A method and system for constructing a double-layer robust optimization scheduling strategy of wind-solar-water storage

Abstract

The application provides a kind of wind light water storage double-layer robust optimization scheduling strategy construction method and system, relating to energy management technical field, including first obtaining the multi-source data of wind farm, photovoltaic power station and cascade hydropower station;Characteristic extraction and integrated prediction are carried out based on the multi-source data, and wind light power prediction result is generated;Further, water power coupling modeling analysis is combined to obtain the power generation capacity of hydropower station and the sequence of adjustable water volume;By constructing multi-layer energy coupling graph and carrying out spectral domain structure optimization, power coupling network representing space-time correlation and ecological sensitivity is formed;Finally, double-layer distributed robust optimization is used to optimize economic target and ecological risk constraint simultaneously, and generate collaborative scheduling strategy.The application effectively solves the time sequence matching problem between wind light fluctuation and water power time lag, improves the renewable energy consumption capacity and water resource utilization efficiency, and reduces the ecological operation risk.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of energy management, in particular to a wind-solar-water-storage double-layer robust optimization scheduling strategy construction method and system. BACKGROUND

[0002] With large-scale grid connection of renewable energy such as wind power and photovoltaic power, the power grid operation faces significant volatility and uncertainty challenges. In the scene of cascade hydropower and wind-solar joint operation, the current scheduling method is mainly based on day or hour scale power prediction and economic scheduling model, focusing on medium and long term power balance and economic benefit optimization. However, this kind of method often ignores the time coupling problem between short time sharp fluctuations of wind-solar output and response lag of hydropower system: that is, the short time peak of wind-solar output may just fall into the ecological sensitive period of downstream river section, or overlap with the delay time window of cascade reservoir discharge propagation, forcing the scheduling system to respond in the form of frequent start-stop of units, additional water abandonment or discharge in the ecological unfavorable period, etc., thereby leading to the decrease of water resource utilization efficiency, the increase of abandoned power rate and the increase of ecological risk.

[0003] Therefore, there is an urgent need for a wind-solar-water-storage double-layer robust optimization scheduling strategy construction method and system to solve the above technical problems. SUMMARY

[0004] The application aims to provide a wind-solar-water-storage double-layer robust optimization scheduling strategy construction method and system to improve the above problems. In order to achieve the above purpose, the technical scheme adopted by the application is as follows:

[0005] In a first aspect, the application provides a wind-solar-water-storage double-layer robust optimization scheduling strategy construction method, comprising:

[0006] obtaining first information, second information and third information, the first information including wind speed and unit operation state data of wind power plants in a target area, irradiance and module temperature data of photovoltaic power stations, the second information including inflow runoff data, reservoir capacity-water level data, historical discharge-water level response observation data and downstream ecological flow constraint data of cascade hydropower stations;

[0007] performing feature extraction and integrated prediction analysis based on the first information to generate power prediction results of wind power generation and photovoltaic power generation;

[0008] performing water power coupling modeling analysis based on the power prediction results of wind power generation and photovoltaic power generation and the second information to obtain power prediction sequence and adjustable water quantity sequence of hydropower stations;

[0009] Based on the power prediction results of the wind power generation and the photovoltaic power generation, the power prediction sequence and the mobilizable water quantity sequence of the hydropower station are used to construct a multi-layer energy coupling graph, and a power coupling network is generated based on the spectrum domain structure optimization analysis of the multi-layer energy coupling graph;

[0010] Based on the power coupling network, a double-layer distributed robust optimization analysis is performed to generate a wind-solar-hydropower-storage comprehensive scheduling strategy.

[0011] In a second aspect, the application further provides a double-layer robust optimization scheduling strategy construction system for wind-solar-hydropower-storage, comprising:

[0012] An acquisition unit is configured to acquire first information, second information and third information, wherein the first information includes wind speed and unit operation state data of a wind farm in a target region, and irradiance and module temperature data of a photovoltaic power station, the second information includes inflow runoff data, reservoir capacity-water level data, historical water release-water level response observation data and downstream ecological flow constraint data of a cascade hydropower station.

[0013] A prediction unit is configured to perform feature extraction and integrated prediction analysis based on the first information to generate power prediction results of wind power generation and photovoltaic power generation.

[0014] An analysis unit is configured to perform water power coupling modeling analysis based on the power prediction results of the wind power generation and the photovoltaic power generation and the second information to obtain a power prediction sequence and a mobilizable water quantity sequence of the hydropower station.

[0015] A modeling unit is configured to construct a multi-layer energy coupling graph based on the power prediction results of the wind power generation and the photovoltaic power generation, the power prediction sequence and the mobilizable water quantity sequence of the hydropower station, and perform spectrum domain structure optimization analysis based on the multi-layer energy coupling graph to generate a power coupling network.

[0016] An optimization unit is configured to perform double-layer distributed robust optimization analysis based on the power coupling network to generate a wind-solar-hydropower-storage comprehensive scheduling strategy.

[0017] The application has the following beneficial effects:

[0018] The present application can effectively avoid the operation risk of the sensitive ecological period, significantly improve the water resource utilization efficiency and renewable energy consumption capacity, and simultaneously consider the economy and safety of system operation, thereby solving the problems of power abandonment, increased water consumption and ecological disturbance caused by time sequence mismatch in the traditional method; the present application also adopts a double-layer distributed robust optimization architecture to simultaneously optimize the long-term operation economy, cascade water consumption and ecological risk probability, and finally generate a coordinated scheduling strategy across the hour scale and day scale, so as to realize the unity of system operation economy and efficient use of water resources under the guarantee of ecological safety constraints.

