Water-light-storage multi-energy complementary collaborative scheduling method based on double-mispeak compensation

Abstract

The application discloses a water-light-storage multi-capability complementary collaborative scheduling method based on double off-peak compensation, comprising the following steps: constructing multiple preset scheduling models including a water-light independent scheduling model, a water-light complementary scheduling model and a water-light-storage complementary scheduling model; collecting photovoltaic predicted output data, water power storage flow data and initial state data of a pumped storage power station in a current period in real time; determining a target scheduling mode in the current period from the multiple preset scheduling models by using a preset conversion efficiency evaluation criterion based on the photovoltaic predicted output data and the initial state data; and solving the preset scheduling model corresponding to the target scheduling mode based on the photovoltaic predicted output data and the water power storage flow data, to generate day-ahead scheduling instructions containing power station output plans. The application dynamically adjusts the complementary strategy by sensing the pumped storage power load state and the power grid pressure, thereby improving the photovoltaic consumption rate and the system power generation capacity.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching technology, and in particular to a multi-energy complementary and coordinated dispatching method based on dual peak-shaving compensation of hydropower, solar power, and energy storage. Background Technology

[0002] Constructing a new power system with new energy sources as the mainstay has become a development trend. Among various renewable energy sources, utilizing hydropower and pumped storage power stations with strong regulation capabilities to complement photovoltaic power with high volatility is an important means to mitigate the randomness of new energy sources and improve the grid's absorption capacity. Researching efficient coordinated dispatch mechanisms for hydro-photovoltaic-storage hybrid energy systems has significant engineering value for ensuring the safe and stable operation of the power grid and improving the utilization rate of clean energy.

[0003] Currently, research on the scheduling of hydro-solar-storage complementary systems mainly focuses on capacity allocation and joint operation strategy optimization. Existing technologies typically construct unified optimization scheduling models covering multiple types of power sources based on deterministic or stochastic programming theories, and solve them directly using large-scale nonlinear programming algorithms. At the operation control level, strategies based on fixed priority rules or single economic indicators (such as fixed electricity price differences or fixed comprehensive cycle efficiency) are mainly used to allocate the output share among hydropower, photovoltaic, and pumped storage to achieve system power balance and absorption targets.

[0004] However, existing dispatching schemes generally suffer from rigid decision-making mechanisms that are difficult to adapt to dynamic operating conditions, and low computational efficiency in unified solutions. Specifically, existing mode-switching logics often rely on static, fixed thresholds, such as only considering the fixed cycle efficiency loss of pumped storage, ignoring the time-varying impact of the real-time state of charge (SOC) of pumped storage power stations and the urgency of grid peak shaving on complementary benefits. The uniform static criteria result in a lack of adaptability in dispatching strategies: when pumped storage capacity is tight or grid peak shaving pressure is high, decisions are still mechanically made based on fixed efficiency, leading to idle or overdrawn pumped storage resources, making it difficult to maximize power generation. Furthermore, traditional methods lack rapid predictive mechanisms, directly performing overall optimization on a complex model with all elements, resulting in high computational dimensionality and slow convergence speed, making it difficult to meet the timeliness requirements of day-ahead dispatching. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-energy complementary and coordinated scheduling method based on dual peak-shifting compensation for water, solar and energy storage, in order to solve at least one of the aforementioned problems in the prior art.

[0006] According to one aspect of this application, a method for coordinated scheduling of hydro-solar-storage multi-energy complementary systems based on dual peak-shifting compensation includes:

[0007] Based on the pre-configured physical parameters of each power station, a variety of pre-set scheduling models are constructed, including independent water-solar scheduling model, water-solar complementary scheduling model, and water-solar-storage complementary scheduling model.

[0008] Real-time collection of photovoltaic power output forecast data, hydropower inflow data, and initial status data of pumped storage power stations for the current time period;

[0009] Based on photovoltaic power output forecast data and initial state data, and using preset conversion efficiency evaluation criteria, the target scheduling mode for the current period is determined from multiple preset scheduling models.

[0010] Based on the photovoltaic power output forecast data and hydropower inflow data, the preset scheduling model corresponding to the target scheduling mode is solved, and the day-ahead scheduling instructions containing the power output plans of each power station are generated.

[0011] Among them, the conversion efficiency evaluation criteria are used to evaluate the conversion relationship between hydropower regulation losses and photovoltaic consumption increments under different dispatch modes.

