Black start control system and method for gas-steam combined cycle unit energy storage

Abstract

This invention relates to the field of power system control and discloses a black start control system and method for gas-steam combined cycle units with energy storage. The system utilizes a data acquisition module to collect real-time status data from the combined cycle unit, energy storage system, and power grid. A feature extraction module jointly processes the unit and energy storage data to generate a joint feature vector for the unit and energy storage, and generates a power grid environment feature vector for the power grid data. A source-side status analysis module analyzes the source side based on this data to derive an energy supply-demand balance index. A grid-side status analysis module assesses the power grid to obtain a power grid quality assessment index. A collaborative calculation module performs collaborative analysis of both data to calculate a comprehensive black start success coefficient. A black start success matching module matches this coefficient with historical operating conditions to obtain a predicted success probability value. A control execution module executes the strategy accordingly to achieve precise control, dynamically adapt to complex operating conditions, predict start-up results in advance, avoid source-grid mismatch, effectively shorten black start time, and improve power supply reliability and grid resilience.

Description

Technical Field

[0001] This invention relates to the field of power system control technology, and more specifically to a black start control system and method for energy storage in gas-steam combined cycle units. Background Technology

[0002] As a key supporting power source for the power grid, the black start capability of gas-steam combined cycle units is crucial for the rapid recovery of the system after a major power outage. Currently, diesel generators or auxiliary battery banks are commonly used as the starting power source for these units. However, the use of energy storage systems to replace traditional auxiliary power sources has become a development trend. By using preset fixed control logic or manual commands, the unit can gradually complete each process from startup, speed-up, grid connection to load operation, thereby achieving the reconstruction and restoration of the local power grid.

[0003] However, the existing technology still has the following drawbacks: Firstly, traditional black start control usually relies on manually preset fixed control logic or static thresholds. This method lacks flexibility and cannot be dynamically adjusted according to the actual health status of the unit, the remaining energy storage capacity, and the real-time load fluctuations of the power grid, which can easily lead to improper control under complex or non-standard operating conditions. Secondly, existing control systems can usually only passively respond to faults or execute instructions step by step, lacking the ability to predict the final result during the startup process. This means that operators often cannot know in advance whether the current startup path will be successful, and can only take remedial measures after startup failure or parameter exceeding limits. This increases the risk of black start failure and delays the valuable time for power grid restoration. Third, traditional control methods often treat generating units and energy storage as independent entities, controlling them separately from the grid environment and setting their own protection thresholds. This approach ignores the strong coupling relationship between energy supply and demand balance and grid environmental quality, which can easily lead to mismatch problems such as the source side having the ability to output but the grid side not receiving it (e.g., substandard power quality) or the grid side having high demand but insufficient support from the source side (e.g., virtual electricity from energy storage). Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a black start control system and method for gas-steam combined cycle units with energy storage, so as to solve the problems existing in the background art.

[0005] This invention provides the following technical solution: a black start control system for energy storage in a gas-steam combined cycle unit, comprising: Data acquisition module: used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data; Feature extraction module: used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data; Source-side state analysis module: Used to perform matching analysis on the source-side state of combined cycle units and energy storage systems based on the generated unit-energy storage joint feature vector, and obtain the energy supply and demand balance index; Grid-side status analysis module: It is used to evaluate and analyze the electrical operating environment of the power grid based on the generated power grid environment feature vector, and obtain the power grid quality assessment index to evaluate the current environmental quality of the power grid; Collaborative calculation module: used to perform collaborative analysis on the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start; Black Start Success Matching Module: This module is used to match and analyze the calculated comprehensive success coefficient of black start with the historical operating condition feature data in the preset historical operating condition database to obtain the predicted success probability value. Control execution module: Executes the corresponding control execution strategy based on the predicted success probability value.

[0006] Preferably, the data acquisition module collects real-time data on the gas turbine speed, shaft torque, exhaust temperature, and steam turbine speed and steam pressure from the source side via sensor interfaces deployed in the gas turbine control unit, steam turbine monitoring system, battery management system, and grid-connected monitoring device, serving as combined cycle unit operating status data; it also collects the state of charge and port voltage and current of the energy storage system, serving as energy storage system status data; and it collects the grid bus voltage amplitude, frequency deviation, and equivalent impedance parameters from the grid side, serving as grid status data. The collected combined cycle unit operating status data, energy storage system status data, and grid status data are subjected to analog-to-digital conversion and moving average filtering to remove high-frequency noise, and are then spatiotemporally aligned and normalized based on timestamps.