[0019] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent from the description, or can be learned by practice of the present application according to the embodiments. The objects and other advantages of the present application can be achieved and obtained by the structure particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows, and it should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation to the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of the drawings.

[0021] Figure 1 The flow chart of the wind, light, water storage double-layer robust optimization scheduling strategy construction method described in the embodiments of the present application;

[0022] Figure 2 The system structure schematic diagram of the wind, light, water storage double-layer robust optimization scheduling strategy construction method described in the embodiments of the present application.

[0023] Marked in the figure: 701, acquisition unit; 702, prediction unit; 703, analysis unit; 704, modeling unit; 705, optimization unit. DETAILED DESCRIPTION

[0024] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application but not all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0025] It should be noted that similar reference numerals and letters refer to similar items throughout the drawings, and thus, once an item is defined in one drawing, it is not necessary to further define and explain it in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", and the like are only used to distinguish descriptions, and cannot be understood as indicating or implying relative importance. Embodiments

[0026] The embodiment provides a double-layer robust optimization scheduling strategy construction method for wind-solar-water storage.

[0027] Referring to Figure 1 , the method includes steps S1, S2, S3, S4 and S5.

[0028] In step S1, first information, second information and third information are acquired, the first information includes wind speed and unit operation state data of a wind farm in a target area, and irradiance and module temperature data of a photovoltaic power station, the second information includes reservoir inflow data, reservoir capacity-water level data, historical water release-water level response observation data and downstream ecological flow constraint data of a cascade hydropower station, and the third information includes water demand data of a target area.

[0029] Understandably, this step, by acquiring multi-source heterogeneous data, provides comprehensive and reliable data support for subsequent high-precision modeling and optimization decisions. Unlike traditional scheduling systems that only collect conventional data such as power and water level, this step specifically collects minute-level / high temporal resolution raw meteorological data (such as wind speed and irradiance) and equipment status data (such as unit operating status and module temperature), rather than simply using readily available power data. This allows subsequent models to capture the short-term fluctuation characteristics of wind and solar power output and the impact of equipment performance degradation from the source. This step also acquires dynamic coupling data of the hydropower link, including not only the static reservoir capacity-water level relationship, but more importantly, historical water release-water level response observation data. This provides the only data foundation for subsequent accurate modeling of hydrodynamic lag response. This step uploads the time-series data of downstream ecological flow constraints, along with system operation data such as electricity demand and power curtailment records, to the database, laying the data foundation for introducing ecological dimension constraints into scheduling decisions.

[0030] Specifically, for example, in a cascade hydropower and wind-solar hybrid power base in Southwest China, monitoring equipment such as meteorological stations, water level gauges, and flow sensors, as well as an automatic water situation monitoring system, are deployed to collect real-time data on wind speed per second at the wind farm, irradiance and backplane temperature of the photovoltaic array per minute, reservoir water level changes at the hydropower station every 5 minutes, and ecological flow monitoring values ​​of the downstream river every 15 minutes.

[0031] Step S2: Based on the first information, perform feature extraction and integrated prediction analysis to generate power prediction results for wind power generation and photovoltaic power generation;

[0032] It is understandable that this step utilizes a variational autoencoder (VAE) to fuse and reduce the dimensionality of the high-dimensional heterogeneous data in the first information. Its unique advantage lies in its ability to learn the complex nonlinear coupling relationship between wind speed, irradiance and equipment operating status in an unsupervised manner, and generate a low-dimensional "wind-solar joint feature representation" containing rich spatiotemporal correlation information. This provides an unprecedented decision-making basis for downstream hydropower stations to determine whether to start the units and in what way (such as continuous full power generation or intermittent peak shaving) to balance this fluctuation. In this step, step S2 includes steps S21 and S22.

[0033] Step S21: Extract the spatiotemporal correlation features from the wind speed and unit operating status data of the wind farm and the irradiance and module temperature data of the photovoltaic power station through the variational autoencoder, and generate a joint wind-solar feature representation.

[0034] It can be understood that this step adopts a model of a variational autoencoder (VAE) to fuse and compress the features of the two different renewable energies of wind and light, so as to extract a low-dimensional latent variable (i.e., a wind-light joint feature representation) capable of uniformly representing the spatiotemporal dynamic characteristics thereof. This is essentially different from the traditional method of simply splicing data or separately processing wind and light data. The variational autoencoder (VAE) can unsupervisedly learn the complex nonlinear spatiotemporal correlation in the data, for example, the spatial correlation of wind speed (the difference in wind speed at different wind turbine sites), the temporal coupling relationship between irradiance and module temperature (temperature rise may lead to a decrease in photovoltaic panel efficiency), and the potential influence of the operating state of a wind turbine (such as shutdown for maintenance) on the overall output.

[0035] In step S22, a nonlinear mapping relationship between wind-light output and meteorological elements is established according to the wind-light joint feature representation and a temporal convolution network, and a prediction bias is corrected based on preset historical power curtailment record data, to obtain a power prediction result of wind power generation and photovoltaic power generation.

