A multi-energy complementary system combined event identification and extreme scenario construction method

Abstract

The application relates to the technical field of power system operation and risk analysis, and discloses a multi-energy complementary system combined event identification and extreme scenario construction method, which comprises the following steps: acquiring meteorological, source-side output, power grid operation and energy storage operation multi-source data and carrying out pretreatment; identifying single events based on quantile threshold value, duration constraint and change rate constraint; uniformly encoding and structurally expressing all single events to form a uniform event label sequence; constructing a combined event library based on the joint occurrence probability between events, time window coupling relationship, timing overlap degree and lag correlation; extracting a historical operation trajectory for the combined event and generating an extreme operation scenario, calculating the extreme degree of the scenario and the system risk contribution degree, and sorting and outputting a key extreme operation scenario set. The application has the advantages that unified event expression of multi-source heterogeneous data and automatic identification of combined events are realized, and an extreme operation scenario with power grid operation risk significance can be generated.

Description

Technical Field

[0001] This invention relates to the field of power system operation and risk analysis technology, and in particular to a method for identifying combined events and constructing extreme scenarios in multi-energy complementary systems. Background Technology

[0002] With the large-scale integration of new energy sources such as wind and solar power, the coupling relationship between power system operation and meteorological conditions is constantly strengthening. Meteorological factors such as wind speed, irradiance, temperature, precipitation, and water inflow not only directly affect the output of wind, solar, and hydropower, but also further influence load levels, system regulation requirements, and grid operation safety margins. In multi-energy complementary systems of hydropower, wind, solar, and energy storage, hydropower regulation capacity, wind and solar power output fluctuations, grid channel constraints, and energy storage state of charge jointly affect the system's supply and demand balance, new energy consumption, and safe operation level.

[0003] Existing methods for constructing extreme scenarios often take a single perspective. For example, they identify extreme weather scenarios based on meteorological quantiles, construct low-output or high-output scenarios based on renewable energy output curves, or define typical operating modes such as "high load, low renewable energy" or "high renewable energy, low load" based on human experience. While these methods can reflect the impact of a single factor or a few factors on system operation to some extent, they still have the following shortcomings:

[0004] First, existing methods typically treat meteorological conditions, source-side output, grid operation status, and energy storage status separately, lacking a unified event expression method for multi-source heterogeneous data, making it difficult to incorporate uncertainties such as low wind, low light, low water, high load, cross-sectional limitations, and low energy storage charge status into a unified analysis framework.

[0005] Second, existing methods rely heavily on manual enumeration to construct combined events, typically setting a few typical scenarios based on experience. This makes it difficult to automatically mine the joint occurrence relationships, temporal overlap relationships, and lag coupling relationships between different events from years of historical data. For example, low wind events, continuous low radiation events, drought events, and high load events may not occur completely simultaneously, but rather appear sequentially within a certain time window and jointly cause supply and demand tensions. Traditional methods that superimpose events at the same time are difficult to accurately characterize such processes.

[0006] Third, existing extreme scenario screening methods often focus on whether the weather or power output itself is extreme, lacking an evaluation of the correlation between these scenarios and grid operation risk indicators. Some extreme weather scenarios may not necessarily cause significant grid risks, while some medium- to high-intensity events may pose high operational risks under conditions of limited transmission channels, unfavorable energy storage status, or insufficient reserves. Therefore, selecting scenarios solely based on weather intensity or new energy output deviations cannot guarantee that the constructed scenarios are relevant to supply and demand balance analysis, new energy consumption assessment, energy storage configuration optimization, and grid safety verification. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for identifying combined events and constructing extreme scenarios in multi-energy complementary systems.

[0008] The objective of this invention is achieved through the following technical solution: a method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system, the method comprising the following steps:

[0009] S1. Obtain multi-source operation data of the target area in historical time periods. The multi-source operation data includes meteorological data, source-side power output data, power grid operation status data and energy storage operation data. Perform time scale unification, missing value repair, outlier processing and normalization on the multi-source operation data to form a multi-source feature sequence of the target area.

[0010] S2. Based on multi-source feature sequences, events are identified for meteorological conditions, source-side power output, power grid operation status, and energy storage operation status, respectively, and continuous operation variables are transformed into single event labels; thus, a multi-source heterogeneous single event set including meteorological events, source-side power output events, power grid operation events, and energy storage status events is obtained.

[0011] S3. All single events identified in S2 are uniformly encoded and structurally represented. The multi-source heterogeneous single event set is uniformly encoded and structurally represented. Each type of single event is mapped to a binary event label, and an event label vector and an event time series matrix are constructed in chronological order to form a unified event label sequence. The unified event label sequence is used to characterize the occurrence state of various single events at different times and serves as the input for subsequent combined event construction.

[0012] S4. Based on the unified event label sequence obtained in S3, analyze the joint occurrence relationship, time window coupling relationship and delayed triggering relationship between different single events, and construct a combined event library under multiple uncertain conditions;

[0013] S5. For the combined events obtained in S4, extract the corresponding historical operation trajectories and combine them with the operation constraints of the multi-energy complementary system of water, wind, solar and storage to generate extreme operation scenarios with complete time series characteristics; calculate the extreme degree of the scenario and the contribution of system risk, construct scenario evaluation indicators by comprehensively considering the probability of occurrence of combined events, event coupling strength, scenario extreme degree and system risk contribution, rank the generated scenarios, and select the top-ranked scenarios as the key extreme operation scenario set.

[0014] Specifically, in S2, meteorological events are identified based on quantile thresholds, as follows:

[0015] For any meteorological variable Based on its historical probability distribution, an event identification threshold is constructed, and a threshold is set. For variables At the quantile level Based on the historical quantiles below, low-value events are defined as follows:

[0016] ;

[0017] In the formula, For variables At any moment Did a low-value event occur? It is at the low quantile level;

[0018] High-value events can be defined as:

[0019] ;

[0020] In the formula, For variables At any moment Has a high-value event occurred? High quantile level;

[0021] The source-side power output events are identified based on a normalized power output threshold and a power output change rate threshold, specifically:

[0022] First, the output power of different types of power supplies is normalized:

[0023] ;

[0024] In the formula, These are wind power, solar power, and hydropower, respectively. For a moment The Power output of similar types; For the first Power supply installed capacity; Contribute to normalization;

[0025] Low output event is defined as:

[0026] ;

[0027] In the formula, For the first Low quantile threshold of normalized output power for similar power sources; This is a low-output event;

[0028] High-output events are defined as:

[0029] ;

[0030] In the formula, For the first High quantile threshold of normalized output power for similar power sources; This is a high-output event;

[0031] A rapid climb event is represented as:

[0032] ;

[0033] In the formula, For the first High quantile threshold of the absolute value of the change in power output of a power source; Identify quantile levels for hill climbing events; This refers to a rapid uphill event; This represents the change in output.