[0007] Preferably, the feature extraction module uses principal component analysis algorithm to perform multi-dimensional spatial mapping on the combined cycle unit operating status data and energy storage system status data, extracts key time-domain features characterizing the unit's start-up capability and frequency-domain features of the energy storage power response, and performs feature coupling through the cross-correlation coefficient between the unit speed and the energy storage port voltage to construct a unit-energy storage joint feature vector reflecting the source-side energy supply and demand coordination characteristics; Meanwhile, the topology identification algorithm is used to perform node state-space analysis on the bus voltage amplitude, frequency deviation and equivalent impedance parameters in the power grid state data, extract the voltage fluctuation mode and frequency stability characteristics of the power grid, and generate a power grid environment feature vector for quantifying the power grid's environmental support capability.

[0008] Preferably, the specific analysis process of the source-side state analysis module includes: A source-side energy supply and demand matching model is constructed. The power supply and demand features in the joint feature vector of the unit energy storage are mapped to the fuzzy evaluation space using a preset membership function. The Euclidean distance between the current feature vector and the preset standard start-up state vector is calculated, and dynamic correction is performed by combining the power supply and demand integral difference within the time-domain sliding window. An energy supply and demand balance index reflecting the energy supply margin of the source side is generated.

[0009] Preferably, the specific analysis process of the network-side state analysis module includes: A power grid environment assessment model based on node state space is constructed. The power grid environment feature vector is mapped to a multi-dimensional quality evaluation space using a preset topology weight matrix. The weighted Mahalanobis distance between the power grid environment feature vector and the ideal stable benchmark vector is calculated by using the power grid voltage fluctuation mode features and frequency stability features contained in the weighted fusion vector. A time-domain decay factor is introduced to dynamically suppress the influence of power grid transient disturbances, resulting in a power grid quality assessment index. The power grid quality assessment index is used to quantitatively characterize the power grid's black start support capability.

[0010] Preferably, the collaborative analysis of the collaborative computing module includes: A source-grid collaborative evaluation model is constructed. Based on the different stages of the black start process, the weight allocation coefficients of the energy supply and demand balance index and the power grid quality assessment index are dynamically determined. The energy supply and demand balance index and the power grid quality assessment index are collaboratively weighted and calculated using a weighted fusion algorithm to obtain an initial comprehensive evaluation value. The initial comprehensive evaluation value is then corrected and calculated with preset start-up risk correction parameters to generate a comprehensive success coefficient for black start, which is used to quantitatively characterize the overall feasibility of black start.

[0011] Preferably, the matching analysis of the successful black start matching module includes: The system calls a preset historical operating condition database, extracts historical operating condition feature vectors that match the current black start process stage, calculates the weighted similarity between the comprehensive success coefficient of the black start and the historical success coefficients contained in the historical operating condition feature vectors, retrieves the target similar operating condition from the historical operating condition database according to the maximum similarity principle, and maps the actual start result corresponding to the target similar operating condition to the predicted success probability value in the current state.

[0012] Preferably, the control execution module constructs a multi-level control strategy response mechanism, compares and analyzes the predicted success probability value with a preset startup probability threshold range level by level, and identifies the security level of the current black start state based on the landing position of the predicted success probability value within the threshold range; wherein, the startup probability threshold range includes a first startup probability threshold and a second startup probability threshold set sequentially from small to large. When the predicted success probability is lower than the first activation probability threshold, the current black start state is determined to be at a high risk level; when the predicted success probability is between the first activation probability threshold and the second activation probability threshold, the current black start state is determined to be at a medium risk level; when the predicted success probability is higher than the second activation probability threshold, the current black start state is determined to be at a low risk level. When identified as a high-risk level, a strong energy storage support strategy is generated, outputting control commands to increase the power output of the energy storage converter to compensate for grid shortages and limiting the rate of increase in the gas turbine fuel regulating valve opening. When identified as a medium-risk level, a smooth regulation strategy is generated, outputting control commands to maintain the power output of the energy storage converter while fine-tuning the gas turbine fuel regulating valve opening to correct energy supply and demand deviations. When identified as a low-risk level, a standard start-up strategy is generated, outputting control commands to adjust the gas turbine fuel regulating valve opening according to a preset curve and controlling the energy storage converter to gradually shut down. By executing the above control commands, closed-loop feedback control of the black start process is achieved.