[0012] According to one aspect of this application, multiple preset scheduling models include a water-solar independent scheduling model, a water-solar complementary scheduling model, and a water-solar-storage complementary scheduling model.

[0013] Among them, the independent scheduling model of hydropower and solar power takes the optimal power generation of the hydropower station as the objective function;

[0014] The hydro-solar complementary dispatch model takes the maximum total power generation of the system during the dispatch period as the objective function, and includes water balance constraints and power transmission channel constraints.

[0015] The hydro-solar-storage complementary scheduling model adds energy balance constraints, water level constraints, and output constraints for pumped storage power stations to the existing hydro-solar complementary scheduling model.

[0016] According to one aspect of this application, the incremental photovoltaic power consumption assessed by the preset conversion efficiency assessment criteria is determined by calculating the complementary power generation of the system.

[0017] The system complementary power generation increase is defined as: the increase in the total power generation of the system obtained by solving the preset scheduling model corresponding to the target scheduling mode, relative to the total power generation of the system obtained by solving the independent scheduling model of water and solar power.

[0018] Solving for the preset scheduling model corresponding to the target scheduling mode includes performing dual peak-shifting compensation:

[0019] Based on photovoltaic power generation forecast data, the power generation schedule of cascade hydropower stations is adjusted to avoid the peak of photovoltaic power generation.

[0020] During periods when photovoltaic power curtailment is detected, the pumped storage power station is controlled to operate in pumping mode to absorb the curtailed photovoltaic power.

[0021] According to one aspect of this application, the hydropower regulation loss involved in the conversion efficiency assessment criteria is quantified as the hydropower compensation photovoltaic power loss rate;

[0022] The hydropower compensation for photovoltaic power loss rate is defined as: the reduction in hydropower generation caused by the consumption of a unit of photovoltaic curtailment, and is calculated as the ratio of the reduction in hydropower generation in the complementary dispatch mode to the increase in photovoltaic grid-connected power generation in the independent dispatch mode.

[0023] The target scheduling mode for the current time period is determined from multiple preset scheduling models, including the application of static decision logic based on preset fixed thresholds:

[0024] Determine whether the power loss rate of hydropower compensating for photovoltaic power is lower than a preset fixed threshold;

[0025] If so, the target scheduling mode will be determined as the water-solar complementary scheduling model;

[0026] If not, or if it is determined that there is photovoltaic curtailment in the current period, then the target scheduling mode will be determined as the hydro-solar-storage complementary scheduling model.

[0027] According to one aspect of this application, based on photovoltaic power generation prediction data and initial state data, a target scheduling mode for the current time period is determined from multiple preset scheduling models, including a three-stage progressive scheduling process of pre-judgment, hierarchical optimization, and coordinated adjustment.

[0028] Phase 1: Based on the photovoltaic power output forecast data and the hydropower operation status under the independent hydropower dispatch mode obtained by solving the independent hydropower dispatch model, calculate the estimated hydropower marginal loss rate;

[0029] Phase 2: Calculate the dynamic loss rate threshold for the current period based on the initial state data, and determine the target scheduling mode based on the comparison between the estimated hydropower marginal loss rate and the dynamic loss rate threshold.

[0030] Phase 3: Solve the preset scheduling model corresponding to the target scheduling mode to obtain the output allocation results of each power station.

[0031] According to one aspect of this application, the estimated marginal loss rate of hydropower is calculated, including:

[0032] The predicted amount of photovoltaic curtailment is determined based on the photovoltaic power output data and the capacity of the power transmission channels.

[0033] Based on the hydropower operation status and the pre-stored optimal hydropower operation efficiency curve, the output adjustment loss coefficient is calculated, which characterizes the degree to which the actual hydropower operation efficiency deviates from the optimal operation efficiency.

[0034] Based on the predicted amount of photovoltaic curtailment, the power output adjustment loss coefficient, and the power output data under the independent hydropower dispatch mode, the estimated marginal loss rate of hydropower is calculated without running the hydropower-solar complementary dispatch model.

[0035] According to one aspect of this application, determining a target scheduling mode includes executing the following logic:

[0036] Determine whether the estimated marginal loss rate of hydropower is less than the dynamic loss rate threshold, and at the same time determine whether the total predicted amount of photovoltaic curtailment is less than the pumped storage capacity of the pumped storage power station.

[0037] If both judgment conditions are met simultaneously, the target scheduling mode is determined to be the hydro-solar complementary scheduling model; otherwise, the target scheduling mode is determined to be the hydro-solar-storage complementary scheduling model.