[0034] The power grid operation events are identified based on load thresholds, cross-sectional utilization thresholds, and reserve capacity thresholds, specifically:

[0035] High load events are defined as:

[0036] ;

[0037] In the formula, For a moment The system load; The high quantile threshold for load; This is a high-load event;

[0038] Low load events are defined as:

[0039] ;

[0040] In the formula, The lower quantile threshold for load; This is a low-load event;

[0041] The cross-section overload event is defined as follows:

[0042] ;

[0043] In the formula, For the utilization rate of key sections; The threshold for judging cross-sectional overload; This is a cross-sectional overload event;

[0044] Insufficient reserves event is defined as:

[0045] ;

[0046] In the formula, For a moment System backup capacity; For a moment The system backup requirements; This is a backup shortage event;

[0047] Channel restricted events are defined as:

[0048] ;

[0049] In the formula, For a moment Available channel capacity; The threshold for determining channel restriction; This is a channel-restricted event;

[0050] The energy storage state events are identified based on the energy storage state of charge threshold and the energy storage available adjustable energy threshold, specifically:

[0051] Energy storage low-charge state event is defined as:

[0052] ;

[0053] In the formula, For a moment The energy storage state of charge; The low state of charge threshold for energy storage; This is a low-charge state event for energy storage.

[0054] Energy storage high-charge state events are defined as follows:

[0055] ;

[0056] In the formula, The high-charge state threshold for energy storage; This is a high-charge state event for energy storage.

[0057] Energy storage can be expressed as discharge energy:

[0058] ;

[0059] In the formula, The minimum state of charge allowed for energy storage; This refers to the rated energy capacity of the energy storage system. It can be used to store energy for discharge.

[0060] Insufficient energy storage discharge support event is represented as:

[0061] ;

[0062] In the formula, This is an event indicating insufficient energy storage discharge support.

[0063] Insufficient energy storage charging absorption is represented as:

[0064] ;

[0065] In the formula, For a moment The energy storage discharge support demand; For a moment The demand for energy storage and charging absorption; Insufficient charging capacity for energy storage.

[0066] Specifically, the event time series matrix is ​​as follows:

[0067] ;

[0068] In the formula, This is an event time series matrix; n represents the total duration of the historical samples; n is the number of single event types. Let be the binary event label for the i-th single event at time t, with a value of 1 or 0, representing whether the single event occurred or did not occur at that time, respectively; where; This indicates that the nth single event is in the... Binary event labels at each moment;

[0069] The sliding accumulation enhancement is expressed as:

[0070] ;

[0071] In the formula, The length of the time window; The duration of the event.

[0072] Specifically, the construction of the combined event library under multiple uncertain conditions in S4 includes:

[0073] A combined event is defined as a set of several single events; simultaneous occurrence combined events are identified and their simultaneous occurrence probability is calculated; a time window coupling discrimination mechanism is introduced to identify time window coupled combined events and calculate the joint occurrence probability of time windows; the temporal overlap and lag correlation between events are calculated; a combined event coupling strength index is constructed by combining the joint occurrence probability of time windows, temporal overlap, and lag correlation; candidate combined events are screened according to the coupling strength threshold to form combined events.

[0074] Specifically, the identification of simultaneous combined events includes:

[0075] For combined events Its occurrence at time t is labeled as follows:

[0076] ;

[0077] In the formula, For combined events At any moment The occurrence label; Let q be the m-th combined event; q is the combined event. The number of single events included; For combined events The j-th single event in; For a single event The event label value at time t; when combining events All single events occur at time t. =1; otherwise, =0;

[0078] Combination events The probability of simultaneous occurrence within a historical period is expressed as:

[0079] ;

[0080] In the formula, This represents the total duration of the historical samples. For combined events The probability of simultaneous occurrence within a historical period.

[0081] The time-window coupled combined event recognition includes:

[0082] Let the time window length be Then a single event At any moment The status of the nearby windows is:

[0083] ;

[0084] In the formula, It is an indicator function; For a single event At any moment The status of nearby windows has changed;

[0085] Combination events The time window coupling label is:

[0086] ;

[0087] In the formula, For combined events Time window coupling label;

[0088] Combination events The probability of the time windows occurring together is:

[0089] ;

[0090] In the formula, This characterizes the probability of multiple events occurring together within a specified time window.

[0091] Specifically, the temporal overlap between the events is as follows:

[0092] ;

[0093] In the formula, The average time overlap between individual events within a combined event; and For two single events; and Representing single events and The event label at time t; T is the total duration of the historical samples;

[0094] For two events and Its lag time The correlation is as follows:

[0095] ;

[0096] In the formula, The time lag is the time delay. This is the function for calculating the correlation coefficient. For two events in lag time The correlation below;

[0097] Within the maximum permissible lag time D, the maximum lag correlation between the two events is:

[0098] ;

[0099] In the formula, The maximum lag correlation between two events;

[0100] Combination events The average maximum lag correlation is:

[0101] ;

[0102] In the formula, For combined events The average maximum lag correlation; The maximum lag correlation between two single events within a preset maximum lag time range; a and b are the sequence numbers of the single events in the combined event, and a < b.

[0103] Specifically, the coupling strength index of the combined events is:

[0104] ;

[0105] In the formula, For combined events The coupling strength; The probability of joint occurrence within a time window; The average temporal overlap of combined events; The average maximum lag correlation of combined events; , , These are the weighting coefficients. ;

[0106] The candidate combination events are filtered as follows:

[0107] ;

[0108] In the formula, This is the threshold for the coupling strength of combined events.

[0109] Specifically, the combined events in the combined event library are divided into four categories: insufficient supply, limited absorption, rapid fluctuation, and insufficient adjustment.

[0110] The supply shortage type combination event includes at least two of the following: low wind event, low radiation event, dry season event, high load event, and low charge state of energy storage event;

[0111] The combined events with limited absorption capacity include at least two of the following: high wind power output events, high photovoltaic power output events, low load events, channel-limited events, and high energy storage state of charge events.

[0112] The rapid fluctuation type of combined events includes at least two of the following: wind power rapid ramp-up event, photovoltaic rapid ramp-up event, load rapid change event, and energy storage insufficient response event;

[0113] The combined events of insufficient regulation include at least two of the following: low regulation capacity of hydropower, insufficient reserve capacity, insufficient available regulation capacity of energy storage, and maintenance constraint events.

[0114] Specifically, the extreme degree of the scenario is calculated as follows:

[0115] For a single runtime variable Define the degree to which it deviates from the normal state:

[0116] ;

[0117] In the formula, This is the historical average. The historical standard deviation; For variables The degree of standardization deviation;

[0118] For combined events Define the extreme intensity of the scene:

[0119] ;

[0120] In the formula, For extreme intensity scenarios; For the first The degree of extreme deviation of class variables from the scene; For the corresponding variable weights;

[0121] The system risk contribution is calculated as follows:

[0122] ;

[0123] In the formula, Contribution to systemic risk; , , , These are the risks of supply and demand gap after normalization, risks of limited new energy consumption, risks of exceeding the limit of cross sections, and risks of insufficient energy storage regulation. , , , The weighting coefficients for the four types of risk indicators mentioned above are respectively, satisfying the following conditions: .

[0124] Specifically, the scenario evaluation metrics are:

[0125] ;

[0126] In the formula, As a scene evaluation metric; The probability of the combined event occurring within a time window; The event coupling strength; For extreme intensity scenarios; Contribution to systemic risk; , , , These are the weighting coefficients corresponding to the probability of occurrence of the combined event time window, the event coupling strength, the extreme intensity of the scenario, and the contribution of system risk, respectively, and satisfying the following conditions: + + + =1.