[0013] To achieve the above objectives, the present invention provides the following technical solution: a black start control method for energy storage in a gas-steam combined cycle unit. Using the aforementioned black start control system for energy storage in a gas-steam combined cycle unit, the method includes the following steps: Step 1: Used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data; Step 2: Used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data; Step 3: Based on the generated combined cycle unit and energy storage joint feature vector, perform matching analysis on the source-side state of the combined cycle unit and energy storage system to obtain the energy supply and demand balance index; Step 4: Based on the generated power grid environment feature vector, the electrical operating environment on the power grid side is evaluated and analyzed to obtain the power grid quality assessment index, which is used to assess the current environmental quality of the power grid. Step 5: Used to perform a collaborative analysis of the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start; Step 6: This step involves matching the calculated black start comprehensive success coefficient with historical operating condition feature data in a preset historical operating condition database to obtain a predicted success probability value. Step 7: Execute the corresponding control execution strategy based on the predicted success probability value.

[0014] The technical effects and advantages of this invention are as follows: (1) The unit health and energy storage status are captured in real time through the data acquisition module and the feature extraction module, and the real-time black start comprehensive success coefficient is generated by the collaborative computing module. This breaks the constraints of fixed logic and realizes the dynamic adjustment of the control strategy according to the actual working conditions, effectively solving the problem of improper control of traditional technology under complex non-standard working conditions.

[0015] (2) By setting up a black start success matching module, the system uses the historical working condition database to perform matching analysis on the calculated comprehensive success coefficient to output the success probability prediction value, giving the system the ability to predict the start result in advance, thus solving the risk that the existing technology cannot predict the start path and can only passively remedy after failure.

[0016] (3) The energy supply and demand balance index is calculated by the source-side state analysis module and the power grid quality assessment index is calculated by the grid-side state analysis module. Multi-dimensional collaborative analysis is carried out through the collaborative calculation module to fully consider the source-grid coupling relationship, thereby accurately matching the source-side output and the grid-side demand, solving the mismatch problem caused by the source-grid separation control in traditional technology. Attached Figure Description

[0017] Figure 1 This is a system structure block diagram of the present invention.

[0018] Figure 2 This is a diagram illustrating the method steps of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. In addition, the forms of the various structures described in the following embodiments are merely illustrative. The gas-steam combined cycle unit energy storage black start control system and method involved in the present invention are not limited to the structures described in the following embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 The embodiment shown provides a black start control system for a gas-steam combined cycle unit with energy storage, including: Data acquisition module: used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data.

[0021] In this embodiment, the data acquisition module collects real-time data on the gas turbine speed, shaft torque, exhaust temperature, and steam turbine speed and steam pressure from the source side via sensor interfaces deployed in the gas turbine control unit, steam turbine monitoring system, battery management system, and grid-connected monitoring device, serving as combined cycle unit operating status data; it also collects the state of charge and port voltage and current of the energy storage system, serving as energy storage system status data; and it collects the grid bus voltage amplitude, frequency deviation, and equivalent impedance parameters from the grid side, serving as grid status data. The collected combined cycle unit operating status data, energy storage system status data, and grid status data are subjected to analog-to-digital conversion and moving average filtering to remove high-frequency noise, and are then spatiotemporally aligned and normalized based on timestamps.

[0022] Specifically, analog-to-digital conversion and moving average filtering refer to: converting analog voltage and current signals collected by sensors into digital sequences at a preset sampling frequency (e.g., 2000 times per second), and then smoothing the digital sequences using a five-point moving average algorithm to remove high-frequency electromagnetic interference noise from the signals; spatiotemporal alignment refers to: given that source-side unit data, energy storage data, and grid-side data come from different acquisition devices, resulting in clock synchronization issues, the system uses the Network Time Protocol (NTP) to calibrate the timestamps of each data source, and uniformly interpolates and resamples data from different acquisition frequencies (e.g., millisecond-level electrical quantities and second-level thermal quantities) to a 100-millisecond time base, constructing a multi-dimensional data matrix at the same moment; and normalization refers to: using the Min-Max normalization method to map data of different dimensions and orders of magnitude to... Within the interval, the calculation formula is: ,in These are the original sampled values. and These are the preset upper and lower limits for the parameter, respectively, thereby eliminating calculation deviations caused by differences in magnitude.

