Power dispatching optimization method based on optical energy storage

Abstract

The invention relates to the technical field of energy dispatching, in particular to a power dispatching optimization method based on optical energy storage, which comprises the following steps: extracting photovoltaic output and charge state fluctuation based on partitions, synchronously identifying abnormal intervals, analyzing direction matching and energy connectivity, screening abnormal units, and extracting linkage units in combination with load and power utilization characteristics. Evaluating power flow deviation, determining an unbalance level, matching a response list, and outputting an automatic linkage scheduling adjustment instruction set. According to the method, photovoltaic output and energy storage state fluctuation are identified in partitions, the accuracy of capturing an abnormal space-time coupling phenomenon is enhanced, the identification depth of scheduling imbalance hidden danger is improved, regional load distribution and power utilization behavior consistency judgment are combined, dynamic load feature evaluation of abnormal units is completed, the abnormal units are scheduled in a layered mode according to risk levels, and the scheduling efficiency is improved. The scheduling range is accurately adjusted, and the response precision and the handling capacity of the scheduling mechanism to the dynamic difference of the load side under new energy fluctuation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy dispatching, and in particular to a power dispatching optimization method based on solar energy storage. Background Art

[0002] The field of energy dispatch technology involves methods and systems for optimizing the configuration and allocation of power resources. Its core issues include energy coordination between the generation and load sides, stable operation and control of the power grid system, prediction and management of renewable energy access, and the role of energy storage systems in power distribution. This technology achieves the scientific allocation of electricity in both time and space through the unified perception of multi-source energy information, the formulation of dispatch strategies, and the implementation of control measures. The field of energy dispatch technology covers multiple aspects such as energy supply and demand balance prediction, energy transmission and distribution path selection, energy storage participation plan arrangement, and dispatch response mechanism formulation. Each link relies on in-depth analysis of actual power grid operation data and reasonable decision-making to achieve coordinated operation of the entire system and improve resource utilization efficiency.

[0003] Traditional power dispatch optimization methods address the imbalance between power supply and demand in different regions during grid operation by constructing deterministic mathematical models and solving them using linear programming methods to formulate daily dispatch plans. This method uses historical load data, power generation capacity boundaries, and unit operating parameters as a basis to establish objective functions and constraints. A single-objective optimization algorithm is then used to determine the unit combination and output plan to complete the system's basic dispatch arrangements. Traditional methods often employ hourly dispatch methods based on minimum cost objectives and distribute load at fixed time granularity. These methods lack the ability to respond to intermittent energy sources, such as photovoltaic output fluctuations, in real time. Furthermore, the limited involvement of energy storage resources makes it difficult to reflect the dynamic load characteristics of a high proportion of renewable energy access.

[0004] Existing dispatching methods rely on deterministic models and a single optimization objective. They respond slowly in scenarios where load fluctuations change rapidly or renewable energy output fluctuates dramatically. They lack the ability to carefully explore the spatiotemporal coupling characteristics, causing the power configuration plan to lag behind the actual fluctuation state. In areas with large-scale photovoltaic access, energy storage resources fail to fully participate in the regulation process, resulting in overload operation in some areas and idle energy storage in others. Dispatching methods are difficult to match load evolution characteristics with system stability goals. If plans are made based only on fixed-granularity data, it is easy for the response range to be too wide or the adjustment to be insufficient, resulting in reduced operating efficiency in local areas and weakened grid risk warning capabilities. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a power dispatch optimization method based on solar energy storage.

[0006] In order to achieve the above objectives, the present invention adopts the following technical solution: a power dispatch optimization method based on solar energy storage, comprising the following steps: S1: Based on the grid-based PV-storage combined units, the PV output variation curve and the energy storage state-of-charge fluctuation trajectory within the unit are extracted. The synchronous abnormal intervals of output fluctuation amplitude and state-of-charge variation trend are identified on the time axis. The corresponding unit number and spatial location information are extracted to generate a candidate set of abnormal units. S2: Recall the photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set, analyze the direction matching of the two at the unit junction, combine the energy flow connectivity of the unit boundary, screen the units with connectivity and abnormal output fluctuations, and form a layer of suspected scheduling imbalance units; S3: Based on the suspected scheduling imbalance unit layer, extract the load distribution matrix and regional power consumption characteristics within the unit, analyze the load fluctuation pattern and the consistency of power demand, screen and match abnormal units, and obtain a set of scheduling imbalance linkage units; S4: According to the position number of the dispatch imbalance linkage unit set, analyze the grid flow distribution trend corresponding to the unit, evaluate the degree of deviation from the grid benchmark operation curve, calibrate the dispatch imbalance level of the unit according to the deviation amplitude, and output a grid unit dispatch risk level list.

[0007] As a further solution of the present invention, the abnormal unit candidate set includes the output fluctuation unit number, the energy storage charge abnormality point, the unit coordinate mark, and the time series abnormality mark; the suspected scheduling imbalance unit layer includes the output direction abnormal unit mark, the unit boundary energy connectivity unit, the unit boundary matching block, and the abnormal output distribution grid; the scheduling imbalance linkage unit set includes the load abnormality continuous unit, the irregular power demand area, the load mutation overlap area, and the linkage abnormal unit number; the power grid unit scheduling risk level list includes the risk level label, the unit response deviation index, the local flow abnormality index, and the operation offset level.

[0008] As a further solution of the present invention, the step of obtaining the abnormal unit candidate set is specifically as follows: S111: Based on the grid-divided photovoltaic and energy storage combined units, the PV output variation curve and the energy storage state of charge fluctuation trajectory within the unit are extracted. The two types of data within the same unit are compared and analyzed to obtain the trend value of the difference between output and state of charge; S112: Based on the output and state of charge difference trend value, identifying the fluctuation amplitude value in the output change curve and the deviation value of the state of charge fluctuation trajectory, superimposing the two types of values ​​over time, extracting the time intervals in which the fluctuation exceeds the reference range and the deviation value exceeds the set limit, and generating a high-frequency abnormal interval time period set; S113: For the high-frequency abnormal interval time period set, matching the corresponding unit numbers and spatial position information, extracting the unit positions where abnormal signals occur, and generating an abnormal unit candidate set.

