Method and system for evaluating success rate of whole-process black start based on network-constructed energy storage

Abstract

The application discloses a whole-process black start success rate evaluation method and system based on network-constructed energy storage, and belongs to the technical field of network-constructed energy storage operation evaluation. In view of the problem that existing black start evaluation methods are focused on offline planning and are difficult to accurately evaluate the whole-process success rate of candidate combinations by fusing real-time operation data, the application obtains real-time operation state information of each candidate combination through multi-source credibility correction, divides the whole process of black start into multiple stages, respectively constructs operation margins, and then obtains stage real-time probabilities by fusing topological characteristics and real-time information correction through an operation graph time sequence probability model, and finally comprehensively evaluates the whole-process success rate according to a series relationship. Real-time quantitative evaluation of the success rate of candidate combinations is realized, accurate decision basis is provided for dispatchers, and the evaluation accuracy is improved.

Description

Technical Field

[0001] This invention relates to the field of grid-type energy storage operation evaluation technology, and in particular to a method and system for evaluating the success rate of black start-up throughout the entire process based on grid-type energy storage. Background Technology

[0002] The statements in this section are merely to provide background information related to the present invention and do not necessarily constitute prior art.

[0003] When a large power grid experiences a complete or widespread blackout, thermal power units typically cannot restore power on their own due to the interruption of their auxiliary power supply. They require black-start power sources to supply power through a starting channel to restore output. Grid-based energy storage, with its voltage source control capabilities, can autonomously establish voltage and frequency under weak grid or islanded conditions, making it an important new type of power source supporting the black start of thermal power units. In actual blackout restoration, dispatchers need to determine the priority execution plan from numerous candidate combinations consisting of grid-based energy storage, starting channels, and target units. The success rate of black start for these candidate combinations is affected not only by the status of the starting channel and the unit's power-on readiness but also by multiple real-time factors such as the energy storage's charge level, the converter's current-limiting status, and the grid control margin. Significant differences in success rates exist between different combinations, and relying solely on pre-outage plans or empirical rules for selection can easily lead to startup failure or restoration delays. Existing research mainly focuses on black-start resource planning and offline recovery path optimization before power outages. It is difficult to integrate real-time measurement and stage feedback to quickly and accurately assess the success rate of candidate combinations throughout the entire process. This results in the inability to generate assessment results that match the actual operating status, making it difficult to provide dispatchers with accurate decision-making basis. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for evaluating the success rate of black start-up based on grid-type energy storage. By using multi-source data correction, multi-stage margin modeling, and time-series probability evaluation of the operation graph, the success rate of candidate combinations can be quantitatively evaluated in real time, thereby improving the accuracy of the evaluation.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: In a first aspect, this invention provides a method for evaluating the success rate of black start-up throughout the entire process based on grid-connected energy storage, comprising: Real-time running data of each candidate black start combination is obtained, and black start running status information is obtained through multi-source credibility correction; The entire black start process is divided into multiple stages, and the runtime margin of each stage is constructed. The black start running status information and the runtime margin of each stage are processed by the runtime graph time series probability model to obtain the prior probability of each stage. The prior probability of each stage is then corrected based on real-time information to obtain the real-time probability of each stage. Based on the real-time probability of each stage, the success rate of black start for each candidate combination throughout the entire process is comprehensively evaluated according to the serial relationship of each stage.

[0006] Furthermore, the candidate black start combination includes grid-type energy storage, a start-up channel, and a target thermal power unit; The real-time operation data includes at least one of the following: dispatch-side operation data, station-end equipment status, energy storage-side operation data, and unit-side startup status.

[0007] Furthermore, the multi-source credibility correction determines the credibility of each data source based on data timeliness, communication quality, reasonableness of exceeding limits, and multi-source consistency, and performs weighted correction on multi-source data with the same state quantity based on the credibility.

[0008] Furthermore, the entire black start process is divided into three stages: the first stage, the second stage, and the third stage. The first-stage operational margin is constructed based on the current-carrying capacity and voltage reactive power regulation capacity of the energy storage converter. The second phase of operational margin is constructed based on power energy support capacity and dynamic quality of plant service busbars; The third-stage operational margin is constructed based on the remaining support time of energy storage and the completeness of the target unit's startup.