[0036] It can be understood that the time sequence convolution network (TCN) is used in this step to learn the complex nonlinear mapping between meteorological elements and output from the wind and light combined feature representation, and the historical power cut record, a system operation feedback data, is innovatively introduced to correct the prediction deviation, and finally a high-precision power prediction result is generated which conforms to the actual operation constraints. This step does not isolate the prediction model from the actual operation of the system, but embeds the historical power cut record as prior knowledge into the model correction link. The time sequence convolution network includes a causal dilated convolution layer, a residual block and a final output layer. The causal dilated convolution layer is used to ensure that the model only uses information before the preset time when predicting the time step, avoiding data leakage. By exponentially increasing the dilated factor, the number of layers required to cover the same length of historical sequence is greatly reduced, but the receptive field is greatly expanded, which enables the model to efficiently capture long-range dependencies from minutes to hours, such as identifying the influence of weather elements (such as cloud movement, rising wind speed) that last for several hours on wind and light output. Then, each residual block contains multiple causal dilated convolution layers, weight normalization, ReLU activation function and Dropout layer. The residual connection not only alleviates the gradient vanishing problem of deep network, but also allows the network to learn the residual mapping between input and output, improving the training stability and model performance. Finally, the final output layer contains two parallel convolution layers to output the parameters of the probability distribution (this step is the mean and variance of the Gaussian distribution). The input in this step is the "wind and light combined feature representation" sequence, which usually has a time resolution of minutes or 15 minutes, and the historical time window length is 2 days. The output is the probability distribution parameters (mean and variance) of the power value at each prediction time point (every 15 minutes) in the next 2 days. The dilated factor sequence uses an exponential growth sequence with a base of 2 to ensure that the network receptive field can cover a long enough historical period. The convolution kernel size is 3 or 5, and the invention uses Dropout and weight normalization to prevent overfitting.

[0037] Step S3, based on the wind power and photovoltaic power prediction results and the second information, water power coupling modeling analysis is performed to obtain a power prediction sequence and a mobilizable water quantity sequence of the hydropower station.

[0038] It can be understood that the step firstly adopts a gated recurrent unit to fuse the probability distribution characteristics (mean and variance) of wind and light power prediction with the reservoir inflow time series data in the second information to generate a "water-power joint input feature", which not only contains water quantity information, but also embeds the uncertainty brought by wind and light fluctuations. Subsequently, this feature is input into an encoder-decoder network with attention mechanism, which learns a large number of historical water release-water level response observation data in the second information to establish a nonlinear mapping relationship between the "water release instruction" and the "downstream water level change in the future multiple periods". The step generates a dynamic capacity sequence that can be directly used for optimization scheduling at the level of hydropower stations, which quantifies the impact of wind and light fluctuations. It converts unstable wind and light resources into clear and reliable regulation capacity of the hydropower system. In the step S3, the step S3 includes the step S31, the step S32 and the step S33.

[0039] Among them, taking the response of a certain cascade hydropower station in southwest China to the sudden increase of wind power as an example: the present application will receive information that the wind power output may rise sharply (probability prediction) in the next 2 hours, and simulate how the reservoir water level will rise (increase water head) and how long the time lag will be if the hydropower station reduces water release at this time to create space for consumption. The model will finally output the upper limit of the power generation of the hydropower station every 15 minutes in the future (for example: 1 hour later, due to the rise of water level, the power generation can increase to 105% of the rated capacity) and the range of water quantity that can be dynamically adjusted in the process under the premise of meeting the downstream ecological flow.

[0040] The step S31, according to the wind power generation and photovoltaic power generation power prediction result and the inflow runoff data of cascade hydropower station, carries out water power coupling feature extraction processing, wherein the output probability prediction result and the inflow runoff time series data are fused by a gated recurrent unit to generate an inflow runoff-power joint input feature;

[0041] It can be understood that the gated recurrent unit in this step reads two input sequences simultaneously: one is the wind and light output prediction sequence containing the power probability distribution parameters (mean and variance) of each future time point, which quantifies the volatility and uncertainty of renewable energy; the other is the actual historical synchronous in- reservoir runoff time series data provided in the second information. Through its gating mechanism, the gated recurrent unit can identify complex patterns such as "when the wind and light output probability prediction shows high-frequency fluctuations in the next 2 hours, how does it affect the reliability of the current runoff prediction" or "how does a sustained wind power generation event superimpose on the seasonal dry period runoff", and finally outputs a low-dimensional, dense "in- reservoir runoff- power generation joint input feature" vector. This joint feature not only contains the original data information, but more importantly, encodes the coupling relationship between wind and light uncertainty and hydrological dynamics, which is very suitable as input for downstream time series models. The gated recurrent unit includes a reset gate control step that determines how much past state should be combined with the current input to capture short-term dependencies or abrupt changes (such as sudden changes in wind and light or sudden increases in runoff), a candidate hidden state calculation step that generates a new candidate state based on the past information filtered by the reset gate and the current input, an update gate control step that determines how much past state should be retained and how much new candidate state should be injected to effectively capture long-term dependencies (such as the persistence of wind and light output and the daily / seasonal cycles of runoff), and a current hidden state update step that outputs the features that integrate historical and current information.

