtaining the required information from the general constant values ​​and preset values ​​of the corresponding step products, the fit and reliability of the tunneling strength assessment basis are ensured. This information can accurately quantify the barrier penetration capability between two points on the joint probability density surface, thus obtaining a clear tunneling strength. At the same time, by observing the changes in tunneling strength values ​​corresponding to changes in different related factors, the differences in the strength of barrier penetration capability between the two points can be clearly reflected, providing accurate quantitative support for the subsequent probability mass redistribution of the joint probability density surface.

[0080] S4. Compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence; In this embodiment of the invention, the step of comparing the standardized data sequence with the benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence includes: Based on the assessment benchmark interval, the positional difference of the standardized data sequence is quantized to obtain the boundary distance vector of the standardized data sequence; Based on the boundary distance vector, a preliminary determination of the limit-crossing state is made on the standardized data sequence to obtain the basic limit-crossing identifier of the standardized data sequence; The standardized data sequence is evaluated for its tendency to exceed limits, resulting in a limit exceedance persistence index sequence. Based on the basic limit violation identifier and the limit violation persistence index sequence, the limit violation pattern of the standardized data sequence is jointly identified to obtain the limit violation characterization of the standardized data sequence.

[0081] Based on the assessment benchmark interval, when quantifying the positional differences of the standardized data sequence, the values ​​in the standardized data sequence are extracted time by time. The distance between the value at each time and the upper and lower boundary values ​​of the assessment benchmark interval is calculated. If the value at that time is within the assessment benchmark interval, the distance is recorded as zero. If the value is higher than the upper boundary, the difference between the value and the upper boundary is calculated. If the value is lower than the lower boundary, the difference between the value and the lower boundary is calculated. The distances corresponding to all times are arranged in chronological order, and the resulting ordered distance set is the boundary distance vector of the standardized data sequence.

[0082] When making a preliminary determination of the limit violation state of the standardized data sequence based on the boundary distance vector, the corresponding distance value in the boundary distance vector is checked at each time step. If the distance value is zero, the standardized data sequence at that time step is determined to be in a non-limit violation state. If the distance value is higher than the upper boundary of the assessment benchmark interval, it is determined to be an upper limit violation state. If the distance value is lower than the lower boundary of the assessment benchmark interval, it is determined to be a lower limit violation state. The limit violation state determination results at each time step are sorted in chronological order, and the resulting ordered determination result set is the basic limit violation identifier of the standardized data sequence.

[0083] When evaluating the trend of exceeding limits in the standardized data sequence, starting from the first moment of the standardized data sequence, the continuous state of the basic exceeding limit indicator is tracked moment by moment. If a moment is in an exceeding limit state, the duration of consecutively being in the same exceeding limit state from that moment forward is counted. At the same time, combined with the cumulative deviation between the standardized data sequence value and the assessment benchmark interval during this duration, an index corresponding to its exceeding limit persistence is assigned to that moment. If the moment is in a non-exceeding limit state, the index is recorded as zero. The indices corresponding to all moments are arranged in chronological order, and the resulting ordered index set is the exceeding limit persistence index sequence of the standardized data sequence.

[0084] Based on the basic limit violation identifier and the limit violation persistence index sequence, when jointly identifying the limit violation pattern of the standardized data sequence, the basic limit violation identifier of that moment is associated with the corresponding limit violation persistence index at each time step. According to the type of the basic limit violation identifier and the magnitude of the limit violation persistence index, the limit violation pattern corresponding to that moment is identified. For example, a short-term single limit violation pattern corresponds to a basic limit violation identifier of a single limit violation and a small index, while a long-term continuous limit violation pattern corresponds to a basic limit violation identifier of consecutive identical limits and a large index. The limit violation pattern, limit violation type, and persistence status corresponding to each moment are integrated into a unified state description. The set of these descriptions arranged in chronological order is the limit violation representation of the standardized data sequence.