[0127] The present invention has the following advantages:

[0128] 1. This invention overcomes the limitations of traditional extreme scenario construction methods that model meteorological factors, renewable energy output, grid operation status, and energy storage status independently. It proposes a unified expression mechanism for multi-source events in hydro-wind-solar-storage complementary systems. By constructing a multi-dimensional feature space encompassing meteorological, source-side output, grid operation, and energy storage status fields, continuous variables such as wind speed, irradiance, water inflow, load level, cross-sectional power flow, and energy storage state of charge are mapped into physically meaningful event label sequences, achieving a unified description of data with different time scales, dimensions, and physical attributes. Simultaneously, an event recognition mechanism based on probability quantile thresholds, duration constraints, and rate of change constraints is introduced to transform operating states such as low wind, low light, low water, high load, cross-sectional heavy load, and insufficient energy storage regulation into standardized event units, providing a unified data foundation for subsequent combined event mining and extreme scenario generation.

[0129] 2. This invention addresses the problem that existing extreme scenario construction methods rely on manual experience to set typical scenarios and struggle to discover the collaborative evolution patterns of multiple factors. It proposes an automatic identification method for combined events that considers time window coupling and event propagation relationships. By analyzing the joint occurrence relationships, temporal overlap characteristics, and sequential triggering relationships among different events, it uncovers deep correlations between meteorological anomalies, source-side power output fluctuations, load changes, grid constraints, and changes in energy storage status, achieving a shift from identifying single extreme factors to identifying multi-factor combined risk events. This method can automatically discover complex risk patterns that traditional methods struggle to cover, such as insufficient renewable energy supply under the combined effects of persistent low winds and low radiation, the absorption risk under the combined effects of high renewable energy output and limited transmission channels, and the supply-demand tension risk under the combined effects of load growth and decreased energy storage regulation capacity.

[0130] 3. This invention proposes a combined event evaluation and extreme scenario screening method for power grid operation risk, overcoming the shortcomings of traditional methods that rely solely on single indicators such as meteorological intensity and renewable energy output deviation to determine extreme scenarios. By establishing the correlation between combined events and supply-demand gaps, renewable energy consumption, grid section safety, and energy storage regulation capacity, and comprehensively considering the probability of event occurrence, degree of sustained impact, coupling strength, and contribution to system operation risk, candidate combined events are ranked and screened. This method can avoid a large number of invalid scenarios that only have meteorological extremes but lack significance for power grid risk, forming a set of high-risk extreme scenarios that better meet the actual operation needs of the power system, and improving the support capability of the scenario construction results for operation analysis, planning decisions, and risk assessment. Attached Figure Description

[0131] Figure 1 This is a schematic diagram of the method for constructing extreme operating scenarios according to the present invention. Detailed Implementation

[0132] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 for explaining the invention and are not intended to limit the invention; that is, the described embodiments are merely some embodiments of the invention, and not all embodiments. The components of the embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0133] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0134] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0135] The present invention will be further described below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0136] like Figure 1 As shown, a method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system is presented. This method includes the following steps:

[0137] S1. Obtain multi-source operation data of the target area in historical time periods. The multi-source operation data includes meteorological data, source-side power output data, power grid operation status data and energy storage operation data. Perform time scale unification, missing value repair, outlier processing and normalization on the multi-source operation data to form a multi-source feature sequence of the target area.

[0138] In this embodiment, meteorological data includes at least one of wind speed, irradiance, temperature, precipitation, and water inflow; source-side power output data includes at least one of hydropower output, wind power output, and photovoltaic power output; grid operation status data includes at least one of load, power flow at key sections, reserve capacity, available channel capacity, unit maintenance status, and line maintenance status; energy storage operation data includes at least one of energy storage state of charge, energy storage charging power, energy storage discharging power, and available energy storage regulation capacity; the above data undergoes time scale unification, missing value repair, outlier processing, and normalization to form the target area at any given time. Multi-source feature vectors:

[0139] ;

[0140] In the formula, For a moment Multi-source feature vectors; This is a meteorological feature vector; This represents the source-side output force characteristic vector; This is a feature vector representing the operating state of the power grid. This is the feature vector of the energy storage operation state.

[0141] The meteorological feature vector is represented as:

[0142] ;

[0143] In the formula, For a moment wind speed; For a moment irradiance; For a moment Temperature; For a moment The amount of precipitation; For a moment The amount of water coming in.

[0144] The source-side output eigenvector is represented as:

[0145] ;

[0146] In the formula, For a moment Wind power output; For a moment Photovoltaic power output; For a moment Hydropower output.

[0147] The power grid operating state feature vector is represented as follows:

[0148] ;

[0149] In the formula, For a moment The system load; For a moment Key cross-sectional tidal current; For a moment The spare capacity; For a moment Available channel capacity; For a moment Maintenance status variables.

[0150] The energy storage operation state feature vector is represented as follows:

[0151] ;

[0152] In the formula, For a moment The energy storage state of charge; For a moment Energy storage charging power; For a moment The energy storage discharge power; For a moment The energy storage can be used to regulate energy.

[0153] To eliminate differences in data units and value ranges, continuous variables can be normalized. For any continuous variable... Its normalized value is expressed as:

[0154] ;

[0155] In the formula, These are the normalized variable values; and Variables Minimum and maximum values ​​in the historical sample; To prevent extremely small positive numbers with a denominator of zero.

[0156] For missing data, linear interpolation, mean imputation of adjacent time periods, or imputation using historical samples of the same type can be used to repair the missing data, depending on the duration of the missing data and the type of variable. For obviously abnormal data, physical constraints and statistical thresholds can be combined for identification and correction. For example, the output of wind power, photovoltaic power, and hydropower should not exceed the corresponding installed capacity, the state of charge of energy storage should be within the allowable operating range, and the power flow at key sections should not exceed the physically feasible range.

[0157] After the above processing, a multi-source feature sequence with consistent time scale, unified dimensions, and reasonable physical constraints is obtained:

[0158] ;

[0159] In the formula, These are multi-source feature sequences within a historical period; This represents the total duration of historical samples. This multi-source feature sequence serves as the foundational input for subsequent single-event identification, event labeling, combined event construction, and extreme scenario generation.

[0160] S2. Based on multi-source feature sequences, event identification is performed on meteorological conditions, source-side power output, grid operation status, and energy storage operation status, respectively, transforming continuous operating variables into single-event labels. These single events include meteorological events, source-side power output events, grid operation events, and energy storage status events. This embodiment identifies typical single events that can affect the operation status of the hydro-wind-solar-storage multi-energy complementary system from multi-source historical data, including meteorological events, source-side power output events, grid operation events, and energy storage status events. Each type of event can be identified through threshold discrimination, duration constraints, and rate of change constraints; specifically...