[0042] Step S32, time lag response modeling processing is performed according to the in- reservoir runoff- power generation joint input feature, an encoder-decoder network is used to learn the mapping relationship between historical water release observation data and water level response observation data, and a water head evolution sequence is output;

[0043] It can be understood that the role of the encoder in this step is to compress the input joint feature sequence (containing wind and light, runoff information) into a context vector that contains the entire sequence information, and the core is to capture pattern features such as "a sustained wind power generation event superimposed on the dry period runoff". The decoder then generates water head values at each future time point based on the context vector. In practical applications, for example, in a certain cascade hydropower and wind-solar complementary base in the southwest, when the model receives a "future 6 hours of sustained high wind power output" joint input feature, the decoder will gradually deduce that if the water release of power station A is reduced at this time to absorb wind power, the reservoir water level of power station A will start to rise (water head increases) after 1 hour, and the in- reservoir flow of its downstream power station B will be reduced, thereby causing the water head of power station B to decrease after 3 hours. The model can output the minute-level or hour-level change sequence of the water head of each power station in this entire dynamic process. This step clearly quantifies the lagging effect of dispatching decisions, providing essential dynamic boundary conditions for the final calculation of the actual power generation capacity and adjustable water volume of the power station.

[0044] Step S33, performing power generation and adjustable water volume collaborative calculation processing according to the water head evolution sequence, wherein the maximum power generation in all preset periods is simulated, the storage capacity-water level data and downstream ecological flow constraint data are coupled to calculate the adjustable water volume range, and finally the power generation prediction sequence and the adjustable water volume sequence of the hydropower station are obtained.

[0045] It can be understood that this step is based on the water head evolution sequence, and the maximum technical output of the hydropower station at each time is calculated according to the dynamic characteristic curve (nonlinear water head-output-flow relationship) of the water turbine. By solving in all preset periods, it is ensured that the results meet the engineering constraints such as unit vibration area and output climbing rate. For example, when the water head rises, the maximum technical output may exceed the rated capacity, and the model of the application can accurately capture this characteristic. The application closely combines physical, ecological constraints and power generation potential, provides clear, reliable and directly executable boundary conditions for subsequent multi-energy complementary optimization, and avoids the risk of violating system safety or ecological constraints of the dispatching instruction. For example, in a certain cascade hydropower and wind-solar complementary base in the southwest, the water head dynamic prediction sequence for the next 6 hours is received. The model calculates that in order to balance the wind power fluctuation, the power station can quickly increase the output to 2200 million kilowatts (based on the maximum capacity under the current water head) within the next 1 hour, but at the same time, in order to meet the constraint that the ecological flow of the Jingjiang section downstream is not less than 5000 cubic meters per second, the water volume for power generation in this period is limited to a certain range (the discharge volume needs to be between 8000 and 10000 cubic meters per second). Finally, it outputs an accurate sequence that gives both the maximum power generation capacity and the water volume adjustment boundary.

[0046] Step S4, constructing a multi-layer energy coupling graph based on the power prediction results of the wind power generation and the photovoltaic power generation, the power generation prediction sequence and the adjustable water volume sequence of the hydropower station, and performing spectral domain structure optimization analysis based on the multi-layer energy coupling graph to generate a power coupling network.

[0047] It can be understood that this step breaks through the point-to-point connection mode considering only power balance in the traditional energy dispatching model, and first constructs a three-dimensional space-time heterogeneous graph containing minute, hour, and day multi-layer time scales, and the edge attributes integrate time lag, power transmission capacity, and ecological sensitivity. The construction process is as follows: first, the wind farm, photovoltaic power station, each cascade hydropower station, and energy storage unit are mapped to the nodes in the graph, and the characteristics of the nodes are defined by the wind power prediction value, the water power generation capacity, and the adjustable water range; then, the edges between the nodes are defined according to the physical connection (power grid line, waterway) and the operation constraint, and the weight attribute of the edge integrates the maximum rate of energy transmission (power transmission capacity), the time required for energy transmission from one end to the other end (time lag), and the potential impact of the transmission behavior on the downstream ecology (ecological sensitivity). In this step, step S4 includes step S41, step S42, and step S43.

[0048] Step S41, the graph structure modeling method maps the wind farm, photovoltaic power station, hydropower station, and energy storage unit to heterogeneous nodes, and constructs an initial multi-layer energy coupling graph containing multiple time scales;

[0049] It can be understood that this step gives each node a dynamic multi-dimensional attribute state: the attributes of wind power nodes and photovoltaic nodes come from power prediction results, which not only contain expected output values, but also contain probability distribution parameters representing fluctuation uncertainty; the attributes of hydropower station nodes are defined by power generation prediction sequence and adjustable water sequence, which contain not only power generation capacity, but also upper and lower bound constraints of regulation capacity; the attributes of energy storage nodes are preset, which include rated power, capacity, and current state of charge. These node attributes will be dynamically updated as the prediction time window rolls over.