[0085] The beneficial effects are as follows: by quantifying the positional differences of standardized data sequences, a boundary distance vector reflecting the distance between the data and the boundary of the assessment benchmark interval is accurately obtained. Based on this vector, a preliminary judgment of the over-limit status is made, forming a basic over-limit identifier that clearly marks the over-limit situation at each time. Subsequently, the over-limit evolution trend of the standardized data sequences is evaluated to obtain an over-limit persistence index sequence that reflects the duration and degree of deviation of the over-limit. Finally, the basic over-limit identifier and the over-limit persistence index sequence are combined to jointly identify the over-limit pattern, generating an over-limit representation that includes the over-limit type, persistence status, and pattern characteristics. This comprehensively and accurately presents the over-limit status of the standardized data sequences, providing detailed and reliable over-limit information support for the subsequent multi-dimensional feature in-depth analysis of the target power plant's operating status.

[0086] S5. Based on the over-limit characterization and the execution error source components, perform multi-dimensional feature deep analysis on the operating state of the target power plant to obtain the dynamic feature representation of the operating state; In this embodiment of the invention, the step of performing multi-dimensional feature depth analysis on the operating state of the target power plant based on the limit violation characterization and the execution error source components to obtain a dynamic feature representation of the operating state includes: The spatiotemporal causal relationship between the out-of-limit representation and the execution error source component is quantified to obtain the causal correlation matrix of the standardized data sequence; Based on the causal correlation matrix, the core fluctuation patterns of the target power plant's operating state are mined to obtain the dominant oscillation mode of the operating state. Based on the time-varying trajectory of the dominant oscillation mode, the stability margin of the operating state is dynamically tracked to obtain the stability margin curve of the operating state. The dynamic feature representation of the operating state is obtained by fusing multi-source features from the causal correlation matrix, the dominant oscillation mode parameters, and the stability margin curve.

[0087] When quantifying the spatiotemporal causal relationship between the out-of-limit representation and the execution error source component, the type and duration of the out-of-limit representation and the magnitude change trend of the execution error source component are extracted at each time step. The out-of-limit representation and the execution error source component at the same time step are matched accordingly. The synchronous change relationship between the two in the time dimension and the degree of mutual influence in the feature dimension are analyzed. The correlation degree at each time step is quantified and assigned. The quantification results of all time steps are arranged in chronological order and feature dimension. The resulting two-dimensional ordered set is the causal correlation matrix of the standardized data sequence.

[0088] Based on the causal correlation matrix, when mining the core fluctuation patterns of the target power plant's operating status, the changes in the quantitative assignments of each dimension in the causal correlation matrix are comprehensively sorted out. The correlation features with the largest numerical changes and the most significant impact on the operating status are identified, and the operating status fluctuation patterns corresponding to these correlation features are extracted. These patterns can reflect the core driving factors and changing trends of the operating status fluctuations. The extracted core fluctuation patterns are the dominant oscillation modes of the operating status.

[0089] When dynamically tracking the stability margin of the operating state based on the time-varying trajectory of the dominant oscillation mode, the changes in the characteristic parameters of the dominant oscillation mode are recorded at each moment and arranged in chronological order to form the time-varying trajectory of the dominant oscillation mode. The stability of the operating state is judged based on the fluctuation amplitude and frequency of the trajectory. The smaller the fluctuation amplitude and the lower the frequency, the higher the stability and the larger the corresponding stability margin value. Conversely, the larger the fluctuation amplitude and the lower the frequency, the smaller the value. The stability margin values ​​at each moment are arranged in chronological order, and the resulting ordered set of values ​​is the stability margin curve of the operating state.

[0090] When performing multi-source feature fusion on the dominant oscillation mode parameters and the stability margin curve of the causal correlation matrix, the core correlation features in the causal correlation matrix, the key fluctuation parameters in the dominant oscillation mode, and the trend features in the stability margin curve are extracted. The three types of features are organically integrated, redundant information is eliminated, and key content that can comprehensively reflect the operating state is retained. The integrated feature set can dynamically present the core attributes of the operating state at different times. The integrated feature set is the dynamic feature representation of the operating state.