[0161] Meteorological events are identified based on quantile thresholds, specifically:

[0162] For any meteorological variable Based on its historical probability distribution, an event identification threshold is constructed, and a threshold is set. For variables At the quantile level Based on the historical quantiles below, low-value events are defined as follows:

[0163] ;

[0164] In the formula, For variables At any moment Did a low-value event occur? For lower quantile levels, 5%, 10%, or 15% can be used depending on project requirements; For variables At the low quantile level Historical quantiles below;

[0165] High-value events can be defined as:

[0166] ;

[0167] In the formula, For variables At any moment Has a high-value event occurred? For a high percentile level, 85%, 90%, or 95% can be used depending on the project requirements; For variables At the high percentile level Historical quantiles below;

[0168] Based on the above methods, typical meteorological events such as low wind events, strong wind events, low radiation events, high temperature events, cold wave events, heavy precipitation events, and drought events can be identified.

[0169] Low wind events are represented as:

[0170] ;

[0171] In the formula, For a moment wind speed; The lower quantile threshold for wind speed; This is a low-wind event;

[0172] A strong wind event is represented as:

[0173] ;

[0174] In the formula, The high percentile threshold for wind speed; This was a strong wind event;

[0175] Low radiation events are represented as:

[0176] ;

[0177] In the formula, For a moment irradiance; The lower quantile threshold of irradiance; This is a low-irradiation event;

[0178] A drought event is represented as:

[0179] ;

[0180] In the formula, For a moment The amount of water coming in; The lower quantile threshold for water inflow; This is a drought event.

[0181] For a multi-energy complementary system integrating hydropower, wind power, solar power, and energy storage, a single moment of low wind, low solar power, or low water levels may not pose a significant operational risk. However, continuous events are more likely to cause supply and demand imbalances or insufficient regulation capacity. Therefore, this embodiment further introduces duration constraints to identify consecutive low or high value events.

[0182] set up For a certain basic event at time An event label is defined as a continuous event if the event occurs continuously within a continuous time window.

[0183] ;

[0184] In the formula, For the event Does it constitute a continuous event? To examine the length of the time window; The minimum duration threshold;

[0185] For example, a series of low-irradiation events can be represented as:

[0186] ;

[0187] In the formula, This is a series of low-irradiation events;

[0188] A series of low wind events is represented as:

[0189] ;

[0190] In the formula, This is a series of low-wind events;

[0191] By constraining the duration, we can avoid misjudging short-term random fluctuations as extreme events, and at the same time, we can more accurately identify process events that have a significant impact on system operation, such as continuous rain, persistent low winds, and persistent drought.

[0192] Source-side power output events are identified based on normalized power output thresholds and power output change rate thresholds, used to characterize the power output levels and change characteristics of hydropower, wind power, and photovoltaic power, specifically:

[0193] First, the output power of different types of power supplies is normalized:

[0194] ;

[0195] In the formula, These are wind power, solar power, and hydropower, respectively. For a moment The Power output of similar types; For the first Power supply installed capacity; Contribute to normalization;

[0196] Low output event is defined as:

[0197] ;

[0198] In the formula, For the first Low quantile threshold of normalized output power for similar power sources; This is a low-output event;

[0199] High-output events are defined as:

[0200] ;

[0201] In the formula, For the first High quantile threshold of normalized output power for similar power sources; This is a high-output event;

[0202] For fluctuating power sources such as wind and solar power, rapid ramp-up events can be further identified. The change in output is defined as:

[0203] ;

[0204] A rapid climb event is represented as:

[0205] ;

[0206] In the formula, This represents the change in output. This refers to a rapid uphill event; For the first High quantile threshold of the absolute value of the change in power output of a power source; Identify quantile levels for hill climbing events.

[0207] The above methods can be used to identify wind power low output events, wind power high output events, photovoltaic low output events, photovoltaic high output events, hydropower low output events, hydropower high output events, and new energy rapid ramp-up events.

[0208] The power grid operation events are identified based on load thresholds, cross-section utilization thresholds, and reserve capacity thresholds. These events characterize risk factors in the power grid's operating state, including high load, low load, heavy cross-section load, insufficient reserve, channel constraints, and maintenance constraints. Specifically:

[0209] High load events are defined as:

[0210] ;

[0211] In the formula, For a moment The system load; The high quantile threshold for load; This is a high-load event;

[0212] Low load events are defined as:

[0213] ;

[0214] In the formula, The lower quantile threshold for load; This is a low-load event;

[0215] The cross-section overload event is defined as follows:

[0216] ;

[0217] ;

[0218] In the formula, For the utilization rate of key sections; The threshold for judging cross-sectional overload; For a moment Key cross-sectional tidal current; Transmission limits for critical sections; This is a cross-sectional overload event;

[0219] Insufficient reserves event is defined as:

[0220] ;

[0221] In the formula, For a moment System backup capacity; For a moment The system backup requirements; This is a backup shortage event;

[0222] Channel restricted events are defined as:

[0223] ;

[0224] In the formula, For a moment Available channel capacity; The threshold for determining channel restriction; This is a channel-restricted event;

[0225] Energy storage state events are identified based on the energy storage state of charge threshold and the energy storage available adjustable energy threshold, used to characterize the energy storage's adjustment capability under different operating scenarios. In scenarios with insufficient supply, a low energy storage state weakens the discharge support capability; in scenarios with limited renewable energy absorption, a high energy storage state weakens the charging absorption capability. Therefore, this embodiment identifies low energy storage state events and high energy storage state events respectively, specifically as follows:

[0226] Energy storage low-charge state event is defined as:

[0227] ;

[0228] In the formula, For a moment The energy storage state of charge; The low state of charge threshold for energy storage; This is a low-charge state event for energy storage.

[0229] Energy storage high-charge state events are defined as follows:

[0230] ;

[0231] In the formula, The high-charge state threshold for energy storage; This is a high-charge state event for energy storage.

[0232] Energy storage can be expressed as discharge energy:

[0233] ;

[0234] In the formula, The minimum state of charge allowed for energy storage; This refers to the rated energy capacity of the energy storage system. It can be used to store energy for discharge.

[0235] When the available discharge energy of the energy storage is lower than the discharge support requirement, an event of insufficient energy storage discharge support is identified:

[0236] ;

[0237] In the formula, This is an event indicating insufficient energy storage discharge support.

[0238] When the available charging space for energy storage is lower than the demand for new energy absorption, identify events indicating insufficient energy storage charging absorption:

[0239] ;

[0240] In the formula, For a moment The energy storage discharge support demand; For a moment The demand for energy storage and charging absorption; Insufficient charging capacity for energy storage;

[0241] Therefore, a multi-source heterogeneous single-event set can be obtained, including meteorological events, source-side power output events, grid operation events, and energy storage status events. Among them, the meteorological events include at least one of low wind events, strong wind events, low radiation events, continuous low radiation events, cold wave events, high temperature events, heavy precipitation events, and drought events; the source-side power output events include at least one of low wind power output events, high wind power output events, low photovoltaic power output events, high photovoltaic power output events, low hydropower output events, high hydropower output events, and rapid ramp-up events of new energy sources; the grid operation events include at least one of high load events, low load events, cross-section heavy load events, insufficient reserve events, channel restriction events, and maintenance constraint events; and the energy storage status events include at least one of low energy storage state events, high energy storage state events, and insufficient energy storage regulation capacity events, providing a foundation for subsequent event labeling, combined event construction, and extreme operation scenario generation.