[0009] Furthermore, including: Using grid-type energy storage, the power station busbars along the startup channel, and the target thermal power unit as nodes, and the connection relationships of lines, transformers, and switches in the startup channel as edges, a time-series probability model of the operation diagram is constructed. Based on the black start running state information and the running margin of each stage, the running graph time-series state features of the candidate black start combination are extracted, and the prior probability of each stage is obtained based on the running graph time-series state features.

[0010] Furthermore, the runtime time-series probability model is trained based on at least one of historical black-start test data, scheduling exercise data, transient simulation data, and accident recovery records, and its training objectives include success / failure label loss and margin monotonicity constraints.

[0011] Furthermore, the comprehensive assessment specifically involves calculating the black start success rate throughout the entire process based on the series relationship between the grid-type energy storage empty charging stage, the auxiliary machine start-up stage, and the thermal power unit start-up stage. Completed stages are treated as confirmed successes; stages that have been confirmed as failures trigger failure determination for the corresponding candidate combinations; and stages that have not yet been completed and have not been confirmed as failures are included in the calculation according to real-time probabilities.

[0012] In a second aspect, the present invention provides a system for evaluating the success rate of black start-up throughout the entire process of grid-based energy storage, comprising: Data acquisition module: used to acquire real-time running data of each candidate black start combination, and obtain black start running status information through multi-source confidence correction; Phase Division and Margin Calculation Module: This module is used to divide the entire black start process into multiple phases and construct the runtime margin for each phase. The probability evaluation module is used to process the black start running status information and the running margin of each stage through the running graph time series probability model to obtain the prior probability of each stage; and to correct it according to the real-time information to obtain the real-time probability of the stage. Comprehensive evaluation module: Used to comprehensively evaluate the success rate of black start of each candidate combination throughout the entire process based on the real-time probability of each stage and the serial relationship of each stage.

[0013] In a third aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the method as described in the first aspect.

[0014] In a fourth aspect, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method as described in the first aspect.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention ensures that the data used in the assessment truly reflects the actual state of the equipment through multi-source reliability correction, avoiding misjudgments caused by a single abnormal data point. By dividing the entire black start process into multiple stages and constructing operational margins for each stage, traditional experience-based judgments are transformed into quantifiable physical support indicators. Furthermore, topological correlation features and stage transmission features are extracted through a time-series probability model of the operation diagram, and the probability is dynamically corrected in combination with real-time information, so that the assessment results can reflect both theoretical support capabilities and respond to sudden changes on site. Finally, the success rate of the entire process is comprehensively assessed according to the series relationship, providing dispatchers with a quantitative basis for decision-making, effectively avoiding start-up failures or recovery delays caused by selecting schemes based on pre-outage plans or experience rules, and improving the accuracy of the success rate assessment of grid-type energy storage black start combinations. Attached Figure Description

[0016] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0017] Figure 1 This is a flowchart of the method according to Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the three-stage black-start operational margin model of the present invention; Figure 3A flowchart for evaluating the success rate of a stage based on the timing state characteristics of the runtime graph; Figure 4 This diagram illustrates the comprehensive calculation of success rate and identification of risk bottlenecks throughout the entire process. Detailed Implementation

[0018] The following detailed description is exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0020] In this embodiment of the invention, the data collection and processing strictly adhere to the requirements of relevant laws and regulations, obtaining informed consent or separate consent from the data subject, and conducting subsequent data use and processing within the scope of laws, regulations, and the data subject's authorization. All data acquisition in this embodiment is based on compliance with laws and regulations and user consent, representing the lawful application of the data.

[0021] Example 1 Compared to the shortcomings of existing technologies, such as the fact that existing research mainly focuses on black-start resource planning and offline recovery path optimization before power outages, it is difficult to integrate real-time measurements, equipment status and stage feedback after power outages to quickly and accurately evaluate the success rate of candidate combinations throughout the entire process. This makes it difficult to form evaluation results that match the actual operating status and provide accurate decision-making basis for dispatchers.