[0038] According to one aspect of this application, the dynamic loss rate threshold is obtained by applying a dual-factor correction to the pumped storage benchmark loss rate. The calculation formula is the product of the pumped storage benchmark loss rate, the capacity availability coefficient, and the peak shaving urgency coefficient. The pumped storage benchmark loss rate is determined based on the comprehensive cycle efficiency of the pumped storage power station.

[0039] The capacity availability factor characterizes the available regulating capacity status of a pumped storage power station at the current moment. Its value is negatively correlated with the current stored energy. It is calculated as follows: the difference between the maximum stored energy of the pumped storage power station and the current stored energy, divided by the difference between the maximum stored energy of the pumped storage power station and the minimum stored energy.

[0040] The peak shaving urgency factor characterizes the pressure state of the power grid in accepting photovoltaic power output during the current period. It is calculated by dividing the portion of the predicted photovoltaic power output during the current period that exceeds the capacity of the power transmission channel by the maximum pumping power of the pumped storage power station, and taking the minimum value of the obtained ratio and the value of 1 as the peak shaving urgency factor.

[0041] According to one aspect of this application, a coordinated adjustment step is also included after phase three:

[0042] Based on the solution results of the preset scheduling model corresponding to the target scheduling mode, the actual hydropower marginal loss rate is calculated.

[0043] Determine whether the deviation between the actual and estimated marginal loss rates of hydropower exceeds a preset allowable deviation threshold.

[0044] If so, update the estimated hydropower marginal loss rate using the actual hydropower marginal loss rate, and return to execution phase two to redetermine the target scheduling mode.

[0045] According to one aspect of this application, based on the predicted amount of photovoltaic curtailment, the power output adjustment loss coefficient, and the power output data under the independent hydropower dispatch mode, the estimated marginal loss rate of hydropower is calculated, specifically through the following method:

[0046] Calculate the difference between the output under the independent hydropower dispatch mode and the preset minimum allowable hydropower output, and combine it with the output adjustment loss coefficient to perform time-period integration or accumulation to obtain the estimated reduction in hydropower generation.

[0047] The estimated marginal loss rate of hydropower is obtained by dividing the estimated reduction in hydropower generation by the sum of the predicted amount of solar power curtailment.

[0048] Beneficial effects: This invention improves the photovoltaic absorption rate and system power generation by using a complementary strategy of sensing the pumped storage charge status and dynamically adjusting grid pressure. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the overall process of the water-solar-storage multi-energy complementary coordinated scheduling method based on dual peak-shifting compensation provided in the embodiments of this application.

[0050] Figure 2 This is a schematic diagram of the three-stage progressive scheduling process of execution pre-judgment, hierarchical optimization, and collaborative adjustment provided in the embodiments of this application.

[0051] Figure 3 A schematic diagram illustrating the process of calculating and estimating the marginal loss rate of hydropower provided in this application embodiment.

[0052] Figure 4 This is a schematic diagram of the execution collaborative adjustment steps provided in the embodiments of this application. Detailed Implementation

[0053] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0054] Example 1 illustrates the overall process of a multi-energy complementary and coordinated scheduling method based on dual peak-shifting compensation, such as... Figure 1 As shown, a general system architecture and process framework are provided for subsequent embodiments, describing a general method that is compatible with multiple scheduling modes. It is applicable to both basic scheduling based on fixed rules and advanced scheduling based on adaptive algorithms, which are detailed in subsequent embodiments.

[0055] The physical system upon which this embodiment relies is a hybrid hydro-solar-storage energy base comprising a cascade hydropower station group, a photovoltaic power station cluster, and a pumped-storage power station. Specifically, the system includes cascade hydropower stations R1, R2, and R3 located from upstream to downstream of the river basin, photovoltaic power stations S1 to S4 distributed in the surrounding area, and supporting pumped-storage power stations. The electricity generated by all the power stations is collected and transmitted through an ultra-high-voltage direct current (UHVDC) transmission channel. To achieve optimized system operation, this embodiment employs the following method:

[0056] Step 101: Based on the pre-configured physical parameters of each power station, construct multiple preset scheduling models, including independent water-solar scheduling model, water-solar complementary scheduling model, and water-solar-storage complementary scheduling model.

[0057] In this embodiment, model construction refers to establishing mathematical expressions in a computer system that describe the operational constraints and optimization objectives of each power station. Multiple preset scheduling models cover three different operational scenarios.