[0091] The beneficial effect is that by quantifying the spatiotemporal causal relationship between the out-of-limit characterization and the execution error source components, a causal correlation matrix that clearly reflects the correlation characteristics between the two is obtained. Based on this matrix, the core fluctuation law of the operating state is mined, and the dominant oscillation mode is extracted. Subsequently, the stability margin of the operating state is dynamically tracked according to the time-varying trajectory of the dominant oscillation mode to form a stability margin curve. Finally, the causal correlation matrix, the dominant oscillation mode parameters, and the stability margin curve are fused with multi-source features to generate a dynamic feature representation that can comprehensively and dynamically reflect the core attributes of the target power plant's operating state, providing accurate and comprehensive operating state information support for subsequent fusion quantitative scoring.

[0092] S6. Perform fusion quantification scoring on the dynamic feature representation to obtain the assessment result of the target power plant; In this embodiment of the invention, the step of performing fusion quantification scoring on the dynamic feature representation to obtain the assessment result of the target power plant includes: Based on the historical data of the target power plant, the dynamic feature representation is evaluated for importance in multiple dimensions to obtain the weight allocation of the dynamic feature representation. Based on the weight allocation, the dynamic feature representation is subjected to feature weighting and aggregation to obtain the comprehensive performance index of the target power plant; Based on the comprehensive performance indicators, the target power plant is scored in a standardized manner to obtain the assessment results of the target power plant.

[0093] Based on the historical data of the target power plant, when determining the multi-dimensional importance of the dynamic feature representation, all feature contents contained in the dynamic feature representation are extracted dimension by dimension. The feature of each dimension is comprehensively compared with the power plant's operation assessment and compliance status in the corresponding period in the historical data. The influence of each dimension feature on the assessment and compliance results is analyzed. The deeper the influence, the higher the importance level of the dimension feature; the shallower the influence, the lower the importance level of the dimension feature. The proportion of each dimension feature in the overall evaluation is determined according to its importance level. The combination of the proportions of all dimension features is the weight allocation of the dynamic feature representation.

[0094] Based on the weight allocation, when performing feature weighted aggregation on the dynamic feature representation, the representational strength of each dimension feature in the dynamic feature representation is adjusted according to the corresponding proportion determined by the weight allocation. Features with high importance are strengthened with a higher proportion, while features with low importance are weakened with a lower proportion. After all dimension features are adjusted, the adjusted feature contents of each dimension are integrated and summarized. The value obtained after summarizing that reflects the comprehensive operating level of the power plant is the comprehensive performance index of the target power plant.

[0095] When performing standardized scoring on the target power plant based on the comprehensive performance indicators, the comprehensive performance indicator values ​​in the historical data of the target power plant are comprehensively reviewed to determine the values ​​that reflect the best and worst operating levels. A complete scoring interval is defined based on these two values. The currently obtained comprehensive performance indicator values ​​are compared with the scoring interval one by one to determine the specific position of the value within the interval. The score at the corresponding position is the assessment result of the target power plant.

[0096] The beneficial effect is that by combining historical data of the target power plant to conduct multi-dimensional importance determination of dynamic feature representation, a weight allocation that fits the actual operation is obtained. Then, based on the weight allocation, the dynamic feature representation is weighted and aggregated to generate a comprehensive performance index that can fully reflect the overall operation level of the power plant. Finally, the comprehensive performance index is standardized and scored based on the scoring interval defined by historical data, resulting in an objective, fair and comparable assessment result, providing an accurate and reliable basis for evaluating the power generation plan execution of the target power plant.