[0022] To address the aforementioned issues, this invention first establishes a multi-source operational status credibility correction model. Based on real-time operational data collected from the dispatching side, station side, energy storage side, and unit side, and considering data timeliness, communication quality, reasonableness of exceeding limits, and multi-source consistency, the same status quantity is weighted and corrected for credibility. This solves the problem of inaccurate assessment caused by an abnormal single data source in the prior art, ensuring that the data used in subsequent assessments truly reflects the actual status of the equipment.

[0023] Based on this, a three-stage black start operation margin model is established, which divides the entire process into three stages: grid-type energy storage air charging, auxiliary machine start-up, and thermal power unit start-up. Based on the dominant operation constraints of each stage, air charging margin, auxiliary machine start-up margin, and unit start-up margin are established respectively. The recovery process that traditionally relies on experience-based judgment is transformed into quantifiable physical support indicators, which solves the problem of lacking quantitative means for supporting the energy storage at each stage in existing technologies.

[0024] Furthermore, a stage success rate evaluation model based on the timing state characteristics of the operation diagram is established. The prior probability of each stage is calculated by extracting topological association features and stage transmission features through the timing probability model of the operation diagram. The model is then dynamically corrected by combining real-time information such as closing feedback, protection actions, converter interlocking, auxiliary machine start-up and shutdown, and unit status. This solves the problem that offline planning in the existing technology cannot respond to actual state changes after power outage, and makes the evaluation results both theoretically supported and capable of on-site situational awareness.

[0025] Finally, a comprehensive evaluation model for the success rate of black start-up throughout the entire process and a risk bottleneck identification model were established. The success rate of the entire process was calculated based on the serial relationship of each stage, and risk bottlenecks were identified based on the failure risk of the incomplete stage. This solved the problem that existing technologies can only evaluate a single link or offline scenario and lack a comprehensive evaluation method for the entire process. It realized a comprehensive evaluation from stage success rate to the success rate of the entire process, and improved the accuracy of the success rate evaluation of grid-type energy storage black start combination.

[0026] Therefore, in a typical embodiment of the present invention, such as Figure 1 As shown, a method for evaluating the success rate of black start based on grid-type energy storage throughout the entire process is disclosed. The specific steps include: S1: Obtain real-time running data of each candidate black start combination, and obtain black start running status information through multi-source credibility correction; S2: Divide the entire black boot process into multiple stages and build the runtime margin for each stage. S3: The black start running status information and the running margin of each stage are processed through the running graph time series probability model to obtain the prior probability of each stage; and corrected according to the real-time information to obtain the real-time probability of the stage; S4: Based on the real-time probability of each stage, comprehensively evaluate the black start success rate of each candidate combination throughout the entire process according to the serial relationship of each stage.

[0027] The following section provides a detailed explanation of the above-mentioned method for evaluating the success rate of black start-up based on grid-type energy storage throughout the entire process, using specific implementation methods as examples.

[0028] In S1, the candidate black start combination includes grid-type energy storage, a start-up channel, and a target thermal power unit; the start-up channel is the power transmission path between the grid-type energy storage and the power receiving point of the target thermal power unit; the success rate of the entire black start process is judged by the target thermal power unit completing the start-up and forming the minimum stable output.

[0029] Real-time operational data includes at least one of dispatch-side operational data, station-side equipment status, energy storage-side operational data, and unit-side startup status; the multi-source credibility correction determines the credibility of each data source based on data timeliness, communication quality, reasonableness of exceeding limits, and multi-source consistency, and performs weighted correction on multi-source data of the same status quantity based on the credibility.

[0030] Real-time operational data from multiple sources are collected for each candidate black-start combination, including grid-connected energy storage, start-up channels, and target thermal power units. This includes the state of charge, converter current limiting values, and grid control parameters on the energy storage side; line voltage, switch positions, and protection signals on the channel side; and auxiliary equipment status, speed, excitation, and plant power information on the unit side. Reliability weights are calculated based on the timeliness, communication quality, reasonableness of exceeding limits, and consistency of each data source. Data from multiple sources for the same state quantity is weighted and fused to reduce the impact of lagging, biased, or conflicting data, while increasing the impact of consistent and reliable data. This forms an information foundation that accurately reflects the actual operating status of each device, providing reliable input for subsequent success rate assessment.