[0023] Feature extraction module: used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data.

[0024] In this embodiment, the feature extraction module uses principal component analysis algorithm to perform multi-dimensional spatial mapping on the combined cycle unit operating status data and energy storage system status data, extracts key time-domain features characterizing the unit's start-up capability and frequency-domain features of the energy storage power response, and performs feature coupling through the cross-correlation coefficient between the unit speed and the energy storage port voltage to construct a unit-energy storage joint feature vector reflecting the source-side energy supply and demand coordination characteristics. Meanwhile, the topology identification algorithm is used to perform node state-space analysis on the bus voltage amplitude, frequency deviation and equivalent impedance parameters in the power grid state data, extract the voltage fluctuation mode and frequency stability characteristics of the power grid, and generate a power grid environment feature vector for quantifying the power grid's environmental support capability.

[0025] It should be specifically explained that the execution process of the principal component analysis algorithm is as follows: First, the collected combined cycle unit operating status data and energy storage system status data are constructed into an original data matrix containing multiple sample points and multidimensional variables, and the matrix is ​​standardized to eliminate dimensional differences; then, the covariance matrix of the standardized matrix is ​​calculated, and the eigenvalues ​​and corresponding eigenvectors of the covariance matrix are solved; the eigenvectors are sorted according to the magnitude of the eigenvalues, and the eigenvectors corresponding to the first few eigenvalues ​​are selected as principal components, so that their cumulative variance contribution rate reaches a preset threshold (such as 85%), thereby mapping the high-dimensional data to a low-dimensional space and extracting the key time-domain features characterizing the unit's start-up capability and the frequency-domain features of the energy storage power response.

[0026] The specific calculation process of feature coupling is as follows: Select the unit speed sequence and the energy storage port voltage sequence within the same time window, calculate the cross-correlation function of the two sequences under different time delays. This function is obtained by calculating the sum of the products of the deviations of the two sequences from their respective means and dividing it by the product of the standard deviations of the two sequences. Traverse all possible time delays, find the correlation coefficient corresponding to the maximum cross-correlation function value, and use it as the coupling index between unit speed and energy storage port voltage. Combine this index with the extracted principal component features to construct the unit-energy storage joint feature vector.

[0027] The topology identification algorithm and the generation process of the power grid environment feature vector are as follows: The node admittance matrix of the power grid is constructed using the collected bus voltage amplitude, phase, and line equivalent impedance parameters; the characteristic equation of the system is solved by eigenvalue decomposition of the node admittance matrix to determine the oscillation modes of the power grid; the voltage fluctuation modes of the power grid are quantitatively analyzed based on the participation factors of each node in different modes; simultaneously, the collected frequency deviation data are statistically analyzed to calculate the mean and variance of the frequency deviation to evaluate frequency stability characteristics; finally, the extracted voltage fluctuation mode parameters are combined with the frequency stability characteristic parameters to generate a power grid environment feature vector used to quantify the power grid's environmental support capability.

[0028] Source-side state analysis module: Based on the generated combined cycle unit and energy storage joint feature vector, it performs matching analysis on the source-side state of the combined cycle unit and energy storage system to obtain the energy supply and demand balance index.

[0029] In this embodiment, the specific analysis process of the source-side state analysis module includes: A source-side energy supply and demand matching model is constructed. The power supply and demand features in the joint feature vector of the unit energy storage are mapped to the fuzzy evaluation space using a preset membership function. The Euclidean distance between the current feature vector and the preset standard start-up state vector is calculated, and dynamic correction is performed by combining the power supply and demand integral difference within the time-domain sliding window. An energy supply and demand balance index reflecting the energy supply margin of the source side is generated.

[0030] It should be specifically explained that constructing the source-side energy supply and demand matching model includes: First, setting a fuzzy evaluation space to characterize the energy supply and demand state, which contains three fuzzy subsets: "severely insufficient," "critically balanced," and "abundant." Then, using a preset half-trapezoidal membership function, the power supply and demand features in the unit's energy storage joint feature vector are mapped to this fuzzy evaluation space to obtain the fuzzy membership degrees of the features. Second, defining a preset standard startup state vector. (This vector is composed of the optimal eigenvalues ​​from historical successful start-up conditions), and will be the real-time generated joint eigenvector of the unit and energy storage. and The input distance calculation formula is used to calculate the Euclidean distance between the two vectors. This distance is obtained by calculating the sum of the squares of the differences between the corresponding feature elements in the two vectors and taking the arithmetic square root, which is used to quantify the degree of deviation between the current state and the standard state. Finally, within a time-domain sliding window (for example, setting the time window to 10 seconds), the difference between the actual output power integral value of the energy storage system and the power integral value required for the combined cycle unit to start up is calculated, which is the supply and demand power integral difference. This difference is used as a correction factor to weight and correct the aforementioned Euclidean distance, generating the final energy supply and demand balance index that reflects the energy supply margin on the source side. The larger the index value, the more sufficient the energy supply on the source side, and the smaller the value, the more unbalanced the supply and demand.