[0242] S3. All multi-source heterogeneous single event sets identified in S2 are uniformly encoded and structurally represented, mapping each type of single event to a binary event label vector; an event time series matrix is ​​constructed, and the original binary events are subjected to sliding accumulation enhancement to characterize the event duration characteristics, forming a unified event label sequence; this unified event label sequence serves as the input for subsequent combined event analysis and extreme scenario construction, specifically:

[0243] Event set definition: All identified single events are uniformly represented as an event set.

[0244] ;

[0245] In the formula, For a collection of events; A collection of meteorological events; This is the set of power output events on the source side. A set of power grid operation events; A set of energy storage status events;

[0246] Suppose that there are a total of Different types of events constitute a unified event space:

[0247] ;

[0248] In the formula, For the first Single event type;

[0249] Event label vector definition, at any time Map all events to binary event label vectors:

[0250] ;

[0251] In the formula, This is a binary event label vector; , indicating an event At any moment Whether it happened; Indicates that an event has occurred; This indicates that the event did not occur;

[0252] Tag mapping rules (unified encoding mechanism) for any event Define its label mapping function:

[0253] ;

[0254] In the formula, It is an indicator function; This is the i-th single event; For a single event The binary event label at time t. When a single event... When the event identification condition is met and the event occurs at time t, =1; otherwise, =0. This mapping achieves the following unified goal: to convert continuous variable events (wind speed, load, SOC, etc.) into discrete event space representations, eliminate differences in different physical dimensions, and unify them into a combined event analysis framework.

[0255] The event time series matrix over the entire historical timescale T is as follows:

[0256] ;

[0257] In the formula, This is an event time series matrix; n represents the total duration of the historical samples; n is the number of single event types. Let be the binary event label for the i-th single event at time t, with a value of 1 or 0, representing whether the single event occurred or not at that time, respectively; where, This indicates that the nth single event is in the... The matrix contains binary event labels at each time step; it serves as the foundational data structure for subsequent combined event mining and probabilistic modeling.

[0258] To characterize the duration of an event, a sliding accumulation enhancement is applied to the original binary event, represented as:

[0259] ;

[0260] In the formula, The length of the time window; This represents the duration and intensity of an event. It is used to address the inability to distinguish between short-term fluctuations and persistent extreme events.

[0261] At time t, the system can be represented as a unified state vector:

[0262] ;

[0263] In the formula, The event has occurred; This represents the duration intensity of the event. It is used for subsequent combined event construction, temporal coupling analysis, and risk contribution calculation.

[0264] This embodiment maps four types of heterogeneous data—meteorological, source-side, power grid, and energy storage—to the same event space; establishes a standardized binary event expression system to provide a unified input for constructing combined events; achieves the computability and composability of events through time-dimensional matrix expression; and provides temporal structure information support for subsequent extreme scenario identification through continuous enhancement expression.

[0265] S4. Based on the unified event label sequence obtained in S3, analyze the joint occurrence relationship, time window coupling relationship, and delayed triggering relationship among different single events to construct a combined event library under multiple uncertain conditions. This embodiment further analyzes the joint occurrence relationship, time window coupling relationship, and delayed triggering relationship among different single events to construct a combined event library under multiple uncertain conditions. The purpose is to automatically mine event combinations with system operation risk significance from historical data, avoiding the reliance on manual experience to enumerate typical scenarios in traditional methods. Constructing a combined event library under multiple uncertain conditions includes:

[0266] A combined event is defined as a set of several single events; simultaneous occurrence combined events are identified and their simultaneous occurrence probability is calculated; a time window coupling discrimination mechanism is introduced to identify time window coupled combined events and calculate the joint occurrence probability of time windows; the temporal overlap and lag correlation between events are calculated; a combined event coupling strength index is constructed by combining the joint occurrence probability of time windows, temporal overlap, and lag correlation; candidate combined events are screened according to the coupling strength threshold to form combined events.

[0267] Basic definition of composite events

[0268] Let the set of single events obtained in S3 be:

[0269] ;

[0270] in, For the first Single event type.

[0271] Any combination of events Represented as a set of several single events:

[0272] ;

[0273] In the formula, The number of individual events contained in the combined event; For combined events The first in A single event.

[0274] For example, a supply shortage combination event is represented as:

[0275] ;

[0276] In the formula, This is a low-wind event; This is a low-irradiation event; This is a drought event; This is a high-load event; This is a low-charge state event for energy storage.

[0277] Simultaneous combination event identification includes:

[0278] For combined events If all the single events it contains occur at the same time If both occur, then the combined event is considered to occur at time [time]. Occurrence. Its occurrence label can be defined as:

[0279] ;

[0280] In the formula, For combined events At any moment The occurrence label; q is a combined event. The number of single events included; This is the m-th combined event; For combined events The j-th single event in; For a single event The event label value at time t. When combining events. All single events occur at time t. =1; otherwise, =0;

[0281] Combination events The probability of simultaneous occurrence within a historical period is expressed as:

[0282] ;

[0283] In the formula, This represents the total duration of the historical samples. For combined events The probability of simultaneous occurrence within a historical period;

[0284] The time-window coupled combined event recognition includes:

[0285] Let the time window length be Then a single event At any moment The status of the nearby windows is:

[0286] ;

[0287] In the formula, It is an indicator function; For a single event At any moment The status of nearby windows has changed;

[0288] Combination events The time window coupling label is:

[0289] ;

[0290] In the formula, For combined events Time window coupling label;

[0291] Combination events The probability of the time windows occurring together is:

[0292] ;

[0293] In the formula, This characterizes the probability of multiple events occurring together within a specified time window.

[0294] Compared to the probability of simultaneous occurrence The probability of time windows occurring together It can identify combined events that occur sequentially but have a combined effect, making it more suitable for scenarios with time-series transmission relationships, such as cold waves, high loads, low renewable energy output, and deterioration of energy storage status.

[0295] In addition to the joint occurrence probability, this embodiment further calculates the temporal overlap between events to measure the degree of overlap between different events in the time dimension. For any two single events... and Its temporal overlap is:

[0296] ;

[0297] In the formula, The average time overlap between individual events within a combined event; and For two single events; and Single event and The event label at time t; T is the total duration of the historical samples.

[0298] Some events have a clear sequential triggering relationship. For example, high temperature or cold wave events may cause peak loads with a lag, continuous rainy weather may cause sustained low photovoltaic output with a lag, and drought events may affect hydropower regulation capacity over a longer time scale.

[0299] To characterize this kind of sequential relationship, this embodiment calculates the lag correlation between events. For two events... and Its lag time The correlation is as follows:

[0300] ;

[0301] In the formula, The time lag is the time delay. This is the function for calculating the correlation coefficient. For two events in lag time The correlation below;

[0302] Within the maximum permissible lag time D, the maximum lag correlation between the two events is:

[0303] ;

[0304] In the formula, This represents the maximum lag correlation between two events; for example, an hourly lag window can be set between meteorological events and new energy output events, an hourly to daily lag window can be set between meteorological events and load events, and a daily to weekly lag window can be set between water inflow events and hydropower output events.

[0305] Combination events The average maximum lag correlation is:

[0306] ;

[0307] In the formula, For combined events The average maximum lag correlation; q is the combined event. The number of single events included; This represents the maximum lag correlation between two individual events within a preset maximum lag time range; a and b are the sequence numbers of the individual events in the combined event, and a < b. This indicator is used to determine whether there is a significant sequential triggering relationship between the individual events within the combined event.