[0031] After a major power outage, a candidate black start combination set is first constructed based on the currently available grid-connected energy storage, available start-up channels, and thermal power units to be started. Let the first... i The candidate black start combinations are represented as follows: (1) in, Indicates the first i A candidate black start combination This indicates the grid-type energy storage in this combination. This indicates the startup channel corresponding to this combination. Indicates the target thermal power unit in this combination. Time. t The set of candidate black start combinations is represented as: (2) in, Indicates time t The following is a set of candidate black-start combinations formed based on real-time availability. NThis indicates the number of candidate combinations. For each candidate black-start combination, real-time operating data of its related equipment is collected. Real-time data sources include the Energy Management System (EMS), Supervisory Control and Data Acquisition (SCADA) system, Wide Area Measurement System (WAMS) and its synchronous phasor measurement unit (PMU), Battery Management System (BMS), Power Conversion System (PCS), Relay Protection and Fault Recording System, Distributed Control System (DCS) for thermal power units, and on-site operation confirmation information.

[0032] The above data is categorized into four states: grid-type energy storage status, start-up channel status, plant auxiliary system status, and target thermal power unit status. Grid-type energy storage status characterizes the energy storage grid support capability; start-up channel status characterizes the operability of the no-load charging line and transformer; plant auxiliary system status characterizes the auxiliary equipment start-up conditions; and target thermal power unit status characterizes the unit start-up progress and the possibility of achieving minimum stable output.

[0033] For the same state variable, different data sources may differ in timescale, refresh cycle, and reliability. This embodiment calculates data reliability based on data timeliness, communication quality, reasonableness of exceeding limits, and multi-source consistency, and then performs weighted correction on multi-source data of the same state variable based on this reliability. (3) In the formula, Indicates time t Next i State quantity in each candidate combination The value after confidence correction; Indicates the first r State variables given by data sources ; This indicates that the data source has state variables. The credibility of data is determined by the timeliness of data, communication quality, reasonableness of exceeding limits, and consistency among multiple sources. R i,a Represents the participating state quantity The number of data sources to be corrected; e To prevent extremely small positive numbers with a denominator of zero.

[0034] It should be noted that in this invention e The specific value is determined based on the units and precision used. e It must be at least 3 to 4 orders of magnitude smaller than all other significant values ​​involved in the calculation in the formula, and is usually taken as 10. -6 Quantity or 10 -8 The magnitude is not specified; this value is merely illustrative and those skilled in the art can select an appropriate, extremely small positive number based on the actual data magnitude, and does not constitute a limitation on the present invention.

[0035] If a data source experiences communication delays, sudden measurement changes that contradict the physical constraints of the equipment, or conflicts with hard states such as protection actions and switch positions, its reliability decreases. Conversely, if a data source is consistent with adjacent measurements, operation ticket execution results, and on-site confirmation information, its reliability increases. After processing using equation (1), a black-start evaluation state matrix is ​​formed. .

[0036] In S2, the entire black start process is divided into three stages: the first stage, the second stage, and the third stage. These stages represent the grid-type energy storage empty-charge stage, the auxiliary equipment start-up stage, and the thermal power unit start-up stage, respectively. The margin in the grid-type energy storage empty-charge stage characterizes the support margin of the grid-type energy storage for the empty-charge process of the start-up channel. The margin in the auxiliary equipment start-up stage characterizes the support margin of the grid-type energy storage for the start-up process of the key auxiliary equipment of the target thermal power unit. The margin in the thermal power unit start-up stage characterizes the degree of matching between the continuous power supply capacity of the grid-type energy storage and the start-up process of the target thermal power unit. The first stage operating margin is constructed based on the current withstand capability and voltage reactive power regulation capability of the energy storage converter; the second stage operating margin is constructed based on the power energy support capability and the dynamic quality of the plant service bus; and the third stage operating margin is constructed based on the remaining support time of the energy storage and the start-up integrity of the target unit.