[0031] Grid-side status analysis module: It is used to evaluate and analyze the electrical operating environment of the grid side based on the generated grid environment feature vector, and obtain the grid quality assessment index to evaluate the current environmental quality of the grid.

[0032] In this embodiment, the specific analysis process of the network-side state analysis module includes: A power grid environment assessment model based on node state space is constructed. The power grid environment feature vector is mapped to a multi-dimensional quality evaluation space using a preset topology weight matrix. The weighted Mahalanobis distance between the power grid environment feature vector and the ideal stable benchmark vector is calculated by using the power grid voltage fluctuation mode features and frequency stability features contained in the weighted fusion vector. A time-domain decay factor is introduced to dynamically suppress the influence of power grid transient disturbances, resulting in a power grid quality assessment index. The power grid quality assessment index is used to quantitatively characterize the power grid's black start support capability.

[0033] Specifically, firstly, a preset topology weight matrix W is set according to the importance of each node in the black-start path of the power grid. The diagonal elements of this matrix correspond to the weight coefficients of each node. The power grid environmental feature vector is multiplied by the topology weight matrix to achieve the mapping from the vector to the multidimensional quality evaluation space. Secondly, an ideal stable reference vector is determined. (i.e., the characteristic state where the grid voltage has no fluctuations and the frequency deviation is zero), and calculate the weighted Mahalanobis distance between the grid environment characteristic vector and the ideal stable reference vector. This calculation is obtained by taking the transpose of the difference vector, the product of the inverse covariance matrix and the difference vector itself, to comprehensively consider the correlation and fluctuation amplitude of voltage fluctuation mode characteristics and frequency stability characteristics; finally, a time-domain attenuation factor is introduced. where e is the natural constant. Here, t is a preset attenuation constant, and t is the time interval from the current moment. This factor is used to weight and penalize historical power grid transient disturbance data to reduce the impact of past disturbances on the current assessment, thereby enabling the calculation through the formula. The power grid quality assessment index is calculated, where The closer the exponent value is to 1, the stronger the black start support capability of the power grid.

[0034] Collaborative calculation module: used to perform collaborative analysis on the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start.

[0035] In this embodiment, the collaborative analysis of the collaborative computing module includes: A source-grid collaborative evaluation model is constructed. Based on the different stages of the black start process, the weight allocation coefficients of the energy supply and demand balance index and the power grid quality assessment index are dynamically determined. The energy supply and demand balance index and the power grid quality assessment index are collaboratively weighted and calculated using a weighted fusion algorithm to obtain an initial comprehensive evaluation value. The initial comprehensive evaluation value is then corrected and calculated with preset start-up risk correction parameters to generate a comprehensive success coefficient for black start, which is used to quantitatively characterize the overall feasibility of black start.

[0036] It should be specifically explained that, firstly, based on the preset stage division rules of the black start process (such as divided into the initial start-up stage, acceleration stage, and pre-grid connection stage), a weight allocation function is defined for each stage. For example, in the initial start-up stage, a higher weight coefficient is assigned to the energy supply and demand balance index. (For example, a value of 0.7) assigns a lower weighting coefficient to the power grid quality assessment index. (e.g., a value of 0.3), and satisfies Subsequently, the initial comprehensive evaluation value was calculated using a weighted summation algorithm. That is, by combining the energy supply and demand balance index with weighting coefficients Multiplication, power grid quality assessment index and weighting coefficient The results are obtained by multiplying and then summing; finally, the preset startup risk correction parameters are obtained. (This parameter is determined based on ambient temperature and humidity and equipment operating time), using a correction formula. Corrective calculations are performed to generate a comprehensive success coefficient for black startup, which is used to quantify the overall feasibility of black startup. The closer the coefficient value is to 1, the higher the feasibility of a successful black start.