[0050] Among them, the edges between the nodes are given three-dimensional quantitative attributes: the first is time lag, which is based on historical observation and water flow propagation data, and quantifies the time required for energy transmission from one end to the other end (for example, the delay time of water release from an upstream reservoir to the reservoir of a downstream power station); the second is power transmission capacity, which is determined by the rated capacity of power grid lines or the water carrying capacity of rivers, and defines the physical transmission limit; the third is ecological sensitivity, which is based on downstream ecological flow constraint data, and quantifies the impact of the transmission behavior represented by the edge on the ecology in a certain period (for example, in the fish spawning period, the sensitivity coefficient of the downstream flow fluctuation will increase significantly). The finally constructed initial multi-layer energy coupling graph actually contains the topological structure of multiple time layers such as minutes, hours, and days, thereby describing the space-time coupling relationship of wind farms, photovoltaic power stations, hydropower stations, and energy storage units.

[0051] For example, in a wind farm, a photovoltaic power station, a hydropower station and an energy storage unit in a certain cascade hydropower and wind-solar complementary base in the southwest, an initial multi-layer graph is constructed, wherein the edge attribute from a wind farm node to a hydropower station node is set as: time lag = 45 minutes (considering the water conduction time), power transmission capacity = 2000 MW (based on the grid line limit), ecological sensitivity = 0.8 (the current period is the dry season and is in the ecological sensitive window).

[0052] In step S42, a spectral domain structure optimization process is performed based on the initial multi-layer energy coupling graph to perform feature decomposition and redundant edge pruning on the connection relationship between the nodes of the initial multi-layer energy coupling graph, and an optimized energy coupling graph is generated.

[0053] It can be understood that in this step, firstly, graph Laplace transformation is performed on the initial multi-layer energy coupling graph to map the connection relationship of the graph in the spatial domain to the spectral domain (frequency domain). By calculating the Laplace matrix of the graph and performing eigenvalue decomposition, a set of eigenvalues and eigenvectors representing the smoothness of the graph connection are obtained. Among them, the eigenvectors corresponding to the smaller eigenvalues represent the low-frequency components (main energy transmission paths) in the graph that change gently and are closely connected, while the high-frequency components corresponding to the larger eigenvalues often represent the connection relationships in the graph that change dramatically, are redundant or have more noise. Subsequently, based on the spectral decomposition result, spectral filtering and redundant edge pruning are performed. By setting a threshold, the connection edges corresponding to the low-frequency components that contribute most to the system are retained, while the redundant edges (such as some standby lines that exist physically but are rarely used in actual scheduling or meaningless long-distance coupling) corresponding to the high-frequency components that have weak influence on the overall energy transmission and balance are pruned. This step significantly reduces the complexity of the subsequent optimization model, improves the calculation efficiency, and not only does not lose key information, but also highlights the most important and essential spatiotemporal energy coupling relationships in the system through mathematical methods, laying a reasonable structural foundation for subsequent generation of high-quality multi-energy complementary power coupling networks.

[0054] For example, in the initial graph of a certain cascade hydropower and wind-solar complementary base in the southwest, dozens of wind farms, photovoltaic power stations and hydropower stations are connected, and the relationship between the nodes is complex. After spectral domain analysis, it is found that the connection edge between a certain wind farm group and a downstream large reservoir power station (although the transmission distance is far and the time lag is 1 hour) is a low-frequency feature in the spectral domain, which reveals that this path is a key long-distance hydropower compensation channel for smoothing wind power fluctuations, and therefore is retained. The edge between another closer pumped storage power station and a certain photovoltaic power station is pruned as a redundant connection because its capacity is extremely small and is rarely used in historical scheduling, and it is a high-frequency noise in the spectral domain.

[0055] Step S43, embedding the power prediction results of wind power and photovoltaic power, the power prediction sequence of the hydropower station and the mobilizable water quantity sequence into network nodes through a graph neural network, calculating edge weights based on downstream ecological flow constraint data, and generating a multi-energy power coupling network containing wind energy, electric energy and water energy.

[0056] It can be understood that this step takes the optimized energy coupling graph as the initial topology, and then takes the power prediction results of wind power and photovoltaic power, the power prediction sequence of the hydropower station and the mobilizable water quantity sequence as dynamic input features of each corresponding node. Through its message passing mechanism, the graph neural network aggregates information from neighbor nodes and combines its own dynamic attributes at each prediction time step (such as rolling execution every 15 minutes), and updates the embedding representation of all nodes. This process makes the state of each node in the graph no longer isolated, but a deep representation that integrates the spatiotemporal correlation information of the entire network. This step provides a dynamic mathematical model that not only represents physical connections but also embeds risk costs for subsequent double-layer optimization, ensuring that the final generated scheduling strategy can actively avoid ecological sensitive operations while pursuing economy.

[0057] Step S5, performing double-layer distributed robust optimization analysis based on the power coupling network to generate a wind-solar-hydro-storage comprehensive scheduling strategy.

[0058] It can be understood that the upper-layer optimization of this step is a slow-time-scale problem with a scale of “week” or “month”, which is based on the entire coupling network and takes the minimization of system long-term expected operation cost and the minimization of ecological risk measure as the joint objective function. It uses stochastic programming to handle the long-term uncertainty of wind and solar output, and outputs a capacity configuration and risk control strategy for multiple future scenarios in a preset time period of months, for example: allocating more regulation capacity reserves to the hydropower station to cope with the autumn wind period, and setting stricter scheduling margins for ecological sensitive periods. The lower-layer optimization is a time-scale problem with a scale of “day” or “hour”, which receives the strategy of the upper layer as boundary conditions, and based on the updated short-term wind and solar power probability prediction, uses the chance constrained programming method to construct a real-time scheduling sub-problem with the minimization of periodic scheduling probability risk (such as power shedding risk) and the minimization of cascade water consumption as the objective. It is responsible for generating a fine, day-based pre-scheduling scheme, for example: during the photovoltaic output peak period from 2 pm to 4 pm tomorrow, specifically arranging the output curves of each power station to ensure that the ecological flow fluctuation limit given by the upper layer is met. This step decomposes and embeds the long-term ecological risk prevention and control target into the short-term economic operation, enabling the system scheduling to change from passive and responsive operation to proactive and preventive optimization. In this step, step S5 includes step S51, step S52 and step S53.