[0037] Specifically, this embodiment divides the black start process of the target thermal power unit supported by grid-type energy storage into three stages, such as... Figure 2 As shown. The first stage is the grid-type energy storage no-load charging stage, where, after establishing a local voltage frequency, the grid-type energy storage charges the unloaded transformers and lines in the start-up channel, enabling the target thermal power unit's auxiliary power receiving point to be energized. The second stage is the auxiliary equipment start-up stage, where the grid-type energy storage supplies power to the target thermal power unit's auxiliary system through the energized start-up channel, supporting the start-up of key auxiliary equipment. The third stage is the thermal power unit start-up stage, where, after the target thermal power unit meets the requirements for auxiliary power and auxiliary equipment, it completes speed-up, excitation, grid connection, or reaches minimum stable output. After the third stage is completed, the main black start process is considered successful.

[0038] Set a completion status and a failure status for each stage. For the... i The first candidate combination kEach stage has a completion status and a failure status. The completion status is denoted as... d i,k ( t ); clearly define the failure status as f i,k ( t When a certain stage is confirmed to be complete by switch status, protection status, PCS status, DCS status, or field confirmation information, then... d i,k ( t The value equals 1; when direct failure evidence such as protection tripping, PCS lockout, auxiliary machine startup failure, or thermal power unit startup failure occurs during this stage, the value is set to 1. f i,k ( t The value is equal to 1. The completion and failure states of the same stage are not both set to 1.

[0039] The decisive constraints during the no-load charging phase are the current-carrying capacity and voltage reactive power regulation capacity of the energy storage converter, and its operating margin is: (4) In the formula, M i,1 ( t This indicates the margin of energy storage capacity during the grid-type energy storage and charging phase. Indicates grid-type energy storage s i The PCS currently allows output current; Indicates candidate launch channel The maximum equivalent current expected to be caused by the hollow charging operation is determined by the dominant term in the transformer inrush current and the line charging current. This indicates the voltage reactive power adjustable capability that the candidate combination can currently provide; This indicates the voltage reactive power regulation capability required to complete the no-load charging operation of this candidate combination; e To prevent extremely small positive numbers with a denominator of zero.

[0040] The decisive constraints during the auxiliary equipment startup phase are the capacity of grid-type energy storage to support the startup power of the auxiliary equipment, and the frequency quality of the plant service bus voltage after the auxiliary equipment is put into operation. Its operating margin is: (5) In the formula, M i,2 ( t () indicates the margin during the auxiliary machine startup phase; Indicates grid-type energy storage s i Current comprehensive power and energy support capabilities available for auxiliary machine startup; Indicates the target thermal power unit g iThe key auxiliary machines are put into operation according to the current startup sequence to meet the comprehensive power and energy requirements. This indicates the expected dynamic offset of the plant busbar caused by the start-up of auxiliary equipment; This indicates the maximum allowable dynamic offset of the plant system; e To prevent extremely small positive numbers with a denominator of zero.

[0041] The decisive constraints during the start-up phase of thermal power units are the remaining support time of grid-type energy storage and the start-up integrity of the target thermal power units. Its operational margin is: (6) In the formula, M i,3 ( t ) indicates the first i The startup margin of each candidate combination of thermal power units; Indicates grid-type energy storage s i The estimated time during which the target unit's power supply can be maintained under current output and current state of charge; Indicates the target thermal power unit g i The estimated remaining time required from the current startup state to reach minimum stable output; Indicates the target thermal power unit g i The completeness of the startup is obtained by comprehensively considering the status of plant power supply, auxiliary equipment, speed, excitation, and protection. This represents the integrity threshold required to determine whether a thermal power unit meets the conditions for continued startup. e To prevent extremely small positive numbers with a denominator of zero.