[0308] Taking into account the probability of joint occurrence of time windows, temporal overlap, and lag correlation, a coupling strength index for combined events is constructed:

[0309] ;

[0310] In the formula, For combined events The coupling strength; The probability of joint occurrence within a time window; The average temporal overlap of combined events; The average maximum lag correlation of combined events; , , These are the weighting coefficients. ; , , It can be set according to the project's priorities. When there is a greater focus on the simultaneous occurrence of multiple events within a time window, it can improve performance. When focusing more on event synchronization and overlap, it can improve... When focusing more on the sequential triggering relationship of events, it can improve... ;

[0311] Candidate combined events are filtered based on their coupling strength. A candidate combined event is included in the combined event library if it meets the following conditions:

[0312] ;

[0313] In the formula, This is the threshold for the coupling strength of combined events.

[0314] To avoid creating too many low-value combined events, this embodiment can further set constraints on the number of events and the event category. That is, combined events. The following conditions must be met:

[0315] ;

[0316] in, and These represent the lower and upper limits for the number of individual events included in a combined event, respectively. Furthermore, a combined event should include at least two types of events from the following categories: meteorological events, source-side power output events, grid operation events, or energy storage status events, to ensure that the combined event reflects the coupling relationship between multiple uncertain factors.

[0317] The combined events in the combined event library are divided into four categories: insufficient supply, limited absorption, rapid fluctuation, and insufficient adjustment.

[0318] Supply-deficient combined events primarily characterize the risk of supply-demand tension caused by the combined effects of factors such as low available output from new energy and hydropower, high load demand, and insufficient energy storage discharge. This type of combined event is represented as follows:

[0319] ;

[0320] In the formula, This is a combination event characterized by insufficient supply. This is a backup shortage event; This is a low-wind event; This is a low-irradiation event; This is a drought event; This is a high-load event; The event is a low-charge state of energy storage; the supply shortage combined event includes at least two of the following: low wind event, low radiation event, dry season event, insufficient reserve event, high load event, and low-charge state of energy storage event.

[0321] The combined event of limited absorption mainly represents the risk of wind and solar curtailment under the combined effects of factors such as high renewable energy generation, low load, limited corridors, heavy cross-section load, and insufficient energy storage and charging space. This type of combined event is represented as:

[0322] ;

[0323] In the formula, This is for the purpose of absorbing constrained portfolio events; This was a strong wind event; This is a high-irradiation event; This is a low-load event; This is a channel-restricted event; This is a cross-sectional overload event; The energy storage high-charge state event; the absorption-limited combination event includes at least two of the following: strong wind event, high radiation event, low load event, channel-restricted event, cross-section heavy load event, and energy storage high-charge state event.

[0324] Rapid fluctuation-type combined events characterize short-term regulation pressures caused by rapid changes in wind power, photovoltaics, or load. These combined events are characterized by:

[0325] ;

[0326] In the formula, This is a rapidly fluctuating combination of events; The incident of rapidly ramping uphill for wind power output; The rapid ramp-up of photovoltaic power generation; This refers to events involving rapid load changes. This is an event indicating insufficient energy storage response.

[0327] The rapid fluctuation type of combined events includes at least two of the following: wind power rapid ramp-up event, photovoltaic rapid ramp-up event, load rapid change event, and energy storage insufficient response event;

[0328] Insufficient regulation combined events mainly characterize the risk of insufficient system regulation capacity under the combined effects of insufficient hydropower regulation capacity, insufficient reserve capacity, insufficient energy storage available regulation capacity, and increased maintenance constraints. This type of combined event is represented as:

[0329] ;

[0330] In the formula, For adjusting insufficient combination events; This is an event indicating low hydropower regulation capacity. For maintenance constraint events; This is a backup shortage event; This is a low-charge state event for energy storage. This is a high-charge state event for energy storage.

[0331] The combined events of insufficient regulation include at least two of the following: low hydropower regulation capacity event, insufficient reserve event, low energy storage state of charge event, high energy storage state of charge event, and maintenance constraint event.

[0332] Through the above screening and classification, a combined event library is formed:

[0333] ;

[0334] In the formula, For combining event libraries; For the first A combination of events; This represents the number of combined events obtained through filtering.

[0335] Each combined event Include at least the following information:

[0336] ;

[0337] In the formula, A collection of single events contained within a composite event; The probability of joint occurrence within a time window; This refers to the temporal overlap degree. The maximum lag correlation; The coupling strength; This is a combined event type.

[0338] Through this step, this embodiment can automatically identify combined events with temporal coupling and operational risk orientation from multi-source historical data, providing a foundation for subsequent risk contribution evaluation and extreme operation scenario generation.

[0339] S5. For the combined events obtained in S4, extract the corresponding historical operating trajectories and, combined with the operating constraints of the hydro-wind-solar-storage multi-energy complementary system, generate extreme operating scenarios with complete temporal characteristics. Calculate the extreme degree and system risk contribution of the scenarios, and construct scenario evaluation indicators by comprehensively considering the probability of occurrence of combined events, event coupling strength, scenario extreme degree, and system risk contribution. Rank the generated scenarios and select the top-ranked scenarios as the key extreme operating scenario set. This embodiment further extracts the corresponding historical operating trajectories for the selected high-risk combined events and, combined with the operating constraints of the hydro-wind-solar-storage multi-energy complementary system, generates extreme operating scenarios with complete temporal characteristics. Unlike traditional methods that directly select historical extreme weather days as typical scenarios, this embodiment reconstructs the historical operating state based on the occurrence conditions and system risk characteristics of the combined events, forming extreme operating scenarios that simultaneously include meteorological conditions, source-side output, grid conditions, and energy storage conditions.

[0340] The resulting combined event library is as follows:

[0341] ;

[0342] For any combination of events The corresponding set of occurrence times is determined based on the sequence of event labels:

[0343] ;

[0344] in: For combined events The corresponding set of historical moments; To determine whether the combined event is true within the time window;

[0345] For each occurrence time Extract its continuous running states before and after:

[0346] ;

[0347] In the formula, This is the initial extreme scenario fragment corresponding to the combined event; The length of the historical state before the event occurred; The duration of the impact after the event occurs; to Each from before the combined event occurred The period from the time of occurrence to the time after Multi-source feature vectors at each time point within a time period. This method can preserve the formation process of combined events, rather than just extracting data at the instant the event occurs.

[0348] For any time t, define the system state vector:

[0349] ;

[0350] In the formula, Weather conditions; Contributing to the water, scenery, and energy sectors; The power grid's operating status; In energy storage operation mode; This is the system state vector;

[0351] The complete extreme operating scenario corresponding to the combined events is represented as follows:

[0352] ;

[0353] In the formula, The duration of the scene; For the first Extreme operating scenarios;

[0354] To avoid selecting scenarios solely based on the probability of combined events, this embodiment further considers the extreme nature of the scenario itself. The extreme nature of the scenario is calculated as follows:

[0355] For a single runtime variable Define the degree to which it deviates from the normal state:

[0356] ;

[0357] In the formula, This is the historical average. The historical standard deviation; For variables The degree of standardization deviation;

[0358] For combined events Define the extreme intensity of the scene:

[0359] ;

[0360] In the formula, For extreme intensity scenarios; For the first The degree of extreme deviation of class variables from the scene; These are the weights of the corresponding variables.