[0037] Black Start Success Matching Module: This module is used to match and analyze the calculated comprehensive success coefficient of black start with the historical operating condition feature data in the preset historical operating condition database to obtain the predicted success probability value.

[0038] In this embodiment, the matching analysis of the successful black start matching module includes: The system calls a preset historical operating condition database, extracts historical operating condition feature vectors that match the current black start process stage, calculates the weighted similarity between the comprehensive success coefficient of the black start and the historical success coefficients contained in the historical operating condition feature vectors, retrieves the target similar operating condition from the historical operating condition database according to the maximum similarity principle, and maps the actual start result corresponding to the target similar operating condition to the predicted success probability value in the current state.

[0039] It should be specifically explained that, firstly, based on the specific stage of the current black start process (such as the gas turbine ignition stage, acceleration stage, or grid connection stage) as the index condition, historical operating condition datasets belonging to the same stage are selected from the preset historical operating condition database; secondly, a weighted similarity calculation formula is set, and the calculated comprehensive success coefficient of the black start is used to... Historical success coefficients contained in historical operating condition feature vectors The comparison is performed using a cosine similarity algorithm, which calculates the dot product of two coefficient vectors divided by the product of their moduli, or calculates the similarity based on the distance after normalization of the coefficient values, thus quantifying the closeness between the current state and each historical state. Finally, based on the principle of maximum similarity, the record with the highest similarity value is retrieved from the selected historical work condition dataset as the target similar work condition. The actual startup result label associated with this target similar work condition in the database is read (e.g., success is marked as 1, failure as 0, or directly recorded historical success rate values). This result label is directly mapped or converted into a success probability prediction for the current state through linear interpolation. .

[0040] Control execution module: Executes the corresponding control execution strategy based on the predicted success probability value.

[0041] In this embodiment, the control execution module constructs a multi-level control strategy response mechanism, compares and analyzes the predicted success probability value with a preset startup probability threshold range level by level, and identifies the security level of the current black start state based on the landing position of the predicted success probability value within the threshold range; wherein, the startup probability threshold range includes a first startup probability threshold and a second startup probability threshold set sequentially from small to large. When the predicted success probability is lower than the first activation probability threshold, the current black start state is determined to be at a high risk level; when the predicted success probability is between the first activation probability threshold and the second activation probability threshold, the current black start state is determined to be at a medium risk level; when the predicted success probability is higher than the second activation probability threshold, the current black start state is determined to be at a low risk level. When identified as a high-risk level, a strong energy storage support strategy is generated, outputting control commands to increase the power output of the energy storage converter to compensate for grid shortages and limiting the rate of increase in the gas turbine fuel regulating valve opening. When identified as a medium-risk level, a smooth regulation strategy is generated, outputting control commands to maintain the power output of the energy storage converter while fine-tuning the gas turbine fuel regulating valve opening to correct energy supply and demand deviations. When identified as a low-risk level, a standard start-up strategy is generated, outputting control commands to adjust the gas turbine fuel regulating valve opening according to a preset curve and controlling the energy storage converter to gradually shut down. By executing the above control commands, closed-loop feedback control of the black start process is achieved.

[0042] It should be specifically noted that the specific settings for the start-up probability threshold range include: setting the first start-up probability threshold to 0.6 (i.e., 60%) and the second start-up probability threshold to 0.85 (i.e., 85%). This value range is based on statistical analysis of historical black start cases and aims to ensure that the system operates within a safety margin. When a high-risk level is determined (the predicted success probability is less than 0.6), the energy storage strong support strategy specifically calculates the current power gap value of the power grid. The output control command sets the power output of the energy storage converter to 100% of its rated capacity or directly equal to it. Simultaneously, digital commands are sent to limit the rate of increase of the gas turbine fuel regulating valve opening to within 0.5% per second to prevent unit overload. When the risk level is determined to be medium (the predicted success probability is between 0.6 and 0.85), the smooth regulation strategy specifically calculates the energy supply and demand deviation through the PI proportional-integral controller, outputs control commands to maintain the power output of the energy storage converter within a fluctuation range of ±5% of the current level, and fine-tunes the opening of the gas turbine fuel regulating valve in steps of 0.1% per second to correct the deviation. When the risk level is determined to be low (the predicted success probability is greater than 0.85%), the standard start-up strategy specifically calls the preset gas turbine start-up rate curve, outputs control commands to control the fuel regulating valve to open according to the curve, and simultaneously sends ramp commands to control the energy storage converter to gradually reduce the output at a rate of 10% of rated power per minute until it stops operating.