[0059] Step S51, based on the power coupling network, an upper layer optimization modeling process is performed, wherein a planning model with a joint optimization of expected operation cost and ecological risk measurement as an objective function is constructed by a stochastic programming method, and a preset period capacity configuration and risk control strategy in months are generated based on preset historical off-grid records and ecological sensitivity indexes as constraint conditions;

[0060] It can be understood that this step is based on the topology of the multi-energy complementary power coupling network, and the node attributes (power station capacity, maximum output) and edge attributes (transmission capacity, time delay, ecological sensitivity) of the network define the physical constraints of the system. The objective function aims to minimize the weighted sum of the long-term expected operation cost (including fuel cost, operation and maintenance cost, off-grid penalty, etc.) and the ecological risk measurement. The ecological risk measurement is a preset value based on the corresponding historical data, for example, during the fish migration period, the average value of the historical data allowed daily flow variation amplitude threshold. The objective function is as follows:

[0061]

[0062] Wherein, is the objective function, is the minimum selection, is the expected total operation cost (including fuel, operation and maintenance, and off-grid cost), is the ecological risk measurement, and is a preset weight coefficient for balancing economic and ecological targets.

[0063] Step S52, according to the preset period capacity configuration and risk control strategy in months, a lower layer scheduling scheme is constructed, an opportunity constraint programming method is used to construct a real-time scheduling sub-problem with scheduling probability risk and minimum cascade water consumption as the target, and a wind-solar-water storage pre-scheduling scheme in days is generated;

[0064] It can be understood that this step constructs a two-objective optimization problem: minimizing the scheduling probability risk and minimizing the cascade water consumption. The scheduling probability risk is a comprehensive index that quantifies the occurrence probability and severity of various adverse scenarios (such as serious off-grid, violation of ecological flow) that may be faced when executing the daily plan; the cascade water consumption minimization reflects the pursuit of economic utilization of hydraulic resources. The opportunity constraint programming is used to handle the randomness in wind and solar power prediction, which allows the constraint condition to be violated at a certain confidence level, for example, "the probability of off-grid rate exceeding 5% should not be higher than 10%" or "the probability of downstream flow exceeding the ecological threshold should not be higher than 5%", and the target optimization formula is as follows:

[0065]

[0066] wherein the spill rate constraint formula and the downstream ecological flow constraint formula are shown as follows:

[0067] wherein, is the scheduling decision variable, and is the multi-objective weight coefficient, is the scheduling probability risk comprehensive index, is the scheduling decision variable set, is an index variable for traversing and referring to each specific hydropower station in the cascade, is the water consumption of the i-th hydropower station, is the probability of the event occurring, is the period spill power, is the period available renewable energy power, is the spill rate threshold value, is the spill risk confidence level, is the period downstream flow, is the ecological flow allowable range, is the ecological overrun risk confidence level, is the set of time periods in which the event occurs, is the set of ecological sensitive periods.

[0068] Step S53, according to the preset period capacity configuration, the risk control strategy and the day-based wind-solar-hydro-storage pre-scheduling scheme, a double-layer collaborative optimization processing is performed to generate a comprehensive scheduling strategy of the wind-solar-hydro-storage system.

[0069] It can be understood that this step first “reports” the day-scale pre-scheduling scheme generated by the lower layer to the upper layer model, and the upper layer model will check whether the scheme completely meets the monthly capacity and risk control strategy (for example, whether the maximum water release in the day plan breaks the upper limit of the flow fluctuation set for the ecological sensitive period in the monthly plan). If the check fails, the upper layer will pass its long-term goal (such as the ecological risk measure) to the lower layer, and the lower layer will adjust its objective function (for example, increase the penalty weight for violating the ecological constraint) accordingly to re-solve the day plan. This process will be iterated until the upper and lower layer schemes reach a “consensus”: that is, the day plan can meet the short-term economic and operation requirements, and completely comply with the long-term strategic risk control boundary. The final output strategy is a cross-time scale strategy set containing multiple scenarios and corresponding scheduling instructions. Embodiments

[0070] As Figure 2 ​As shown, this embodiment provides a two-layer robust optimization scheduling strategy construction system for wind, solar, hydro, and storage. The system includes an acquisition unit 701, a prediction unit 702, an analysis unit 703, a modeling unit 704, and an optimization unit 705.

[0071] The acquisition unit 701 is used to acquire first information, second information and third information. The first information includes wind speed and unit operation status data of wind farms in the target area, and irradiance and module temperature data of photovoltaic power stations. The second information includes inflow runoff data, reservoir capacity-water level data, historical water release-water level response observation data and downstream ecological flow constraint data of cascade hydropower stations.