[0009] As a further solution of the present invention, the steps for obtaining the suspected scheduling imbalance unit layer are specifically as follows: S211: Retrieving the photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set, extracting the projection trajectory of the two at the unit boundary, identifying the distribution number and aggregation degree of the boundary points in the unit, and obtaining a unit boundary matching map; S212: Based on the unit boundary matching map, filter the boundary areas with a higher-than-average degree of aggregation, compare the spatial boundaries of the overall power grid structure map, identify continuous boundary concentrated areas belonging to the same unit, and obtain the abnormal output direction areas within the unit; S213: Call the abnormal output direction area in the unit, calculate the output direction discrete deviation index, perform integrated analysis on the unit boundary matching, output direction discreteness, load distribution balance and energy storage charging and discharging delay, match the units according to the response blocks in the layer, and form a layer of suspected scheduling imbalance units.

[0010] As a further solution of the present invention, the step of obtaining the scheduling imbalance linkage unit set is specifically: S311: Based on the suspected dispatch imbalance unit layer, extract the load distribution matrix and regional power consumption characteristics of the numbered units in the layer, perform time alignment on the data within the unit, identify the load intensity fluctuation characteristics and the power demand consistency offset characteristics, and obtain the local dispatch response feature set of the power grid; S312: Based on the local dispatch response feature set of the power grid, jointly analyze the load fluctuation pattern and the consistency of power demand within the unit, calculate the load-power coupling feature value, select the unit partitions with the load-power coupling degree in the layer, and establish a load-power coordinated response spatial distribution map; S313: Call the load power consumption coordinated response spatial distribution map, cluster the units that exceed the coordinated identification benchmark in the coupling characteristic value layer, mark the unit codes and coordinates corresponding to the continuous abnormal areas, and obtain the scheduling imbalance linkage unit set.

[0011] As a further solution of the present invention, the steps for obtaining the grid unit dispatch risk level list are specifically as follows: S411: extracting the unit power flow distribution curve under the specified number according to the position number of the scheduling imbalance linkage unit set, performing time uniform processing, identifying the unit time power flow change characteristics, and obtaining the unit power flow abnormal change characteristic set; S412: Based on the unit power flow abnormal change feature set, identify the power flow distribution curve of the reference operation stage, compare the current power flow change sequence with the reference curve, identify the unit power flow deviation level, extract and mark the units whose deviation level exceeds the warning limit, and obtain the deviation sudden increase unit set; S413: Based on the deviation sudden increase unit set, the deviation level value of each unit is bound to the position number in the power grid structure space diagram, sorted by risk level, and output a power grid unit scheduling risk level list.

[0012] As a further embodiment of the present invention, the method further comprises step S5: S5: calling the grid unit dispatch risk level list, identifying the corresponding unit number in the grid function diagram, calling the dispatch response unit list, comparing the response level with the grid protection priority sequence, screening the unit numbers whose response coverage needs to be adjusted, and outputting the automatic linkage dispatch adjustment instruction set; The automatic linkage scheduling adjustment instruction set includes an adjustment target unit number, a response level adjustment item, a protection priority comparison item, and a linkage response trigger type.

[0013] As a further solution of the present invention, the steps for obtaining the automatic linkage scheduling adjustment instruction set are specifically as follows: S511: Calling the grid unit dispatch risk level list, extracting the unit number in the grid function map, mapping the unit risk level value with the regional coordinate boundary, identifying the unit information corresponding to the grid protection level, and generating a grid unit risk distribution map; S512: Based on the grid unit risk distribution map, extract the dispatch response unit number and response level, match the unit risk level with the dispatch response level, identify the unit number with insufficient response coverage, and obtain a grid unit response risk disconnection list; S513: According to the grid unit response risk disconnection list and the level number in the grid protection priority sequence, the key unit numbers for which the response coverage needs to be improved are extracted, and the adjustment control parameters linked with the original dispatching unit are output in sequence, and an automatic linkage dispatching adjustment instruction set is output.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by identifying fluctuations in the timing characteristics of photovoltaic output and energy storage status within the partition range, the accuracy of capturing abnormal spatiotemporal coupling phenomena is enhanced, and the output matching differences in connected areas are identified by analyzing the direction of energy flow, thereby improving the depth of identification of scheduling imbalance hazards. Based on the consistency judgment of regional load distribution and electricity consumption behavior, the correlation evaluation of load dynamic characteristics of abnormal units is realized, and further combined with the deviation analysis of the power grid current trend and the operating benchmark curve, the scheduling anomalies are stratified and clustered according to risk levels, so that the scheduling response coverage of risk units is more in line with the system protection requirements, and accurate linkage adjustment of the scheduling range is achieved, effectively enhancing the response accuracy and handling capabilities of the scheduling mechanism to dynamic differences on the load side under the background of new energy fluctuations. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the main steps of the present invention; Figure 2 This is a flow chart for obtaining an abnormal unit candidate set in the present invention; Figure 3 This is a flowchart for obtaining a layer of a suspected scheduling imbalance unit in the present invention; Figure 4 This is a flow chart for obtaining a set of scheduling imbalance linkage units in the present invention; Figure 5 A flowchart for obtaining a list of grid unit dispatch risk levels in the present invention; Figure 6 This is a flow chart for obtaining the automatic linkage scheduling adjustment instruction set in the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0018] Example 1 See also Figure 1The present invention provides a technical solution: a method for optimizing power dispatch based on solar energy storage, comprising the following steps: S1: Based on the grid-based PV-storage combined units, the PV output variation curve and the energy storage state-of-charge fluctuation trajectory within the unit are extracted. The synchronous abnormal intervals of output fluctuation amplitude and state-of-charge variation trend are identified on the time axis. The corresponding unit number and spatial location information are extracted to generate a candidate set of abnormal units. S2: The photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set are retrieved, and their directional matching at the unit boundary is analyzed. Combined with the energy flow connectivity at the unit boundary, units with good connectivity and abnormal output fluctuations are screened to form a layer of suspected scheduling imbalance units. S3: Based on the suspected unbalanced scheduling unit layer, extract the load distribution matrix and regional power consumption characteristics within the unit, analyze the load fluctuation pattern and the consistency of power demand, screen and match abnormal units, and obtain the set of unbalanced scheduling linkage units; S4: Based on the location number of the dispatch imbalance linkage unit set, analyze the grid flow distribution trend corresponding to the unit, evaluate the degree of deviation from the grid benchmark operation curve, calibrate the dispatch imbalance level of the unit according to the deviation amplitude, and output a grid unit dispatch risk level list; S5: Call the grid unit dispatch risk level list, identify the corresponding unit number in the grid function diagram, call the dispatch response unit list, compare the response level with the grid protection priority sequence, screen the unit numbers that need to adjust the response coverage, and output the automatic linkage dispatch adjustment instruction set.