[0043] like Figure 2 The embodiment shown provides a black start control method for energy storage in a gas-steam combined cycle unit, including the following steps: Step 1: Used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data; Step 2: Used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data; Step 3: Based on the generated combined cycle unit and energy storage joint feature vector, perform matching analysis on the source-side state of the combined cycle unit and energy storage system to obtain the energy supply and demand balance index; Step 4: Based on the generated power grid environment feature vector, the electrical operating environment on the power grid side is evaluated and analyzed to obtain the power grid quality assessment index, which is used to assess the current environmental quality of the power grid. Step 5: Used to perform a collaborative analysis of the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start; Step 6: This step involves matching the calculated black start comprehensive success coefficient with historical operating condition feature data in a preset historical operating condition database to obtain a predicted success probability value. Step 7: Execute the corresponding control execution strategy based on the predicted success probability value.

[0044] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0045] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A black start control system for energy storage in a gas-steam combined cycle unit, characterized in that, include: Data acquisition module: used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data; Feature extraction module: used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data; Source-side state analysis module: Used to perform matching analysis on the source-side state of combined cycle units and energy storage systems based on the generated unit-energy storage joint feature vector, and obtain the energy supply and demand balance index; Grid-side status analysis module: It is used to evaluate and analyze the electrical operating environment of the power grid based on the generated power grid environment feature vector, and obtain the power grid quality assessment index to evaluate the current environmental quality of the power grid; Collaborative calculation module: used to perform collaborative analysis on the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start; Black Start Success Matching Module: This module is used to match and analyze the calculated comprehensive success coefficient of black start with the historical operating condition feature data in the preset historical operating condition database to obtain the predicted success probability value. Control execution module: Executes the corresponding control execution strategy based on the predicted success probability value.

2. The gas-steam combined cycle unit energy storage black start control system according to claim 1, characterized in that, The data acquisition module collects real-time data on the source side, including gas turbine speed, shaft torque, exhaust temperature, and steam turbine speed and steam pressure, through sensor interfaces deployed in the gas turbine control unit, steam turbine monitoring system, battery management system, and grid-connected monitoring device, as combined cycle unit operating status data; it also collects the state of charge, port voltage, and current of the energy storage system as energy storage system status data; and it collects the grid bus voltage amplitude, frequency deviation, and equivalent impedance parameters on the grid side as grid status data. The collected combined cycle unit operating status data, energy storage system status data, and grid status data are subjected to analog-to-digital conversion and moving average filtering to remove high-frequency noise, and are then spatiotemporally aligned and normalized based on timestamps.

3. The gas-steam combined cycle unit energy storage black start control system according to claim 2, characterized in that, The feature extraction module uses principal component analysis algorithm to perform multi-dimensional spatial mapping on the combined cycle unit operating status data and energy storage system status data, extracts key time-domain features characterizing the unit's start-up capability and frequency-domain features of the energy storage power response, and performs feature coupling through the cross-correlation coefficient between the unit speed and the energy storage port voltage to construct a combined unit-energy storage feature vector reflecting the source-side energy supply and demand coordination characteristics. Meanwhile, the topology identification algorithm is used to perform node state-space analysis on the bus voltage amplitude, frequency deviation and equivalent impedance parameters in the power grid state data, extract the voltage fluctuation mode and frequency stability characteristics of the power grid, and generate a power grid environment feature vector for quantifying the power grid's environmental support capability.

4. The gas-steam combined cycle unit energy storage black start control system according to claim 3, characterized in that, The specific analysis process of the source-side state analysis module includes: A source-side energy supply and demand matching model is constructed. The power supply and demand features in the joint feature vector of the unit energy storage are mapped to the fuzzy evaluation space using a preset membership function. The Euclidean distance between the current feature vector and the preset standard start-up state vector is calculated, and dynamic correction is performed by combining the power supply and demand integral difference within the time-domain sliding window. An energy supply and demand balance index reflecting the energy supply margin of the source side is generated.