[0042] In step S3, the previously obtained black start operation status information and operation margins of each stage are input into the operation diagram time-series probability model. The operation diagram time-series probability model constructs an operation diagram with grid-type energy storage, the busbars of the power plants along the start-up channel and the target thermal power unit as nodes, and the connection relationships of lines, transformers and switches in the start-up channel as edges, to characterize the electrical connection relationships of candidate black start combinations. The topological association features between each device and the transmission features between stages are extracted through graph convolution and gated cyclic units, and the prior success probability of each stage is calculated quantitatively based on the current margin in the absence of real-time feedback. On this basis, real-time signals such as closing feedback, protection action, converter lockout, auxiliary machine start-up and shutdown and unit speed are further connected to the field to perform Bayesian correction on the prior probability: if the received signal supports the smooth progress of the stage, the probability is increased; if an abnormality or failure sign is detected, the probability is decreased, thereby forming a stage real-time success probability that integrates offline model deduction and online real-time situation, so that the evaluation results can reflect both theoretical support capabilities and quickly respond to dynamic changes in the actual execution process.

[0043] The runtime graph temporal probability model includes a graph convolutional layer, a graph attention layer, a gated recurrent unit, and a probability output layer. It is used to extract the topological association features and stage transfer features of candidate black-start combinations, and output the prior probability of each stage based on the runtime state information and the runtime margin of each stage.

[0044] This embodiment addresses each candidate black-start combination. c i An operation diagram is constructed based on the electrical connection relationship between its grid-type energy storage, start-up channel, and target thermal power unit. The nodes in the operation diagram represent the operating status of grid-type energy storage, the busbars of power plants along the start-up channel, and the target thermal power units. The edges in the operation diagram represent the connection relationships of lines, transformers, and switches. Based on the operation diagram, the confidence-corrected operation state matrix, and the state characteristics of the preceding stages, the [previous stage] is updated. Timing state characteristics of the runtime graph for each candidate combination: (7) In the formula, h i ( t () indicates time t The timing state characteristics of the running graph; Indicates parameters i The runtime time-series probability model; This indicates the current running graph of the black boot process; Indicates the first i The running state matrix of candidate combinations; This represents the timing characteristics of the runtime graph at the previous evaluation time, used to preserve the impact of preceding operations on the current stage. Δ t Indicates the evaluation time interval.

[0045] Furthermore, based on the timing state characteristics of the runtime graph, the first... k Prior success probability at each stage: (8) In the formula, π i,k ( t ) indicates the first i Candidate combination number k The prior success probability of the stage; σ( () is the Sigmoid function; α k w k , and r k For the first k The parameters of the model to be trained for each stage; Pool( The ) indicates a pooling operation, used to compress the temporal representations of multiple nodes into a stage representation. Indicates the first i Among the candidate combinations, the first k A set of nodes related to a specific stage. i,v ( t ) represents a node v At any moment t The timing state characteristics of the running graph; M i,k ( t ) indicates the first i Candidate combination number k Phase operational margin; s i,k ( t ) indicates the first i Candidate combination number k The state vector of a stage.

[0046] During the model training phase, the runtime graph time-series probabilistic model can be trained by constructing training samples based on at least one of the following: historical black-start test data, scheduling exercise data, transient simulation data, and accident recovery records.

[0047] Its training objectives include success and failure label loss and margin monotonicity constraint, wherein the margin monotonicity constraint is used to ensure that the prior probability of the corresponding stage does not decrease when the stage running margin increases.

[0048] Specifically, to ensure that the model output conforms to engineering principles, a margin consistency constraint is introduced: (9) In the formula, This represents the model training loss function. i This represents the set of parameters to be trained for the runtime graph temporal probability model; Indicates sample n In the stage k The actual execution result is 1 for success and 0 for failure; π n,k Δ represents the prior success probability output by the model. M This represents the margin disturbance. π n,k ( M n,k )and π n,k ( M n,k +Δ M ) represent the model output probabilities with the original margin and the margin increased, respectively; Indicates the weight of the margin monotonicity constraint, [ ] + This indicates that only the positive values ​​are retained. The first term is the success / failure label loss; the second term is the margin monotonicity constraint, used to ensure that the success probability of the corresponding stage does not decrease when the stage margin increases.

[0049] The system is then corrected based on real-time information to obtain the real-time probability of each stage. The real-time information includes at least one of the following: closing feedback, protection action status, energy storage converter current limiting or blocking status, auxiliary machine start-up status, auxiliary machine tripping status, target thermal power unit speed status, excitation status, grid connection status, and minimum stable output achievement status. When the real-time information supports a stage's success, the real-time probability of the corresponding stage is increased; when the real-time information supports a stage's failure, the real-time probability of the corresponding stage is decreased.