[0361] To ensure that the generated scenario not only reflects extreme weather conditions but also carries significance for power grid operation risks, this embodiment further calculates the system risk indicators corresponding to the scenario. The system risk contribution is calculated as follows:

[0362] Define the system's net supply and demand gap:

[0363] ;

[0364] In the formula, For a moment The supply and demand gap; It provides power to conventional power sources; This refers to the energy storage discharge power;

[0365] Supply and demand risks in various scenarios:

[0366] ;

[0367] In the formula, Risks related to supply and demand in various scenarios.

[0368] The risk of renewable energy absorption, and the renewable energy acceptance capacity are expressed as follows:

[0369] ;

[0370] In the formula, To ensure the acceptance of new energy sources; Energy storage charging capability; For external delivery capacity; To ensure that the power supply output is operational;

[0371] Wasted electricity from renewable energy sources:

[0372] ;

[0373] in,

[0374] ;

[0375] In the formula, For the abandoned electricity of new energy sources; Wind power output at time t With photovoltaic power sum.

[0376] Risks of power curtailment in various scenarios:

[0377] ;

[0378] In the formula, To mitigate the risk of power curtailment in certain scenarios.

[0379] Power grid security risks

[0380] The cross-section exceeds the limit:

[0381] ;

[0382] In the formula, For cross-sections exceeding the limit; Transmission limits for critical sections;

[0383] Scene cross-sectional risks:

[0384] ;

[0385] In the formula, Risks associated with cross-sectional areas of the scene.

[0386] Energy storage regulation risks

[0387] Energy storage discharge support capability:

[0388] ;

[0389] In the formula, To support the energy storage discharge capability;

[0390] Energy storage charging absorption capacity:

[0391] ;

[0392] In the formula, Energy storage charging absorption capacity;

[0393] Energy storage regulation gap:

[0394] ;

[0395] In the formula, This represents a gap in energy storage regulation. The energy storage discharge support gap indicates that the available discharge energy from energy storage cannot meet the system's supply-demand gap support requirements. The energy storage charging absorption gap indicates that the available charging space for energy storage cannot meet the demand for new energy absorption.

[0396] Scenarios Energy Storage Risks:

[0397] ;

[0398] In the formula, Risks associated with energy storage in various scenarios.

[0399] Normalize different types of risk indicators:

[0400] ;

[0401] In the formula, These correspond to the risks of supply and demand gaps, new energy consumption, cross-section exceeding limits, and energy storage regulation, respectively.

[0402] Further calculate the system risk contribution:

[0403] ;

[0404] In the formula, Contribution to systemic risk; , , , These are the risks of supply and demand gap after normalization, risks of limited new energy consumption, risks of exceeding the limit of cross sections, and risks of insufficient energy storage regulation. , , , The weighting coefficients for the four types of risk indicators mentioned above are respectively, satisfying the following conditions: .

[0405] Sorting of extreme operating scenarios

[0406] Taking into account: the probability of combined events, the degree of event coupling, the extreme nature of the scenario, and the contribution of the system to risk;

[0407] Construct the final scenario evaluation metrics:

[0408] ;

[0409] In the formula, As a scene evaluation metric; The probability of the combined event occurring within a time window; The event coupling strength; For extreme intensity scenarios; Contribution to systemic risk; , , , These are the weighting coefficients corresponding to the probability of occurrence of the combined event time window, the event coupling strength, the extreme intensity of the scenario, and the contribution of system risk, respectively, and satisfying the following conditions: + + + =1.

[0410] according to:

[0411] ;

[0412] The generated scenarios are sorted, and the top-ranked scenarios are selected as the key extreme running scenario set.

[0413] Through this step, this embodiment realizes the transformation from combined events to simulable extreme operating scenarios, so that the generated scenario simultaneously satisfies: extreme meteorological conditions, abnormal power output from the source side, correlation with grid operation risks, and constraints on energy storage regulation capabilities.

[0414] Therefore, the scenarios generated by this method can be directly used for subsequent supply and demand balance analysis, new energy consumption assessment, energy storage capacity optimization, and grid safety verification.

[0415] Through the above steps, an extreme operating scenario library containing the following information is ultimately formed:

[0416] ;

[0417] In the formula, For corresponding combined events; This is time-series data for extreme operating scenarios; Rate the priority of each scenario; For combined event types; Contribution to systemic risk.

[0418] The resulting database of extreme operating scenarios can serve as the basic input for the operation analysis of multi-energy complementary systems of hydro, wind, solar, and energy storage, supply and demand balance verification, new energy consumption assessment, energy storage configuration optimization, and grid risk early warning.

[0419] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any person skilled in the art can make many possible variations and modifications to the technical solution of the present invention, or modify it into equivalent embodiments, without departing from the scope of the present invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technology of the present invention without departing from the scope of the present invention are within the protection scope of the present invention.

Claims

1. A method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system, characterized in that: The method includes the following steps: S1. Obtain multi-source operation data of the target area in historical time periods. The multi-source operation data includes meteorological data, source-side power output data, power grid operation status data and energy storage operation data. Perform time scale unification, missing value repair, outlier processing and normalization on the multi-source operation data to form a multi-source feature sequence of the target area. S2. Based on multi-source feature sequences, events are identified for meteorological conditions, source-side power output, power grid operation status, and energy storage operation status, respectively, and continuous operation variables are transformed into single event labels; thus, a multi-source heterogeneous single event set including meteorological events, source-side power output events, power grid operation events, and energy storage status events is obtained. S3. All multi-source heterogeneous single event sets identified in S2 are uniformly encoded and structurally represented. Each type of single event is mapped to a binary event label, and an event label vector and an event time series matrix are constructed in chronological order to form a unified event label sequence. The unified event label sequence is used to characterize the occurrence state of various single events at different times and serves as the input for subsequent combined event construction. S4. Based on the unified event label sequence obtained in S3, analyze the joint occurrence relationship, time window coupling relationship and delayed triggering relationship between different single events, and construct a combined event library under multiple uncertain conditions; S5. For the combined events obtained in S4, extract the corresponding historical operation trajectories and combine them with the operation constraints of the multi-energy complementary system of water, wind, solar and storage to generate extreme operation scenarios with complete time series characteristics; calculate the extreme degree of the scenario and the contribution of system risk, construct scenario evaluation indicators by comprehensively considering the probability of occurrence of combined events, event coupling strength, scenario extreme degree and system risk contribution, rank the generated scenarios, and select the top-ranked scenarios as the key extreme operation scenario set.

2. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: In S2, meteorological events are identified based on quantile thresholds, specifically as follows: For any meteorological variable Construct an event identification threshold based on its historical probability distribution, and set... For variables At the quantile level Based on the historical quantiles below, low-value events are defined as follows: ； In the formula, For variables At any moment Did a low-value event occur? It is at the low quantile level; High-value events can be defined as: ； In the formula, For variables At any moment Has a high-value event occurred? High quantile level; The source-side power output events are identified based on a normalized power output threshold and a power output change rate threshold, specifically: First, the output power of different types of power supplies is normalized: ； In the formula, These are wind power, solar power, and hydropower, respectively. For a moment The Power output of similar types; For the first Power supply installed capacity; Contribute to normalization; Low output event is defined as: ； In the formula, For the first Low quantile threshold of normalized output power for similar power sources; This is a low-output event; High-output events are defined as: ； In the formula, For the first High quantile threshold of normalized output power for similar power sources; This is a high-output event; A rapid climb event is represented as: ； In the formula, For the first High quantile threshold of the absolute value of the change in power output of a power source; Identify quantile levels for hill climbing events; This refers to a rapid uphill event; This represents the change in output. The power grid operation events are identified based on load thresholds, cross-sectional utilization thresholds, and reserve capacity thresholds, specifically: High load events are defined as: ； In the formula, For a moment The system load; The high quantile threshold for load; This is a high-load event; Low load events are defined as: ； In the formula, This is a low-load event; The lower quantile threshold for load; The cross-section overload event is defined as follows: ； In the formula, For the utilization rate of key sections; The threshold for judging cross-sectional overload; This is a cross-sectional overload event; The insufficiency event is defined as: ； In the formula, For a moment System backup capacity; For a moment The system backup requirements; This is a backup shortage event; Channel restricted events are defined as: ； In the formula, For a moment Available channel capacity; The threshold for determining channel restriction; This is a channel-restricted event; The energy storage state events are identified based on the energy storage state of charge threshold and the energy storage available adjustable energy threshold, specifically: Energy storage low-charge state event is defined as: ； In the formula, For a moment The energy storage state of charge; The low state of charge threshold for energy storage; This is a low-charge state event for energy storage. Energy storage high-charge state events are defined as follows: ； In the formula, The high-charge state threshold for energy storage; This is a high-charge state event for energy storage. Energy storage can be expressed as discharge energy: ； In the formula, The minimum state of charge allowed for energy storage; This refers to the rated energy capacity of the energy storage system. It can be used to store energy for discharge. Insufficient energy storage discharge support event is represented as: ； In the formula, This is an event indicating insufficient energy storage discharge support. An event of insufficient energy storage charging absorption is represented as: ； In the formula, For a moment The energy storage discharge support demand; For a moment The demand for energy storage and charging absorption; Insufficient charging capacity for energy storage.

3. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: The event time series matrix is ​​as follows: ； In the formula, This is an event time series matrix; n represents the total duration of the historical samples; n is the number of single event types. Let be the binary event label of the i-th single event at time t, with a value of 1 or 0, representing whether the single event occurred or did not occur at that time, respectively; For the nth single event in the th... Binary event labels at each moment; The sliding accumulation enhancement is expressed as: ； In the formula, The length of the time window; The duration of the event.

4. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: The S4 section describes the construction of a combined event library under multiple uncertain conditions, which includes: A combined event is defined as a set of several single events; simultaneous occurrence combined events are identified and their simultaneous occurrence probability is calculated; a time window coupling discrimination mechanism is introduced to identify time window coupled combined events and calculate the joint occurrence probability of time windows; the temporal overlap and lag correlation between events are calculated; a combined event coupling strength index is constructed by combining the joint occurrence probability of time windows, temporal overlap, and lag correlation; candidate combined events are screened according to the coupling strength threshold to form combined events.

5. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 4, characterized in that: The identification of simultaneous combined events includes: For combined events Its occurrence at time t is labeled as follows: ； In the formula, For combined events At any moment The occurrence label; Let be the m-th combined event; q be the combined event. The number of single events included; For a single event The event tag value at time t; Combination events The probability of simultaneous occurrence within a historical period is expressed as: ； In the formula, This represents the total duration of the historical samples. For combined events Simultaneous occurrence probability within a historical time period The time-window coupled combined event recognition includes: Let the time window length be Then a single event At any moment The status of the nearby windows is: ； In the formula, It is an indicator function; For a single event At any moment The status of nearby windows has changed; Combination events The time window coupling label is: ； In the formula, For combined events Time window coupling label; Combination events The probability of the time windows occurring together is: ； In the formula, This characterizes the probability of multiple events occurring together within a specified time window.

6. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 4, characterized in that: The temporal overlap between the events is: ； In the formula, The average time overlap between individual events within a combined event; and For two single events; and Representing single events and The event label at time t; T represents the total duration of the historical samples; For two events and Its lag time The correlation is as follows: ； In the formula, The time lag is the time delay. This is the function for calculating the correlation coefficient. For two events in lag time The correlation below; Within the maximum permissible lag time D, the maximum lag correlation between the two events is: ； In the formula, The maximum lag correlation between two events; Combination events The average maximum lag correlation is: ； In the formula, For combined events The average maximum lag correlation; q is the combined event. The number of single events included; The maximum lag correlation between two single events within a preset maximum lag time range; a and b are the sequence numbers of the individual events in the combined event, and a < b.

7. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 4, characterized in that: The coupling strength index of the combined events is: ； In the formula, For combined events The coupling strength; The probability of joint occurrence within a time window; The average temporal overlap of combined events; The average maximum lag correlation of combined events; , , These are the weighting coefficients. ; The candidate combination events are filtered as follows: ； In the formula, This is the threshold for the coupling strength of combined events.

8. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: The combined events in the combined event library are divided into four categories: insufficient supply, limited absorption, rapid fluctuation, and insufficient adjustment. The supply shortage type combination event includes at least two of the following: low wind event, low radiation event, low water event, high load event, and low charge state of energy storage event; The combined events with limited absorption capacity include at least two of the following: high wind power output events, high photovoltaic power output events, low load events, channel-limited events, and high energy storage state of charge events. The rapid fluctuation type of combined events includes at least two of the following: wind power rapid ramp-up event, photovoltaic rapid ramp-up event, load rapid change event, and energy storage insufficient response event; The combined events of insufficient regulation include at least two of the following: low regulation capacity of hydropower, insufficient reserve capacity, insufficient available regulation capacity of energy storage, and maintenance constraint events.

9. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: The extreme degree of the scenario is calculated as follows: For a single runtime variable Define the degree to which it deviates from the normal state: ； In the formula, This is the historical average. The historical standard deviation; For variables The degree of standardization deviation; For combined events Define the extreme intensity of the scene: ； In the formula, For extreme intensity scenarios; For the first The degree of extreme deviation of class variables from the scene; For the corresponding variable weights; The system risk contribution is calculated as follows: ； In the formula, Contribution to systemic risk; , , , These are the risks of supply and demand gap after normalization, risks of limited new energy consumption, risks of exceeding the limit of cross sections, and risks of insufficient energy storage regulation. , , , For the weighting coefficients, satisfying .

10. The method for identifying combined events and constructing extreme scenarios in a multi-energy complementary system according to claim 1, characterized in that: The evaluation metrics for the scenario are: ； In the formula, As a scene evaluation metric; The probability of the combined event occurring within a time window; The event coupling strength; For extreme intensity scenarios; Contribution to systemic risk; , , , These are the weighting coefficients corresponding to the probability of occurrence of the combined event time window, the event coupling strength, the extreme intensity of the scenario, and the contribution of system risk, respectively, and satisfying the following conditions: + + + =1.

Images