5. The gas-steam combined cycle unit energy storage black start control system according to claim 3, characterized in that, The specific analysis process of the network-side status analysis module includes: A power grid environment assessment model based on node state space is constructed. The power grid environment feature vector is mapped to a multi-dimensional quality evaluation space using a preset topology weight matrix. The weighted Mahalanobis distance between the power grid environment feature vector and the ideal stable benchmark vector is calculated by using the power grid voltage fluctuation mode features and frequency stability features contained in the weighted fusion vector. A time-domain decay factor is introduced to dynamically suppress the influence of power grid transient disturbances, resulting in a power grid quality assessment index. The power grid quality assessment index is used to quantitatively characterize the power grid's black start support capability.

6. The gas-steam combined cycle unit energy storage black start control system according to claim 5, characterized in that, The collaborative analysis of the collaborative computing module includes: A source-grid collaborative evaluation model is constructed. Based on the different stages of the black start process, the weight allocation coefficients of the energy supply and demand balance index and the power grid quality assessment index are dynamically determined. The energy supply and demand balance index and the power grid quality assessment index are collaboratively weighted and calculated using a weighted fusion algorithm to obtain an initial comprehensive evaluation value. The initial comprehensive evaluation value is then corrected and calculated with preset start-up risk correction parameters to generate a comprehensive success coefficient for black start, which is used to quantitatively characterize the overall feasibility of black start.

7. The gas-steam combined cycle unit energy storage black start control system according to claim 6, characterized in that, The matching analysis of the successful black boot matching module includes: The system calls a preset historical operating condition database, extracts historical operating condition feature vectors that match the current black start process stage, calculates the weighted similarity between the comprehensive success coefficient of the black start and the historical success coefficients contained in the historical operating condition feature vectors, retrieves the target similar operating condition from the historical operating condition database according to the maximum similarity principle, and maps the actual start result corresponding to the target similar operating condition to the predicted success probability value in the current state.

8. The gas-steam combined cycle unit energy storage black start control system according to claim 7, characterized in that, The control execution module constructs a multi-level control strategy response mechanism, compares and analyzes the predicted success probability value with the preset startup probability threshold range level by level, and identifies the security level of the current black start state based on the landing position of the predicted success probability value within the threshold range; wherein, the startup probability threshold range includes a first startup probability threshold and a second startup probability threshold set sequentially from small to large. When the predicted success probability is lower than the first activation probability threshold, the current black start state is determined to be at a high risk level; when the predicted success probability is between the first activation probability threshold and the second activation probability threshold, the current black start state is determined to be at a medium risk level; when the predicted success probability is higher than the second activation probability threshold, the current black start state is determined to be at a low risk level. When identified as a high-risk level, a strong energy storage support strategy is generated, outputting control commands to increase the power output of the energy storage converter to compensate for grid shortages and limiting the rate of increase in the gas turbine fuel regulating valve opening. When identified as a medium-risk level, a smooth regulation strategy is generated, outputting control commands to maintain the power output of the energy storage converter while fine-tuning the gas turbine fuel regulating valve opening to correct energy supply and demand deviations. When identified as a low-risk level, a standard start-up strategy is generated, outputting control commands to adjust the gas turbine fuel regulating valve opening according to a preset curve and controlling the energy storage converter to gradually shut down. By executing the above control commands, closed-loop feedback control of the black start process is achieved.

9. A black start control method for energy storage in a gas-steam combined cycle unit, using the black start control system for energy storage in a gas-steam combined cycle unit as described in any one of claims 1-8, characterized in that, Includes the following steps: Step 1: Used to acquire real-time operating status data of combined cycle units, energy storage system status data, and power grid status data; Step 2: Used to perform joint feature extraction on the acquired combined cycle unit operating status data and energy storage system status data, generate a joint feature vector of unit and energy storage, and generate a grid environment feature vector on the acquired grid status data; Step 3: Based on the generated combined cycle unit and energy storage joint feature vector, perform matching analysis on the source-side state of the combined cycle unit and energy storage system to obtain the energy supply and demand balance index; Step 4: Based on the generated power grid environment feature vector, the electrical operating environment on the power grid side is evaluated and analyzed to obtain the power grid quality assessment index, which is used to assess the current environmental quality of the power grid. Step 5: Used to perform a collaborative analysis of the energy supply and demand balance index and the power grid quality assessment index, and calculate the comprehensive success coefficient of black start; Step 6: This step involves matching the calculated black start comprehensive success coefficient with historical operating condition feature data in a preset historical operating condition database to obtain a predicted success probability value. Step 7: Execute the corresponding control execution strategy based on the predicted success probability value.

Images