[0050] Specifically, during operation, the prior success probability is adjusted based on direct evidence from real-time information: (10) In the formula, p i,k ( t ) indicates the first i The candidate combination of the first k Real-time success probability after real-time information correction at each stage; π i,k ( t ) represents the prior success probability obtained from equation (8); Indicates that the current real-time information supports the first i The candidate combination of the first k The likelihood of success at this stage; Indicates that the current real-time information supports the first i The candidate combination of the first k The likelihood of a phase failure; e To prevent extremely small positive numbers with a denominator of zero; the entire calculation process is as follows Figure 3 As shown. If the no-load charging operation has been completed and the protection has not acted, the success likelihood of the first stage is increased; if the PCS is blocked, the protection trips, or the auxiliary equipment trips, the failure likelihood of the corresponding stage is increased.

[0051] In step S4, based on the real-time probabilities of each stage, and following the sequential logic that air charging, auxiliary machine startup, and thermal power unit startup must all succeed in sequence, the overall black start success rate of each candidate combination is comprehensively calculated. Specifically, the comprehensive evaluation involves: calculating the overall black start success rate according to the sequential relationship between the grid-type energy storage air charging stage, auxiliary machine startup stage, and thermal power unit startup stage; treating completed stages as confirmed successes; triggering failure judgment for corresponding candidate combinations for stages that have clearly failed; and participating in the calculation based on their real-time probabilities for stages that have not yet been completed and have not yet clearly failed.

[0052] Specifically, this embodiment limits the overall black start success rate to the success probability of three consecutive stages: the grid-type energy storage empty-charge stage, the auxiliary equipment start-up stage, and the thermal power unit start-up stage. If the target thermal power unit achieves minimum stable output, the black start process is considered successful, and its overall black start success rate is calculated using the following formula: (11) In the formula, P BS,i ( t ) indicates the first i Candidate combination time t The success rate of black booting throughout the entire process; k =1 indicates the empty charging stage of grid-type energy storage. k =2 indicates the auxiliary machine startup phase. k =3 indicates the start-up phase of the thermal power unit; d i,k ( t ) indicates the first i The candidate combination of the first k Has the phase been successfully completed? f i,k ( t ) indicates the first i The candidate combination of the first k Has the phase clearly failed? p i,k ( t ) indicates the first i The candidate combination of the first k The real-time success probability of each stage after real-time information correction. The calculation logic of Equation (11) is as follows: if any stage has clearly failed, the success rate of the whole process is 0; if a stage has been successfully completed, the stage is treated as a confirmed success and the success rate of the whole process is not reduced again; if a stage has not been completed and has not failed, the stage participates in the calculation of the success rate of the whole process according to its real-time success probability.

[0053] To identify the main risk stages affecting the success rate of candidate combinations, the calculation of the first stage is performed. i The candidate combination of the first k Contribution of risk bottlenecks to each unfinished stage: (12) In the formula, B i,k ( t ) indicates the first i The candidate combination of the first k The contribution of each unfinished stage to the current risk of failure. q This indicates the summation number in the denominator used to iterate through the three stages; p i,k ( t) indicates the first i The candidate combination of the first k Real-time success probability at each stage; ln p i,k ( t The larger the value, the more likely that this stage is to become a risk bottleneck; e To prevent extremely small positive numbers with a denominator of zero, the flowchart of the whole process black start success rate calculation and risk bottleneck identification model is as follows: Figure 4 As shown, this index is used to comprehensively evaluate candidate black-start combinations formed after a power outage. The larger the index value, the higher the number of the corresponding candidate combinations. k The greater the contribution of a stage to the current risk of black start failure.