[0019] The abnormal unit candidate set includes the output fluctuation unit number, energy storage charge abnormality point, unit coordinate mark, and time series abnormality mark. The suspected scheduling imbalance unit layer includes the output direction abnormal unit mark, unit boundary energy connectivity unit, unit boundary matching block, and abnormal output distribution grid. The scheduling imbalance linkage unit set includes load abnormality continuous unit, irregular power demand area, load mutation overlap area, and linkage abnormal unit number. The grid unit scheduling risk level list includes risk level label, unit response deviation index, local flow abnormality index, and operation offset level. The automatic linkage scheduling adjustment instruction set includes adjustment target unit number, response level adjustment item, protection priority comparison item, and linkage response trigger type.

[0020] See also Figure 2 ,The specific steps for obtaining the abnormal unit candidate set are: S111: Based on the grid-divided photovoltaic and energy storage combined units, the PV output variation curve and the energy storage state of charge fluctuation trajectory within the unit are extracted. The two types of data within the same unit are compared and analyzed to obtain the trend value of the difference between output and state of charge; Based on the PV-storage combined units divided by the grid, data extraction operations are performed on each preset PV-storage combined unit. For example, for the PV-storage combined unit "Region A-GS01" in a certain area of ​​the North China Power Grid, which includes a PV power station with a rated power of 50 MW and an energy storage power station with a rated capacity of 20 MWh and a maximum charge and discharge power of 10 MW, the real-time PV power generation output data sequence and energy storage charge state data sequence of this unit within a specified time period (for example, from 08:00 to 18:00 on July 15) are extracted. The photovoltaic power generation output data is recorded at a sampling frequency of once per minute, and the energy storage charge state data is also recorded at a sampling frequency of once per minute. For example, at 08:00, the photovoltaic output is 15 MW and the energy storage charge state is 70%. At 08:01, the photovoltaic output is 15.2 MW and the energy storage charge state is 69.8%. And so on. The photovoltaic power generation output change curve and the energy storage charge state fluctuation trajectory in the same unit are compared and analyzed. The photovoltaic power generation output change curve is calculated by calculating the output between adjacent sampling points. For example, from 15 MW to 15.2 MW, the output change is 0.2 MW. The energy storage state of charge fluctuation trajectory is obtained by calculating the difference in SOC values ​​between adjacent sampling points. For example, from 70% to 69.8%, the state of charge change is -0.2%. For these two change sequences, the absolute difference within each minute is calculated, for example, |0.2-(-0.2)|=0.4. The absolute difference of each minute is accumulated and the average value of each 15 minutes is obtained to form a preliminary difference degree sequence. Further, by The preliminary difference sequence is subjected to exponential smoothing, and the smoothing coefficient is set to 0.2. The setting of this smoothing coefficient is based on historical data analysis. For example, in the operating data of nearly 1,000 sets of photovoltaic storage units, it was found through least squares fitting that a smoothing coefficient of 0.2 can make the difference trend value moderately sensitive to short-term fluctuations, while having good tracking capabilities for long-term trends, thereby obtaining the output and state of charge difference trend value of the photovoltaic storage combined unit within a specified time period. For example, the final trend value is 0.35%.

[0021] S112: Based on the trend value of the difference between output and state of charge, identify the fluctuation amplitude value in the output change curve and the deviation value of the state of charge fluctuation trajectory, superimpose the two types of values ​​over time, extract the time intervals where the fluctuation exceeds the reference range and the deviation value exceeds the set limit, and generate a high-frequency abnormal interval time period set; Based on the trend value of the difference between output and state of charge, the identification of the fluctuation amplitude value is based on the calculation of the sliding window standard deviation of the photovoltaic output data series. For example, using a 10-minute sliding window, if the output standard deviation within a window exceeds the preset baseline fluctuation threshold, the threshold is set to 5% of the average photovoltaic output. The setting of this threshold is based on the statistical analysis of photovoltaic output fluctuations under normal grid operation. For example, under three consecutive months of normal meteorological conditions, through a large sample statistics of historical output data, it was found that 95% of the time the photovoltaic output standard deviation did not exceed 5% of its average value. Therefore, this value is set as the baseline fluctuation threshold. For example, if the average photovoltaic output in a certain period is 40 MW, the baseline fluctuation threshold is 2 MW. When the photovoltaic output standard deviation reaches 3 MW within a 10-minute window, it is considered that the fluctuation exceeds the baseline range. At the same time, the deviation value of the energy storage state of charge fluctuation trajectory is identified. The deviation value is obtained by calculating the absolute difference between the actual energy storage state of charge value and the predicted state of charge value calculated based on the historical average charge / discharge efficiency. For example, if the actual SOC is 65%, the predicted SOC is 65%. OC is 68%, and the deviation is 3%. When this deviation exceeds the preset limit, the limit is set to 2.5% of SOC. For example, 2.5% is set based on the healthy operating range of the energy storage battery and the tolerance of the scheduling strategy. For example, the normal charge and discharge deviation range recommended by the battery manufacturer is usually within + / - 2%. Considering the scheduling flexibility, an additional 0.5% margin is added. For example, when the SOC deviation reaches 3.2% in a certain period of time, it is considered that the deviation exceeds the set limit, and the time when the photovoltaic output fluctuation exceeds the benchmark range is set as 2.5%. The time interval is superimposed with the time interval when the energy storage state of charge deviation exceeds the set limit. For example, if the photovoltaic fluctuation exceeds the range from 09:30 to 09:40, and the energy storage deviation exceeds the limit from 09:35 to 09:45, the superimposed abnormal time interval is 09:35-09:40. All superimposed abnormal time intervals are counted to generate a high-frequency abnormal interval time period set. The standard for determining high frequency is set to 3 or more abnormal intervals within 1 consecutive hour, or 15 or more abnormal intervals within 24 consecutive hours.

[0022] S113: For the high-frequency abnormal interval time period set, matching the corresponding unit number and spatial location information, extracting the unit location where the abnormal signal occurs, and generating an abnormal unit candidate set; For the high-frequency abnormal interval time period set, for example, the identified unit "region A-GS01" has a high-frequency abnormal interval from 10:00 to 11:00, and the unit number and spatial location information corresponding to the abnormal time period set are matched. The unit number is "GS01", and its spatial location information is obtained by reading the coordinate data in the power grid geographic information. For example, the core substation coordinates of the unit are 116.32 east longitude and 39.91 north latitude. The unit location where the abnormal signal occurs is extracted. For example, the geographical location where the abnormality occurs is confirmed to be the "GS01" unit in the North China power grid area. The unit is located near the main load center in the power grid topology, for example, within the power supply range of the regional load center P32 substation. The unit number "GS01" and its coordinates "(116.32E, 39.91N)" and the abnormal time period set are packaged to generate an abnormal unit candidate set.