[0072] The prediction unit 702 is used to perform feature extraction and integrated prediction analysis based on the first information to generate power prediction results for wind power generation and photovoltaic power generation.

[0073] Analysis unit 703 is used to perform hydrodynamic coupling modeling analysis based on the power prediction results of wind power generation and photovoltaic power generation and the second information to obtain the power generation prediction sequence and the adjustable water volume sequence of the hydropower station.

[0074] Modeling unit 704 is used to construct a multi-layer energy coupling diagram based on the power prediction results of wind power generation and photovoltaic power generation, the power generation prediction sequence of hydropower station and the adjustable water volume sequence, and to perform spectral domain structure optimization analysis based on the multi-layer energy coupling diagram to generate a power coupling network.

[0075] The optimization unit 705 is used to perform a two-layer distributed robust optimization analysis based on the power coupling network to generate a comprehensive scheduling strategy for wind, solar, hydro, and storage.

[0076] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0077] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0078] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for constructing a robust optimization scheduling strategy of wind-solar-water storage dual-layer, characterized in that, The method comprises the following steps: acquiring first information and second information, the first information comprising wind speed and unit operating state data of a wind farm in a target area, and irradiance and module temperature data of a photovoltaic power station, and the second information comprising reservoir inflow data of a cascade hydropower station, reservoir capacity-water level data, historical discharge-water level response observation data, and downstream ecological flow constraint data; performing feature extraction and integrated prediction analysis based on the first information to generate power prediction results of wind power generation and photovoltaic power generation; performing water power coupling modeling analysis based on the power prediction results of wind power generation and photovoltaic power generation and the second information to obtain a power prediction sequence and a movable water volume sequence of the hydropower station; constructing a multi-layer energy coupling graph based on the power prediction results of wind power generation and photovoltaic power generation, the power prediction sequence and the movable water volume sequence of the hydropower station, and performing spectral domain structure optimization analysis based on the multi-layer energy coupling graph to generate a power coupling network; performing double-layer distributed robust optimization analysis based on the power coupling network to generate a wind-solar-hydro-storage comprehensive dispatching strategy. 2.The method of claim 1, wherein The feature extraction and integrated prediction analysis based on the first information comprises: extracting spatiotemporal correlation features in the wind speed and unit operating state data of the wind farm and the irradiance and module temperature data of the photovoltaic power station through a variational autoencoder to generate a wind-solar combined feature representation; establishing a nonlinear mapping relationship between wind-solar output and meteorological elements according to the wind-solar combined feature representation and a time series convolution network, and correcting prediction bias based on preset historical curtailed power record data to obtain the power prediction results of wind power generation and photovoltaic power generation.

3. The method of claim 1, wherein The water power coupling modeling analysis based on the power prediction results of wind power generation and photovoltaic power generation and the second information comprises: performing water power-electricity coupling feature extraction processing according to the power prediction results of wind power generation and photovoltaic power generation and the reservoir inflow data of the cascade hydropower station, wherein an output probability prediction result and reservoir inflow time series data are fused through a gated recurrent unit to generate reservoir inflow-power generation combined input features; performing time lag response modeling processing according to the reservoir inflow-power generation combined input features, and learning a mapping relationship between historical discharge observation data and water level response observation data using an encoder-decoder network to output a water head evolution sequence; performing power generation power and adjustable water volume collaborative calculation processing according to the water head evolution sequence, wherein the maximum power generation power in all preset time periods is simulated, and the reservoir capacity-water level data and downstream ecological flow constraint data are coupled to calculate the adjustable water volume range, and finally the power prediction sequence and the movable water volume sequence of the hydropower station are obtained.

4. The method of claim 1, wherein The multi-layer energy coupling graph is constructed based on the power prediction results of wind power generation and photovoltaic power generation, the power prediction sequence and the movable water volume sequence of the hydropower station, and the spectral domain structure optimization analysis is performed based on the multi-layer energy coupling graph, which comprises: mapping the wind farm, the photovoltaic power station, the hydropower station, and the energy storage unit into heterogeneous nodes based on a graph structure modeling method to construct an initial multi-layer energy coupling graph containing multiple time scales; Performing spectral domain structure optimization processing based on the initial multi-layer energy coupling graph, decomposing the connection relationship between nodes of the initial multi-layer energy coupling graph, and pruning redundant edges to generate an optimized energy coupling graph; The power prediction results of the wind power generation and the photovoltaic power generation, the power generation power prediction sequence of the hydropower station and the adjustable water volume sequence are embedded into network nodes through a graph neural network, edge weights are calculated based on downstream ecological flow constraint data, and a multi-energy power coupling network containing wind energy, electric energy and water energy is generated.

5. The method of claim 1, wherein Based on the power coupling network, a double-layer distributed robust optimization analysis is performed to generate a wind-solar-hydro-storage comprehensive scheduling strategy, including: Based on the power coupling network, an upper-layer optimization modeling process is performed, in which a planning model with an expected operation cost and ecological risk measurement joint optimization as an objective function is constructed by a stochastic programming method, and a preset time period capacity configuration and risk control strategy in months is generated based on preset historical off-grid records and ecological sensitivity indicators as constraint conditions; According to the preset time period capacity configuration and risk control strategy in months, a lower-layer scheduling scheme is constructed, an opportunity constraint programming method is used to construct a real-time scheduling sub-problem with scheduling probability risk and minimum cascade water consumption minimization as an objective, and a wind-solar-hydro-storage pre-scheduling scheme in days is generated; According to the double-layer cooperative optimization processing of the preset time period capacity configuration and risk control strategy in months and the wind-solar-hydro-storage pre-scheduling scheme in days, a comprehensive scheduling strategy of the wind-solar-hydro-storage system is generated.