[0097] Please see Figure 2 In another embodiment of the present invention, a power generation plan assessment system based on spatiotemporal consistency regularization is provided. This system can be used to implement the above-mentioned power generation plan assessment method based on spatiotemporal consistency regularization. Specifically, the power generation plan assessment system based on spatiotemporal consistency regularization includes a data module, an analysis module, a deduction module, an identification module, an analysis module, and a scoring module.

[0098] The data module is used to perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain a standardized data sequence of the target power plant. The analysis module is used to receive the standardized data sequence output by the data module, perform correlation entropy blind source separation on the deviation data of the standardized data sequence, and obtain the planning error source component and the execution error source component of the deviation data. The deduction module is used to receive the planning error source component output by the analysis module, and perform dynamic confidence interval deduction on the assessment benchmark value of the target power plant based on the planning error source component to obtain the assessment benchmark interval of the target power plant. The identification module is used to receive the standardized data sequence output by the data module and the assessment benchmark interval output by the inference module, respectively, and compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence. The parsing module is used to receive the limit violation characterization output by the identification module and the execution error source component output by the analysis module, respectively, and perform multi-dimensional feature deep analysis on the operating state of the target power plant based on the limit violation characterization and the execution error source component to obtain the dynamic feature representation of the operating state. The scoring module is used to receive the dynamic feature representation output by the parsing module, perform fusion quantification scoring on the dynamic feature representation, and obtain the assessment result of the target power plant.

[0099] This invention provides a terminal device comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or function. The processor described in this embodiment can be used in the operation of a power generation plan assessment method based on spatiotemporal consistency regularization, including: The real-time operation data and power generation plan data of the target power plant are spatiotemporally consistent and normalized to obtain a standardized data sequence for the target power plant. Correlation entropy blind source separation is performed on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data. Based on the planning error source component, dynamic confidence interval extrapolation is performed on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval for the target power plant. A time-by-time variable comparison is performed between the standardized data sequence and the assessment benchmark interval to obtain the limit-crossing characterization of the standardized data sequence. Based on the limit-crossing characterization and the execution error source component, multi-dimensional feature deep analysis is performed on the operating status of the target power plant to obtain a dynamic feature representation of the operating status. A fusion quantitative scoring is performed on the dynamic feature representation to obtain the assessment result of the target power plant.

[0100] Please see Figure 3The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When executed by the processor 61, the computer program 63 implements the power generation plan assessment method based on spatiotemporal consistency regularization in this embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when executed by the processor 61, the computer program 63 implements the functions of each model / unit in the power generation plan assessment system based on spatiotemporal consistency regularization in this embodiment. To avoid repetition, these details are not elaborated here.

[0101] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 3 This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0102] The processor 61 may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0103] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the computer device 60.

[0104] Furthermore, the memory 62 may include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0105] Please see Figure 4 The terminal device is an electronic device 600, which is manifested in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.

[0106] The storage unit stores program code, which can be executed by the processing unit 610 to perform the steps described in the method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 610 can perform actions such as... Figure 1 The steps are shown in the figure.

[0107] Storage unit 620 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 6201 and / or cache memory 6202, and may further include a read-only memory (ROM) 6203.

[0108] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0109] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0110] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem). This communication can be performed via input / output interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network, wide area network, and / or public network, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0111] Example 4 This invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the terminal device and extended storage media supported by the terminal device; it can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which can be one or more computer programs (including program code). More specific examples of the computer-readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical fiber, portable compact disk read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0112] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium can also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, radio frequency, etc., or any suitable combination thereof.

[0113] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0114] One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the power generation plan assessment method based on spatiotemporal consistency regularization in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps: The real-time operation data and power generation plan data of the target power plant are spatiotemporally consistent and normalized to obtain a standardized data sequence for the target power plant. Correlation entropy blind source separation is performed on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data. Based on the planning error source component, dynamic confidence interval extrapolation is performed on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval for the target power plant. A time-by-time variable comparison is performed between the standardized data sequence and the assessment benchmark interval to obtain the limit-crossing characterization of the standardized data sequence. Based on the limit-crossing characterization and the execution error source component, multi-dimensional feature deep analysis is performed on the operating status of the target power plant to obtain a dynamic feature representation of the operating status. A fusion quantitative scoring is performed on the dynamic feature representation to obtain the assessment result of the target power plant.