[0023] See also Figure 3 The specific steps for obtaining the suspected scheduling imbalance unit layer are as follows: S211: Recalling the photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set, extracting the projection trajectory of the two at the unit boundary, identifying the distribution number and aggregation degree of the intersection points in the unit, and obtaining the unit boundary matching map; The information of the unit in the abnormal unit candidate set is called, including the unit number "GS01", and the photovoltaic output direction and energy storage charge and discharge capacity of the unit are extracted. The photovoltaic output direction vector is obtained by calculating the product of the photovoltaic output change rate and its geographical direction (for example, the south direction is 0 degrees and the east direction is 90 degrees). For example, if the output change rate is +0.5 MW / min and the geographical direction is 180 degrees, the direction vector is 0.5×cos(180°). The energy storage charge and discharge capacity is obtained by calculating the energy storage charge and discharge power (charging is positive and discharging is negative). ) and the virtual direction of the battery pack in the spatial layout (for example, pointing inward when charging and pointing outward when discharging). For example, if the charging and discharging power is -5 MW (discharging) and the virtual direction is 90 degrees, then the vector is -5×cos(90°). The projection trajectory of the two at the unit junction is extracted. The unit junction is defined as the transmission line interface point between the photovoltaic storage unit and the adjacent grid unit (for example, another photovoltaic storage unit or a traditional power generation unit). At the junction point, the photovoltaic output direction is calculated based on the angle between the line flow direction and the vector. The projection of the energy storage charge and discharge capacity is used. For example, if the angle between the photovoltaic output direction vector and the line direction at the junction is 30 degrees and the vector magnitude is 0.5, then the projection is 0.5×cos(30°). The projection value is plotted as a trajectory over time to identify the number of junction points distributed in the unit. For example, it is counted that there are 15 electrical junction points between the "GS01" unit and the surrounding 10 power grid units, and the degree of aggregation of the junction points is identified. The degree of aggregation is determined by calculating the spatial distance and electrical connection density of the junction points on the power grid geographical topology map. For example, by calculating the Euclidean distance between each junction point and its nearest neighbor junction point, and grouping and counting the distances, a threshold for high aggregation (greater than the 75th percentile) is set. For example, after counting the aggregation degrees of the junction points of all units, the median is 50 meters, and the threshold for aggregation degrees above the average level is set to less than 50 meters. For example, among the 15 junction points of the "GS01" unit, the nearest neighbor distance of 8 junction points is less than 40 meters, which indicates that their aggregation degree is high. Based on the quantity and aggregation degree information, a unit junction matching map is obtained.

[0024] S212: Based on the unit boundary matching map, filter the boundary areas with higher-than-average aggregation, compare the spatial boundaries of the overall grid structure map, identify the continuous boundary concentration areas belonging to the same unit, and obtain the abnormal output direction areas within the unit; Based on the unit boundary matching map, we screened out boundary areas with higher-than-average aggregation. For example, in the boundary matching map of the "GS01" unit, we identified three boundary points (J1, J2, and J3) in its southern region, all of which were less than 40 meters apart, and much lower than the average distance (75 meters) of all boundary points in the unit. Therefore, we determined that its aggregation was higher than the average level. By comparing the spatial boundaries of the overall structure diagram of the power grid, for example, we retrieved the layout diagram of the lines and substations of the North China Power Grid with voltage levels of 220kV and above, and spatially superimposed the screened high-aggregation boundary areas with the overall structure diagram. , identify continuous and concentrated intersection areas belonging to the same unit. For example, it is found that the aggregated intersection points J1, J2, and J3 form a continuous area that is not separated by unit lines on the overall structure diagram, and all belong to the geographical coverage of the "GS01" unit. For example, this area is geographically located below the main transmission line between the "GS01" unit and the adjacent load center P32 substation, thereby confirming that this area is an abnormal output direction area within the unit. For example, this area is marked as "GS01-Southern Abnormal Zone", and this zone is determined to be an abnormal output direction area within the unit.

[0025] S213: Call the abnormal output direction area within the unit, using the formula: ; Calculate the output direction discrete deviation index, conduct an integrated analysis of unit boundary matching, output direction discreteness, load distribution balance, and energy storage charge and discharge delay, match units based on the response blocks in the layer, and form a layer of suspected dispatch imbalance units; in, Represents the discrete deviation index of the output direction, represents the number of units involved in the analysis, Representative The actual output value of each unit, Represents the average output value of all units, Representative The variance of the original output value of each unit, Representative The directional fluctuation amplitude of each unit in the scheduling process, Representative The influencing factor of the directional coupling between a unit and adjacent units; The output direction anomaly zoning within the unit is called, for example, the obtained "GS01-Southern anomaly zoning", and the output direction discrete deviation index is calculated using the formula, where; The output direction discrete deviation index comprehensively measures the degree of deviation between the output of the PV-storage unit and the average output of its region. It also takes into account the unit's own output fluctuations, directional swings during the scheduling process, and the coupling effect with adjacent units. Its purpose is to identify units whose output behavior is significantly inconsistent with the surrounding environment. Represents the number of units involved in the analysis. For example, in the current analysis scenario, consider the "GS01" unit and its five adjacent solar storage units. ; Representative The actual output value of each unit, in megawatts (MW), is obtained through the real-time monitoring system. For example, unit 1 (GS01): MW; Unit 2: MW; Unit 3: MW; Unit 4: MW; Unit 5: MW; Unit 6: MW; Represents the average output value of all participating analysis units, calculated as The sum divided by , for example, to calculate: ; Representative The variance of the original output value of each unit, in units of This value is obtained by statistical analysis of the unit output data in the past 24 hours. For example, for unit 1 (GS01), the variance of its output in the past 24 hours is ; Representative The directional fluctuation amplitude of each unit during the scheduling process is obtained by monitoring the standard deviation of the unit output direction change after the scheduling instruction is issued. The unit is angle (°) or dimensionless value. For example, when the scheduling requires an increase in output, the deviation between the actual unit output direction and the expected direction is normalized to the range of (0, 1]. For example, for unit 1 (GS01), ; Representative The influence factor of the directional coupling degree between a unit and the adjacent unit is calculated by The Pearson correlation coefficient between the output direction of the unit and the output direction of the adjacent unit is quantified. For example, for unit 1 (GS01), the correlation coefficient with the output direction of the adjacent unit is 0.75. The conversion formula is: , map it to the range (0, 1], then: ; The operation logic in the formula is: First, calculate the absolute deviation between the actual output of each unit and its regional average output , the deviation directly reflects the degree of deviation between the unit output and the overall regional trend; then, a correction factor is calculated ,in and Accumulation reflects the inherent instability of the unit itself (large output fluctuations and large scheduling response swings), and Subtract, indicating that the higher the coupling with the adjacent units, the more consistent their behavior, so a smaller "tolerance" should be given when calculating the deviation (that is, the smaller the denominator, the larger the overall deviation); Finally, the weighted deviation of each unit is accumulated and divided by the number of units involved in the analysis , get the average discrete deviation index .