6. A wind-solar-water storage double-layer robust optimization scheduling strategy construction system, characterized in that, including: An acquisition unit is configured to acquire first information and second information, the first information including wind speed and unit operating state data of a wind farm in a target region, and irradiance and module temperature data of a photovoltaic power station, and the second information including inflow runoff data, reservoir capacity-water level data, historical water release-water level response observation data and downstream ecological flow constraint data of a cascade hydropower station; A prediction unit is configured to perform feature extraction and integrated prediction analysis based on the first information to generate power prediction results of wind power generation and photovoltaic power generation; An analysis unit is configured to perform hydrodynamic coupling modeling analysis based on the power prediction results of wind power generation and photovoltaic power generation and the second information to obtain a power generation power prediction sequence and an adjustable water volume sequence of a hydropower station; A modeling unit is configured to construct a multi-layer energy coupling graph based on the power prediction results of wind power generation and photovoltaic power generation, the power generation power prediction sequence and the adjustable water volume sequence of the hydropower station, and perform spectral domain structure optimization analysis based on the multi-layer energy coupling graph to generate a power coupling network; An optimization unit is configured to perform a double-layer distributed robust optimization analysis based on the power coupling network to generate a wind-solar-hydro-storage comprehensive scheduling strategy.

7. The wind-solar-water storage double-layer robust optimization scheduling strategy construction system according to claim 6, characterized in that, The prediction unit includes: A first prediction subunit is configured to extract spatiotemporal correlation features in wind speed and unit operating state data of a wind farm, and irradiance and module temperature data of a photovoltaic power station through a variational autoencoder to generate wind-solar joint feature representation. The second prediction unit is configured to establish a nonlinear mapping relationship between wind and light output and meteorological elements according to the wind and light combined feature representation and the time sequence convolution network, and correct prediction deviation based on preset historical power curtailment record data, to obtain power prediction results of wind power generation and photovoltaic power generation.

8. The wind-solar-water dual-layer robust optimization scheduling strategy construction system according to claim 6, characterized in that, The analysis unit comprises: The first analysis subunit is configured to perform hydro-power coupling feature extraction processing according to the power prediction results of wind power generation and photovoltaic power generation and reservoir inflow data of the cascade hydropower station, wherein the output probability prediction results and the inflow time series data are fused through a gating cycle unit to generate inflow-runoff-power combined input features; The second analysis subunit is configured to perform time lag response modeling processing according to the inflow-runoff-power combined input features, learn a mapping relationship between historical water release observation data and water level response observation data by using an encoder-decoder network, and output a water head evolution sequence; The third analysis subunit is configured to perform power generation and adjustable water volume collaborative calculation processing according to the water head evolution sequence, wherein the maximum power generation in all preset time periods is simulated, and the reservoir capacity-water level data and downstream ecological flow constraint data are coupled to calculate the adjustable water volume range, so as to finally obtain a power prediction sequence and an adjustable water volume sequence of the hydropower station. 9.The system according to claim 6, wherein, The modeling unit comprises: The first modeling subunit is configured to map wind farms, photovoltaic power stations, hydropower stations and energy storage units into heterogeneous nodes based on a graph structure modeling method, and construct an initial multi-layer energy coupling graph containing multiple time scales; The second modeling subunit is configured to perform spectral domain structure optimization processing based on the initial multi-layer energy coupling graph, decompose the connection relationship between nodes of the initial multi-layer energy coupling graph, and prune redundant edges to generate an optimized energy coupling graph; The third modeling subunit is configured to embed the power prediction results of wind power generation and photovoltaic power generation, the power prediction sequence and the adjustable water volume sequence of the hydropower station into network nodes by using a graph neural network, calculate edge weights based on downstream ecological flow constraint data, and generate a multi-energy power coupling network containing wind energy, electrical energy and water energy.

10. The system according to claim 6, wherein, The optimization unit comprises: The first optimization subunit is configured to perform upper-layer optimization modeling processing based on the power coupling network, wherein a planning model with an expected operation cost and ecological risk metric joint optimization as an objective function is constructed by using a stochastic programming method, and preset historical power curtailment records and ecological sensitivity indicators are used as constraint conditions to generate a preset time period capacity configuration and risk control strategy in months; The second optimization subunit is configured to construct a lower-layer scheduling scheme according to the preset time period capacity configuration and risk control strategy in months, construct a real-time scheduling sub-problem with scheduling probability risk and minimum cascade water consumption minimization as an objective by using an opportunity constraint programming method, and generate a wind, light, water and storage pre-scheduling scheme in days; The third optimization subunit is configured to perform double-layer collaborative optimization processing according to the preset time period capacity configuration in months, the risk control strategy, and the wind, light, water and storage pre-scheduling scheme in days, to generate a comprehensive scheduling strategy of the wind, light, water and storage system.

Images