[0115] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0116] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0117] Experimental scenario setting Three power plants of different types (thermal power plant A, hydropower plant B, and renewable energy power plant C) were selected as experimental subjects. Real-time operational data and power generation plan data for 30 consecutive days were used as experimental samples, including core indicators such as power generation, unit load, and operating time. The experiment was divided into two groups: the control group used existing technology (fixed benchmark + single limit-breaking index scoring), and the experimental group used the method of this invention. The superiority of this invention was verified by comparing the key indicators of the two groups.

[0118] Experimental data results

[0119] The experimental group significantly improved the time synchronization rate, format uniformity rate, and outlier removal accuracy of standardized data through spatiotemporal consistency regularization, increasing them by 15.5% compared to the control group, thus completely solving the problem of insufficient data preprocessing accuracy.

[0120] The experimental group used the relevant entropy blind source separation technology to accurately separate the sources of planning and execution errors, achieving an accuracy rate of 92.4%, which is 26.7% higher than the control group, and realizing the accurate definition of assessment responsibilities.

[0121] The dynamic assessment benchmark range of the experimental group was adjusted in real time according to the fluctuation of the planning error, and its adaptability to the actual operation scenario was improved by 22.9% compared with the control group, effectively avoiding unfair assessment.

[0122] A questionnaire survey of relevant personnel from three power plants showed that the acceptance rate of the experimental group's assessment results reached 91%, which was 23% higher than that of the control group, indicating that the fairness and scientific nature of the method of this invention are more easily accepted.

[0123] The optimization suggestions based on multi-dimensional feature analysis output in the experimental group are more targeted. After implementation, the power plant's power generation plan execution efficiency increased by an average of 27.2%, which is much higher than that of the control group, verifying the supporting role of the invention in operation optimization.

[0124] This invention achieves high-precision data standardization through spatiotemporal consistency regularization, ensuring data reliability; it precisely separates planning and execution error sources using relevant entropy blind source separation, enabling accurate error attribution and providing support for responsibility delineation; it deduces dynamic assessment benchmark intervals based on planning error source components, improving the adaptability of benchmarks to actual scenarios; and it comprehensively characterizes the actual operating level of power plants through multi-dimensional feature deep analysis and fusion quantitative scoring. Overall, it solves the problems of insufficient data preprocessing, coarse error attribution, fixed benchmarks, and singular evaluation in existing technologies, significantly improving the scientific rigor, fairness, and comprehensiveness of power generation plan assessment, while providing precise direction for subsequent operational optimization and enhancing the overall efficiency of power generation plan execution.

[0125] In summary, this invention provides a power generation plan assessment method and system based on spatiotemporal consistency regularization. Through spatiotemporal consistency regularization, it achieves high-precision standardized processing of real-time operational data and power generation plan data, ensuring data quality. Combined with relevant entropy blind source separation technology, it can accurately separate the planning error source components and execution error source components in deviation data, achieving accurate error attribution and providing reliable technical support for defining assessment responsibilities, significantly improving the scientific validity and credibility of assessment results. Based on the planning error source components, it derives a dynamic assessment benchmark interval, allowing the assessment boundary to be dynamically adjusted according to the fluctuation characteristics of planning errors, enhancing the adaptability of the assessment benchmark to actual operating scenarios. Through multi-dimensional feature deep analysis and fusion quantitative scoring, it comprehensively correlates execution errors with operating status characteristics, accurately characterizing the actual operating level of the power plant, while providing precise directions for subsequent operation optimization, improving the overall efficiency of power generation plan execution.