[0026] The innovation of this indicator is that it takes into account 、 、 These three parameters correct the output deviation so that It not only reflects the simple deviation of unit output from the average, but more importantly, it can identify units with "abnormal" deviations, that is, units whose deviations still appear "outstanding" after considering their own stability and coupling with neighbors, thereby improving the accuracy of scheduling imbalance identification; Example calculation of the correction factor denominator for unit 1 (GS01): Correction factor denominator = ; Calculations for Unit 1 (GS01): ; Next, continue calculating the parameters of additional cells to fill the entire summation term.

[0027] Table 1: Unit parameters involved in the calculation

[0028] As shown in Table 1, the parameter values ​​of the six units involved in the calculation are listed. Based on the above parameter values, the denominator and absolute term of the correction factor of each unit are calculated one by one: Unit 2: Denominator = , ; Unit 3: Denominator = , ; Unit 4: Denominator = , ; Unit 5: Denominator = , ; Unit 6: Denominator = , ; Sum all unit terms: ; Final calculation : ; Further integrated analysis is conducted on unit boundary matching, output direction dispersion, load distribution balance and energy storage charge and discharge delay. The output direction dispersion is directly calculated using the above The load distribution balance is obtained by calculating the root mean square of the deviation between the real-time load of each load point in the unit and the total load. For example, if this value is greater than a preset threshold (for example, 10%), the load is considered to be unbalanced. The energy storage charging and discharging delay is measured by monitoring the time interval from the issuance of the scheduling instruction to the actual response of the energy storage and comparing it with the preset response standard time (for example, 100 milliseconds). If the delay exceeds 50 milliseconds, it is considered that there is a delay. Unit matching is performed according to the response block in the layer. For example, "GS01-Southern Abnormal Zoning" is displayed in the analysis. If the value is high, the load distribution is unbalanced, and the energy storage response is delayed, it will be matched with the abnormal attributes to form a suspected dispatch imbalance unit layer. For example, the "GS01" unit will be marked as a suspected dispatch imbalance unit and highlighted in red on the power grid geographic information system.

[0029] See also Figure 4 ,The specific steps for obtaining the scheduling imbalance linkage unit set are: S311: Based on the suspected dispatch imbalance unit layer, extract the load distribution matrix and regional power consumption characteristics of the numbered units in the layer, perform time alignment on the data within the unit, identify the load intensity fluctuation characteristics and power demand consistency deviation characteristics, and obtain the local dispatch response feature set of the power grid; Based on the suspected scheduling imbalance unit layer, including unit "GS01", the load distribution matrix and regional power consumption characteristics of the numbered units in the layer are extracted. The load distribution matrix is ​​constructed by real-time collection of load data of each user node in unit "GS01" (for example, once every 5 minutes, in kilowatts). For example, in a monitoring period, the load of user 1 is 100kW, the load of user 2 is 150kW, and the load of user 3 is 80kW. The regional power consumption characteristics are obtained by analyzing the historical load data and typical load curves of the area where the unit is located, such as the peak load characteristics of the noon peak on weekdays. The data in the unit is time-aligned to ensure that all time series data such as photovoltaic output, energy storage status, and load data are synchronized at intervals of 5 minutes. For example, if a certain type of data is sampled for 1 minute, it is aggregated to 5-minute intervals by averaging or interpolation to identify load intensity fluctuation characteristics. The load intensity fluctuation characteristics are obtained by calculating the standard deviation of the load data in each 5-minute time slice. For example, if the load in a certain 5-minute slice is from The load fluctuates from 100kW to 120kW and then to 105kW, with a standard deviation of 8.66kW. When this standard deviation exceeds 10% of the average load in that time slice, the load intensity fluctuation is considered significant. For example, if the average load is 100kW, a standard deviation exceeding 10kW is considered significant. The power demand consistency deviation feature is also identified. This feature is obtained by comparing the actual load data with the historical average power consumption pattern or predicted power consumption pattern of the same type of area. For example, if the actual load curve is continuously higher than the predicted load curve during a certain period of time and the deviation exceeds 5% of the predicted value, for example, the predicted load is 1000kW and the actual load is 1080kW, with a deviation of 8%, then the consistency deviation feature is considered significant. For example, in the "GS01" unit, the actual load continuously exceeds the predicted value by more than 8% between 2:00 PM and 4:00 PM on a certain day. The local grid dispatch response feature set is obtained, which contains the load intensity fluctuation data and power demand consistency deviation data of the "GS01" unit during this period.