[0054] Example 2 In a typical embodiment of the present invention, a full-process black start success rate evaluation system based on grid-type energy storage is provided, comprising: Data acquisition module: used to acquire real-time running data of each candidate black start combination, and obtain black start running status information through multi-source confidence correction; Phase Division and Margin Calculation Module: This module is used to divide the entire black start process into multiple phases and construct the runtime margin for each phase. The probability evaluation module is used to process the black start running status information and the running margin of each stage through the running graph time series probability model to obtain the prior probability of each stage; and to correct it according to the real-time information to obtain the real-time probability of the stage. Comprehensive evaluation module: Used to comprehensively evaluate the success rate of black start of each candidate combination throughout the entire process based on the real-time probability of each stage and the serial relationship of each stage.

[0055] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the method of Embodiment 1.

[0056] Example 4 The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method of Embodiment 1.

[0057] The steps and methods involved in Embodiments 3 and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0058] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0060] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for evaluating the success rate of black start-up throughout the entire process of grid-based energy storage, characterized in that, include: Real-time running data of each candidate black start combination is obtained, and black start running status information is obtained through multi-source credibility correction; The multi-source credibility correction determines the credibility of each data source based on data timeliness, communication quality, reasonableness of exceeding limits, and multi-source consistency, and performs weighted correction on multi-source data with the same state quantity based on the credibility. The entire black start process is divided into multiple stages, and the operating margin for each stage is constructed: the entire black start process is divided into the first stage, the second stage, and the third stage; the operating margin for the first stage is constructed based on the current carrying capacity and voltage reactive power regulation capacity of the energy storage converter; the operating margin for the second stage is constructed based on the power energy support capacity and the dynamic quality of the plant service bus. The remaining support time of energy storage and the start-up integrity of the target unit are used to construct the third stage of operation margin; the black start operation status information and the operation margin of each stage are processed through the operation diagram time series probability model to obtain the prior probability of each stage: the grid-type energy storage, the power station bus along the start-up channel and the target thermal power unit are used as nodes, and the connection relationship of lines, transformers and switches in the start-up channel is used as edges to construct the operation diagram time series probability model; Based on the black start running status information and the running margin of each stage, the running graph time-series state features of the candidate black start combination are extracted, and the prior probability of each stage is obtained based on the running graph time-series state features; and the real-time probability of each stage is obtained by correcting it according to the real-time information. Based on the real-time probability of each stage, the success rate of the entire black start of each candidate combination is comprehensively evaluated according to the series relationship of each stage. The comprehensive evaluation is specifically as follows: the success rate of the entire black start is calculated according to the series relationship of the grid-type energy storage empty charging stage, the auxiliary machine start-up stage and the thermal power unit start-up stage; for the completed stage, it is treated as a confirmed success; for the stage that has been clearly failed, the failure judgment of the corresponding candidate combination is triggered; for the stage that has not yet been completed and has not been clearly failed, it participates in the calculation according to the real-time probability.

2. The method for evaluating the success rate of black start-up throughout the entire process based on grid-type energy storage as described in claim 1, characterized in that, The candidate black start combination includes grid-type energy storage, a start-up channel, and a target thermal power unit; The real-time operation data includes at least one of the following: dispatch-side operation data, station-end equipment status, energy storage-side operation data, and unit-side startup status.

3. The method for evaluating the success rate of black start-up throughout the entire process based on grid-type energy storage as described in claim 1, characterized in that, The runtime time-series probability model is trained based on at least one of historical black-start test data, scheduling exercise data, transient simulation data, and accident recovery records. Its training objectives include success and failure label loss and margin monotonicity constraints.

4. A system for evaluating the success rate of black start-up throughout the entire process based on grid-connected energy storage, employing the method for evaluating the success rate of black start-up throughout the entire process based on grid-connected energy storage as described in any one of claims 1-3, characterized in that... include: Data acquisition module: used to acquire real-time running data of each candidate black start combination, and obtain black start running status information through multi-source confidence correction; Phase Division and Margin Calculation Module: This module is used to divide the entire black start process into multiple phases and construct the runtime margin for each phase. The probability evaluation module is used to process the black start running status information and the running margin of each stage through the running graph time series probability model to obtain the prior probability of each stage; and to correct it according to the real-time information to obtain the real-time probability of the stage. Comprehensive evaluation module: Used to comprehensively evaluate the black start success rate of each candidate combination throughout the entire process based on the real-time probability of each stage and the serial relationship of each stage.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1-3.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1-3.

Images