[0126] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0127] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0128] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0129] In the embodiments provided by this invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0130] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0131] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0132] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random-access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0133] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0134] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0135] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0136] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A power generation plan assessment method based on spatiotemporal consistency regularization, characterized in that, Includes the following steps: S1. Perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain the standardized data sequence of the target power plant; S2. Perform correlation entropy blind source separation on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data; S3. Based on the planned error source components, perform dynamic confidence interval extrapolation on the assessment benchmark value of the target power plant to obtain the assessment benchmark interval of the target power plant; S4. Compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence; S5. Based on the over-limit characterization and the execution error source components, perform multi-dimensional feature deep analysis on the operating state of the target power plant to obtain the dynamic feature representation of the operating state; S6. Perform fusion quantification scoring on the dynamic feature representation to obtain the assessment results of the target power plant.

2. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S1, the real-time operating data and power generation plan data of the target power plant are spatiotemporally consistent and normalized to obtain a standardized data sequence of the target power plant, including: S101. Perform precise timestamp correction on the real-time operation data and power generation plan data of the target power plant to obtain the time synchronization data pair of the target power plant; S102. Standardize the format of the time synchronization data pair to obtain a unified data sequence of the time synchronization data pair; S103. Remove outliers from the unified data sequence to obtain the standardized data sequence of the target power plant.

3. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S2, correlation entropy blind source separation is performed on the deviation data of the standardized data sequence to obtain the planning error source component and the execution error source component of the deviation data, including: S201. Perform time-frequency distribution mapping on the deviation data of the standardized data sequence to obtain the time-frequency energy distribution spectrum of the deviation data; S202. Based on the power generation plan data, perform periodic pattern feature matching on the time-frequency energy distribution map to obtain the dominant mode component of the deviation data; S203. Track the sudden abnormal process of the deviation data to obtain the transient impact event sequence of the deviation data; S204. Based on the dominant mode component and the transient impact event sequence, perform correlation entropy constraint projection separation on the deviation data to obtain the planning error source component and the execution error source component of the deviation data.

4. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 3, characterized in that, Step S204 is as follows: Time-frequency feature analysis is performed on the dominant mode component and the transient impact event sequence to obtain the time-frequency distribution pair of the deviation data; Based on the time-frequency distribution pairs, calculate the projection separation weight of the time-frequency energy distribution map; Based on the projection separation weight, a weighted mapping matching is performed on the time-frequency energy distribution map to obtain the planning error projection component and the execution error projection component of the time-frequency energy distribution map. The planning error projection component and the execution error projection component are subjected to inverse time-frequency transformation to obtain the planning error source component and the execution error source component of the deviation data.

5. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S3, based on the planned error source components, a dynamic confidence interval is derived for the target power plant's assessment benchmark value to obtain the target power plant's assessment benchmark interval, including: S301. Reconstruct the kernel density distribution of the planning error source components to obtain the joint probability density surface of the planning error source components; S302. Based on the joint probability density surface, perform contour cluster adaptive extraction on the planning error source components to obtain the probability contour cluster set of the planning error source components; S303. Map the probability equal-height clusters to a preset assessment tolerance risk level to obtain the selected probability equal-height clusters of the planning error source components; S304. Based on the envelope boundary of the selected probability equal-high cluster, the assessment benchmark value of the target power plant is dynamically divided to obtain the assessment benchmark interval of the target power plant.

6. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 5, characterized in that, Step S302 specifically includes: A topological invariant analysis is performed on the joint probability density surface to obtain the Betti number sequence of the joint probability density surface; Based on the Betti number sequence, a topological constraint field is constructed on the joint probability density surface to obtain the dynamic decision potential field of the joint probability density surface. The potential energy function of the joint probability density surface is obtained by performing a negative logarithmic transformation on the probability density values ​​on the joint probability density surface. The variance and kurtosis statistical characteristics of the planning error source components are normalized to obtain the equivalent quality parameters of the planning error source components. Based on the dynamic decision potential field, the barrier penetration capability of the probability distribution of the planning error source components is evaluated to obtain the tunneling intensity between two points in the joint probability density surface. Based on the tunneling intensity, the probability mass of the joint probability density surface is redistributed to obtain the enhanced probability density distribution of the joint probability density surface. By contour closure of the enhanced probability density distribution, the probability equal-height clusters of the planning error source components are obtained.

7. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S4, the standardized data sequence is compared with the benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence, including: S401. Based on the assessment benchmark interval, the standardized data sequence is subjected to position difference quantization to obtain the boundary distance vector of the standardized data sequence; S402. Based on the boundary distance vector, perform a preliminary determination of the limit-crossing state of the standardized data sequence to obtain the basic limit-crossing identifier of the standardized data sequence; S403. Evaluate the out-of-limit evolution trend of the standardized data sequence to obtain the out-of-limit persistence index sequence of the standardized data sequence; S404. Based on the basic limit violation identifier and the limit violation persistence index sequence, perform joint limit violation pattern identification on the standardized data sequence to obtain the limit violation characterization of the standardized data sequence.

8. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S5, based on the limit violation characterization and the execution error source components, a multi-dimensional feature depth analysis is performed on the operating state of the target power plant to obtain a dynamic feature representation of the operating state, including: S501. The spatiotemporal causal relationship between the out-of-limit representation and the execution error source component is quantified to obtain the causal correlation matrix of the standardized data sequence. S502. Based on the causal correlation matrix, the core fluctuation pattern of the operating state of the target power plant is mined to obtain the dominant oscillation mode of the operating state. S503. Based on the time-varying trajectory of the dominant oscillation mode, the stability margin of the operating state is dynamically tracked to obtain the stability margin curve of the operating state. S504. Perform multi-source feature fusion on the causal correlation matrix, the dominant oscillation mode parameters, and the stability margin curve to obtain a dynamic feature representation of the operating state.

9. The power generation plan assessment method based on spatiotemporal consistency regularization according to claim 1, characterized in that, In step S6, the dynamic feature representation is fused and quantified to obtain the assessment result of the target power plant, including: S601. Based on the historical data of the target power plant, perform multi-dimensional importance determination on the dynamic feature representation to obtain the weight allocation of the dynamic feature representation; S602. Based on the weight allocation, perform feature weighting and aggregation on the dynamic feature representation to obtain the comprehensive performance index of the target power plant; S603. Based on the comprehensive performance indicators, the target power plant is scored in a standardized manner to obtain the assessment results of the target power plant.

10. A power generation plan assessment system based on spatiotemporal consistency regularization, characterized in that, include: The data module is used to perform spatiotemporal consistency normalization on the real-time operation data and power generation plan data of the target power plant to obtain a standardized data sequence of the target power plant. The analysis module is used to receive the standardized data sequence output by the data module, perform correlation entropy blind source separation on the deviation data of the standardized data sequence, and obtain the planning error source component and the execution error source component of the deviation data. The deduction module is used to receive the planning error source component output by the analysis module, and perform dynamic confidence interval deduction on the assessment benchmark value of the target power plant based on the planning error source component to obtain the assessment benchmark interval of the target power plant. The identification module is used to receive the standardized data sequence output by the data module and the assessment benchmark interval output by the inference module, respectively, and compare the standardized data sequence with the assessment benchmark interval at each time step to obtain the limit-crossing characterization of the standardized data sequence. The parsing module is used to receive the limit violation characterization output by the identification module and the execution error source component output by the analysis module, respectively, and perform multi-dimensional feature deep analysis on the operating state of the target power plant based on the limit violation characterization and the execution error source component to obtain the dynamic feature representation of the operating state. The scoring module is used to receive the dynamic feature representation output by the parsing module, perform fusion quantification scoring on the dynamic feature representation, and obtain the assessment result of the target power plant.

Images