[0030] S312: Based on the local dispatch response feature set of the power grid, a joint analysis is performed on the load fluctuation pattern and the consistency of power demand within the unit, using the formula: ; Calculate the load-power coupling characteristic value, screen the unit partitions of the load-power coupling degree in the layer, and establish the load-power coordinated response spatial distribution map; in, Represents the load power coupling characteristic value, Represents the number of time slices involved in the analysis, Representative Actual load data of the unit within a time slice, Representative The unit predicts the electricity demand data within the time slice, Representative Actual load data of the unit within a time slice, Representative The unit predicts the electricity demand data within the time slice, Representative Unit load response delay adjustment factor within a time slice; Based on the local dispatch response feature set of the power grid, for example, the load intensity fluctuation characteristics and power demand consistency deviation characteristics of the "GS01" unit, a joint analysis of the load fluctuation pattern and power demand consistency within the unit is performed, and the load-power coupling characteristic value is calculated using the following formula: in, : Load-power coupling characteristic value, which measures the degree of matching between the actual load and the predicted power demand of the photovoltaic and energy storage combined unit in a specific time period. The larger the value, the worse the matching and the lower the coupling degree, and vice versa. : The number of time slices involved in the analysis. For example, if each time slice is 5 minutes, then the total number of time slices in a day is: ; : No. The actual load value (MW) of each time slice is collected in real time by the SCADA system; : No. The predicted power demand value (MW) for each time slice is generated by the short-term load forecasting system; : The standard deviation of the actual load data in the time slice (MW), reflecting the degree of load fluctuation; : The standard deviation of the predicted electricity demand in this time slice (MW), reflecting the forecast volatility; : The load response delay factor of the time slice, which measures the load response lag to the dispatch instruction or electricity price signal. For example, if the average response delay is 20 seconds, then ; Operation logic: Calculate the load forecast deviation for each time slice: , calculate the weighting factor: , multiply the deviation by the weighting factor and sum them to get the overall coupling eigenvalue The innovation of this indicator is the introduction of load fluctuation ( ), predicting volatility ( ) and response delay ( ), thereby more comprehensively reflecting the actual coupling relationship between load and electricity consumption; Assume that the parameters for a certain time slice (for example, 10:00–10:05) are as follows: MW, MW, MW, MW, ; Calculate the weighting factors: ; The contribution value of this time slice is: ; Assume another time slice (e.g. 15:00–15:05) with the parameters: MW, MW, MW, MW, ; Weighting factors: ; The contribution value of this time slice is: ; If we only consider the above two time slices: ; The cell partition of the load coupling degree in the filter layer, for example, the calculated Sort the values ​​and set the coupling eigenvalues The high coupling threshold is 1000. For example, this threshold is obtained through historical scheduling experience and system simulation. If it is lower than this value, it means the coupling is good, and if it is higher than this value, it means the coupling is poor. If the value exceeds this threshold, it is considered that the load-power coupling is low, and a load-power coordinated response spatial distribution map is established. For example, on the power grid map, Units with values ​​below 1000 are marked in green, and units with values ​​above 1000 are marked in yellow, which intuitively shows the load power coordinated response of each unit.

[0031] S313: Call the load power consumption coordinated response spatial distribution map, cluster the units that exceed the coordinated identification benchmark in the coupling characteristic value layer, mark the unit codes and coordinates corresponding to the continuous abnormal areas, and obtain the scheduling imbalance linkage unit set; Call the load power consumption coordinated response spatial distribution map, where the "GS01" unit The value is calculated to be 61.02 (assuming this is a simplified example, the actual calculated value is 1250). Cluster the cells in the coupled eigenvalue layer that exceed the collaborative identification benchmark. The collaborative identification benchmark is set to >1000. This benchmark value is obtained through a large amount of historical data analysis and expert experience calibration. For example, in the operating data of the past year, when the unit When the value exceeds 1000, the probability of scheduling imbalance in the subsequent scheduling of the unit increases significantly to more than 80%, so this value is set as the benchmark and all Cells with values ​​greater than 1000, for example, the "GS01" cell The value is 1250, which meets this condition. Then, the units are clustered according to their geographical locations. For example, the "GS01" unit is clustered with the adjacent "GS02" unit and "PV03" unit. If the values ​​of all the units exceed 1000 and are geographically adjacent to each other (for example, the spatial distance is less than 5 kilometers), the three units are clustered into one category, and the unit codes and coordinates corresponding to the continuous abnormal area are marked. For example, the clustered area is marked as the "dispatching imbalance area in southern North China", and the codes "GS01, GS02, PV03" and the core coordinates "(116.32E, 39.91N), (116.35E, 39.90N), (116.30E, 39.92N)" of all the units in the area are listed to obtain a set of dispatching imbalance linkage units. For example, this set contains multiple units such as "GS01, GS02, PV03", indicating that the units have a low coupling degree in load power consumption and need to be coordinated.

[0032] See also Figure 5 ,The steps for obtaining the grid unit dispatch risk level list are as follows: S411: extracting the unit power flow distribution curve under the specified number according to the position number of the scheduling imbalance linkage unit set, performing time uniform processing, identifying the unit time power flow change characteristics, and obtaining the unit power flow abnormal change characteristic set; Based on the fact that the dispatch imbalance linkage unit set includes the "GS01" unit and its number, the unit power flow distribution curve with the specified number is extracted. The unit power flow distribution curve refers to the real-time active power and reactive power data of the lines connected to the unit. For example, for the "GS01" unit, the power flow data of the 220kV outgoing line connected to the upper power grid is extracted every 5 minutes for the past 24 hours. Time unification is performed to ensure that the timestamps of all power flow data are consistent. For example, all data is unified to a standard 5-minute sampling interval through interpolation or sample alignment. The unit time power flow variation characteristics are identified. The unit time power flow variation characteristics are obtained by calculating the rate of change of the power flow value every 5 minutes. For example, if the active power flow changes from 100MW to 110MW in a certain 5-minute period, the change rate is +2MW / minute. The unit power flow abnormal change feature set is obtained. This feature set contains the power flow change rate series of the "GS01" unit over the 24-hour period. For example, during a specific period, the power flow change rate series of the unit exhibits abnormal fluctuations, such as continuous changes exceeding ±5MW / minute.

[0033] S412: Based on the unit power flow abnormal change feature set, identify the power flow distribution curve of the reference operation stage, compare the current power flow change sequence with the reference curve, identify the unit power flow deviation level, extract and mark the units whose deviation level exceeds the warning limit, and obtain the deviation sudden increase unit set; According to the abnormal change feature set of the unit tidal current, the tidal current distribution curve of the benchmark operation stage is identified. The tidal current distribution curve of the benchmark operation stage refers to the historical average tidal current curve under the condition of no faults and no abnormal dispatch instructions. For example, the tidal current data of typical sunny working days in the past month are averaged to obtain the benchmark tidal current curve of the unit. The current tidal current change sequence is compared with the reference curve. For example, the tidal current change rate sequence within the current 24 hours is compared point by point with the tidal current curve of the benchmark operation stage to identify the deviation level of the unit tidal current. The deviation level is determined based on the percentage deviation between the actual tidal current and the benchmark tidal current. For example, if the actual tidal current deviates from the benchmark tidal current by more than 1%, the deviation level will be less than 1%. The absolute value of the deviation is 0-5% for "normal deviation", 5-10% for "mild deviation", 10-20% for "moderate deviation", and more than 20% for "severe deviation". The classification is based on the power grid operation safety regulations and dispatching experience. For example, a deviation of 20% is close to the line transmission limit and belongs to the high-risk range. The units with deviation levels exceeding the warning limit are extracted and marked. The warning limit is set to "moderate deviation" and above, that is, the deviation exceeds 10%. For example, if the flow deviation of the "GS01" unit reaches 15% in a certain period of time, its deviation level is "moderate deviation", which exceeds the warning limit, and it is marked and included in the deviation sudden increase unit set.

[0034] S413: Based on the deviation sudden increase unit set, the deviation level value of each unit is bound to the position number in the power grid structure space diagram, sorted by risk level, and output a grid unit scheduling risk level list; According to the deviation sudden increase unit set, including the "GS01" unit, the deviation level value of each unit is bound to the position number in the power grid structure space diagram. For example, the "moderate deviation" level of the "GS01" unit is associated with its position number "GS01-123" and coordinates "(116.32E, 39.91N)". For example, the information after binding is "unit number: GS01, location: 220kV substation M area, deviation level: moderate deviation". Sorted by risk level, the risk level is sorted according to: severe deviation (Highest risk) > Moderate deviation > Mild deviation. For example, if the "GS03" unit is "Severe Deviation" and the "GS01" is "Moderate Deviation", the risk level of "GS03" is higher than that of "GS01". A list of grid unit dispatch risk levels is output, which clearly lists the risk level, unit number, and specific location information of all units with sudden deviations. For example, "GS03: Severe Deviation, (116.35E, 39.90N), GS01: Moderate Deviation, (116.32E, 39.91N)".

[0035] See also Figure 6 ,The specific steps for obtaining the automatic linkage scheduling adjustment instruction set are: S511: Calling the grid unit dispatch risk level list, extracting the unit number in the grid function map, mapping the unit risk level value with the regional coordinate boundary, identifying the unit information corresponding to the grid protection level, and generating a grid unit risk distribution map; Call the grid unit dispatch risk level list, which includes the "GS01" unit as "moderate deviation", and extract the unit number in the grid function diagram. The grid function diagram is an abstract grid logical topology diagram, in which the unit number is independent of its actual physical location and only represents its logical identification in dispatch. For example, the "GS01" unit is numbered "LCSU-001" in the grid function diagram. Map the unit risk level value with the regional coordinate boundary. For example, map the "moderate deviation" risk level of "GS01" to the geographic coordinate boundary of the region to which it belongs (for example, the Z2 sub-region of the North China region). For example, the boundary of the Z2 sub-region is [(116.0E, 39.5N), (117.0E, 40.5N)]. The risk level value can be quantified into a numerical value, for example, severe Deviation = 3, moderate deviation = 2, slight deviation = 1, thereby realizing numerical mapping and identifying the unit information corresponding to the power grid protection level. The power grid protection level refers to the importance level of different areas or equipment in the power grid. For example, core hub stations, important load centers, and general load areas correspond to different protection levels (for example, first-level protection zones, second-level protection zones, and third-level protection zones). The mapped risk information is superimposed on the preset power grid protection level layer. For example, it is identified that the Z2 sub-area where "GS01" is located belongs to the second-level protection zone. The risk level of the "GS01" unit is associated with the protection level to generate a power grid unit risk distribution map. The map displays the risk levels of various areas of the power grid in a visual form. For example, different shades of color are used to indicate high and low risks, and the protection level areas are marked.

[0036] S512: Based on the grid unit risk distribution map, the dispatch response unit numbers and response levels are extracted, the unit risk levels are matched with the dispatch response levels, the unit numbers with insufficient response coverage are identified, and a grid unit response risk disconnection list is obtained; Based on the grid unit risk distribution map, the dispatch response unit refers to the equipment or unit that can provide output regulation or load response under the grid dispatch instruction. Its response level is obtained through a comprehensive evaluation of parameters such as its adjustable power, response speed and response duration. For example, the response level is divided into "high response" (for example, the ability to provide 20% of the rated power within 5 minutes), "medium response" (for example, 10% within 10 minutes) and "low response" (for example, 5% within 30 minutes). Match the unit risk level with the dispatch response level. For example, compare the "medium deviation" risk level (quantitative value of 2) of the "GS01" unit with its current "medium response" level. If the risk level of the "GS01" unit is 2, and its dispatch response level is "medium response" (quantitative value of 2), this indicates that the matching degree is not yet Yes, but if a unit with a risk level of 3 (severe deviation) has a dispatch response level of "low response" (quantitative value of 1), it means there is a mismatch, and the unit numbers with insufficient response coverage are identified. The criterion for insufficient response coverage is: when the difference between the unit risk level quantification value and the response level quantification value is greater than or equal to 1, for example, the risk level is "severe deviation" (3), and the response level is "medium response" (2) or "low response" (1), the difference is 1 or 2, and both are considered to have insufficient response coverage. For example, if the "GS03" unit is found to be "severe deviation" but its response capability is only "low response", its number "GS03" is identified as insufficient response coverage, and a list of grid unit response risk disconnections is obtained, which clearly lists the unit numbers with high risk levels but insufficient dispatch response capabilities.

[0037] S513: Based on the grid unit response risk disconnection list and the level numbers in the grid protection priority sequence, the key unit numbers for which response coverage needs to be improved are extracted, and adjustment control parameters for linkage with the original dispatching unit are output in sequence, along with an automatic linkage dispatch adjustment instruction set. According to the grid unit response risk disconnection list, including the "GS03" unit, according to the level number in the grid protection priority sequence, the grid protection priority sequence is preset, for example, ultra-high voltage transmission channels and core city power supply areas belong to the highest priority (priority 1), important industrial parks belong to the second highest priority (priority 2), and general load areas belong to the lower priority (priority 3), extract the key unit numbers that need to improve the response coverage, for example, the "GS03" unit belongs to the priority 1 area, it is considered to be a key unit that needs to give priority to improving the response coverage, even if its risk level is "moderate deviation", it will be given priority because it is in a high priority protection area, and if another "GS04" unit is also "medium If the system is in priority 3, then GS03 will have a higher priority and will sequentially output adjustment control parameters linked to the original dispatch unit. These adjustment control parameters include: energy storage charging and discharging power adjustment instructions (for example, requiring the energy storage in the GS03 unit to increase its discharge power by 10MW within 5 minutes), photovoltaic output limit instructions (for example, requiring the photovoltaic power station in the GS03 unit to limit its output to below 30MW), and load-side response instructions (for example, requiring the controllable load in the area to be reduced by 5MW within 10 minutes). The system will then output an automatic linkage dispatch adjustment instruction set. For example, the final instruction set includes a series of coordinated control commands for the GS03 unit and its associated dispatch units to ensure stable grid operation.

[0038] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for optimizing power dispatch based on solar energy storage, characterized in that: The following steps are involved: S1: Based on the grid-based PV-storage combined units, the PV output variation curve and the energy storage state-of-charge fluctuation trajectory within the unit are extracted. The synchronous abnormal intervals of output fluctuation amplitude and state-of-charge variation trend are identified on the time axis. The corresponding unit number and spatial location information are extracted to generate a candidate set of abnormal units. S2: Recall the photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set, analyze the direction matching of the two at the unit junction, combine the energy flow connectivity of the unit boundary, screen the units with connectivity and abnormal output fluctuations, and form a layer of suspected scheduling imbalance units; S3: Based on the suspected scheduling imbalance unit layer, extract the load distribution matrix and regional power consumption characteristics within the unit, analyze the load fluctuation pattern and the consistency of power demand, screen and match abnormal units, and obtain a set of scheduling imbalance linkage units; S4: According to the position number of the dispatch imbalance linkage unit set, analyze the grid flow distribution trend corresponding to the unit, evaluate the degree of deviation from the grid benchmark operation curve, calibrate the dispatch imbalance level of the unit according to the deviation amplitude, and output a grid unit dispatch risk level list.

2. The power dispatch optimization method based on solar energy storage according to claim 1, characterized in that: The abnormal unit candidate set includes the output fluctuation unit number, the energy storage charge abnormality point, the unit coordinate mark, and the time series abnormality mark. The suspected scheduling imbalance unit layer includes the output direction abnormal unit mark, the unit boundary energy connectivity unit, the unit boundary matching block, and the abnormal output distribution grid. The scheduling imbalance linkage unit set includes the load abnormality continuous unit, the irregular power demand area, the load mutation overlap area, and the linkage abnormal unit number. The grid unit scheduling risk level list includes the risk level label, the unit response deviation index, the local power flow abnormality index, and the operation offset level.

3. The power dispatch optimization method based on solar energy storage according to claim 1, characterized in that: The steps for obtaining the abnormal unit candidate set are specifically as follows: S111: Based on the grid-divided photovoltaic and energy storage combined units, the PV output variation curve and the energy storage state of charge fluctuation trajectory within the unit are extracted. The two types of data within the same unit are compared and analyzed to obtain the trend value of the difference between output and state of charge; S112: Based on the output and state of charge difference trend value, identifying the fluctuation amplitude value in the output change curve and the deviation value of the state of charge fluctuation trajectory, superimposing the two types of values ​​over time, extracting the time intervals in which the fluctuation exceeds the reference range and the deviation value exceeds the set limit, and generating a high-frequency abnormal interval time period set; S113: For the high-frequency abnormal interval time period set, matching the corresponding unit numbers and spatial position information, extracting the unit positions where abnormal signals occur, and generating an abnormal unit candidate set.

4. The power dispatch optimization method based on solar energy storage according to claim 3 is characterized in that: The steps for obtaining the suspected scheduling imbalance unit layer are as follows: S211: Retrieving the photovoltaic output direction and energy storage charge and discharge capacity of the units in the abnormal unit candidate set, extracting the projection trajectory of the two at the unit boundary, identifying the distribution number and aggregation degree of the boundary points in the unit, and obtaining a unit boundary matching map; S212: Based on the unit boundary matching map, filter the boundary areas with a higher-than-average degree of aggregation, compare the spatial boundaries of the overall power grid structure map, identify continuous boundary concentrated areas belonging to the same unit, and obtain the abnormal output direction areas within the unit; S213: Call the abnormal output direction area in the unit, calculate the output direction discrete deviation index, perform integrated analysis on the unit boundary matching, output direction discreteness, load distribution balance and energy storage charging and discharging delay, match the units according to the response blocks in the layer, and form a layer of suspected scheduling imbalance units.

5. The power dispatch optimization method based on solar energy storage according to claim 4, characterized in that: The steps for obtaining the scheduling imbalance linkage unit set are specifically as follows: S311: Based on the suspected dispatch imbalance unit layer, extract the load distribution matrix and regional power consumption characteristics of the numbered units in the layer, perform time alignment on the data within the unit, identify the load intensity fluctuation characteristics and the power demand consistency offset characteristics, and obtain the local dispatch response feature set of the power grid; S312: Based on the local dispatch response feature set of the power grid, jointly analyze the load fluctuation pattern and the consistency of power demand within the unit, calculate the load-power coupling feature value, select the unit partitions with the load-power coupling degree in the layer, and establish a load-power coordinated response spatial distribution map; S313: Call the load power consumption coordinated response spatial distribution map, cluster the units that exceed the coordinated identification benchmark in the coupling characteristic value layer, mark the unit codes and coordinates corresponding to the continuous abnormal areas, and obtain the scheduling imbalance linkage unit set.

6. The power dispatch optimization method based on solar energy storage according to claim 5, characterized in that: The steps for obtaining the grid unit dispatch risk level list are specifically as follows: S411: extracting the unit power flow distribution curve under the specified number according to the position number of the scheduling imbalance linkage unit set, performing time uniform processing, identifying the unit time power flow change characteristics, and obtaining the unit power flow abnormal change characteristic set; S412: Based on the unit power flow abnormal change feature set, identify the power flow distribution curve of the reference operation stage, compare the current power flow change sequence with the reference curve, identify the unit power flow deviation level, extract and mark the units whose deviation level exceeds the warning limit, and obtain the deviation sudden increase unit set; S413: Based on the deviation sudden increase unit set, the deviation level value of each unit is bound to the position number in the power grid structure space diagram, sorted by risk level, and output a power grid unit scheduling risk level list.

7. The power dispatch optimization method based on solar energy storage according to claim 1, characterized in that: The method further comprises step S5: S5: calling the grid unit dispatch risk level list, identifying the corresponding unit number in the grid function diagram, calling the dispatch response unit list, comparing the response level with the grid protection priority sequence, screening the unit numbers whose response coverage needs to be adjusted, and outputting the automatic linkage dispatch adjustment instruction set; The automatic linkage scheduling adjustment instruction set includes an adjustment target unit number, a response level adjustment item, a protection priority comparison item, and a linkage response trigger type.

8. The power dispatch optimization method based on solar energy storage according to claim 7, characterized in that: The steps for obtaining the automatic linkage scheduling adjustment instruction set are specifically as follows: S511: Calling the grid unit dispatch risk level list, extracting the unit number in the grid function map, mapping the unit risk level value with the regional coordinate boundary, identifying the unit information corresponding to the grid protection level, and generating a grid unit risk distribution map; S512: Based on the grid unit risk distribution map, extract the dispatch response unit number and response level, match the unit risk level with the dispatch response level, identify the unit number with insufficient response coverage, and obtain a grid unit response risk disconnection list; S513: According to the grid unit response risk disconnection list and the level number in the grid protection priority sequence, the key unit numbers for which the response coverage needs to be improved are extracted, and the adjustment control parameters linked with the original dispatching unit are output in sequence, and an automatic linkage dispatching adjustment instruction set is output.

Images