[0058] Specifically, the independent hydro-solar dispatch model corresponds to the operational state where each energy source operates independently and is typically used as a benchmark for evaluating complementary benefits. The hydro-solar complementary dispatch model corresponds to the joint operation of hydropower and photovoltaics, utilizing the regulation capabilities of hydropower to mitigate photovoltaic fluctuations. The hydro-solar-storage complementary dispatch model further introduces pumped storage power stations, utilizing their bidirectional regulation capabilities of pumping water to fill valleys and peak power generation to jointly absorb curtailed photovoltaic power with hydropower.

[0059] When constructing the above model, it is necessary to input the physical parameters of each power station, such as the installed capacity, reservoir capacity curve, and unit characteristics of hydropower stations, the installed capacity of photovoltaic power stations, and the reservoir capacity and conversion efficiency of pumped storage power stations. The model is usually constructed using nonlinear programming methods and can be solved using mathematical optimization software such as LINGO (an optimization software) and its built-in generalized reduced gradient method (GRG).

[0060] Step 102: Real-time collection of photovoltaic power output forecast data, hydropower inflow data, and initial status data of pumped storage power station for the current time period.

[0061] In this embodiment, data acquisition is the foundation of day-ahead scheduling. Photovoltaic power output forecast data typically originates from meteorological forecasting systems, including the theoretical photovoltaic power generation sequence for each time period within the next 24 hours. Hydropower inflow data comes from hydrological monitoring systems, including predicted inflow runoff values ​​for each cascade hydropower station. The initial state data for pumped storage power stations mainly refers to the upper reservoir's water storage or energy storage at the start of the scheduling cycle, determining the subsequent regulation capacity boundary of the pumped storage power station. For example, when the system performs day-ahead scheduling calculations at 00:00 each day, it collects the above data as optimization boundary conditions.

[0062] During data acquisition, if data is missing or abnormal, the system employs the following fault-tolerance strategies: For photovoltaic power output forecast data, if data for the current period is missing, it will be filled using data from the previous period or the historical average for the same period; for hydropower inflow data, if communication is interrupted, the most recent valid data will be used and a warning will be triggered; for the initial state data of pumped storage power stations, if the data is abnormal, such as the stored energy exceeding the physical boundary, the state at the end of the previous scheduling cycle will be used as the initial value. When data quality is severely insufficient, the system can degrade to a conservative independent hydro-solar scheduling mode to ensure safety.

[0063] Step 103: Based on the photovoltaic power output forecast data and initial state data, the target scheduling mode for the current period is determined from multiple preset scheduling models using preset conversion efficiency evaluation criteria.

[0064] The conversion efficiency assessment criterion is a judgment logic used to weigh the costs of hydropower regulation against the benefits of photovoltaic power integration. Since hydropower regulation may lead to head loss or potential water curtailment risks, and pumped storage, while flexible, suffers from cycle efficiency losses, the system cannot blindly pursue power integration but needs to assess whether the current conversion efficiency is worthwhile.

[0065] The process of determining the target scheduling mode is essentially answering the question of whether pure hydropower regulation or the introduction of pumped storage is more optimal under the current operating conditions. If the evaluation results show that the marginal cost of hydropower regulation is low, the target scheduling mode is determined to be a hydro-solar complementary scheduling model; if the evaluation results show that the amount of solar power curtailment is huge or the cost of hydropower regulation is too high, the target scheduling mode is determined to be a hydro-solar-storage complementary scheduling model. This mechanism avoids the unnecessary losses caused by forcibly starting pumped storage when it is not needed, and also avoids the curtailment of solar power due to insufficient regulation when pumped storage is needed. The specific form of the conversion efficiency evaluation criteria will be detailed in subsequent embodiments, which can be either a static rule based on a fixed threshold or a dynamic adaptive algorithm based on real-time state awareness.

[0066] Step 104: Based on the photovoltaic power output forecast data and hydropower inflow data, solve the preset scheduling model corresponding to the target scheduling mode and generate a day-ahead scheduling instruction containing the power output plan of each power station.

[0067] In this embodiment, once the target scheduling model is determined, a specific mathematical optimization problem is selected. The system substitutes the collected photovoltaic power output forecast data and hydropower inflow data into the model and uses an optimization solver to find the optimal solution that satisfies all constraints. This optimal solution is the scheduling plan that includes the power output values ​​of each cascade hydropower station, photovoltaic power station, and pumped storage power station for each time period within the next 24 hours. The day-ahead scheduling instructions are then issued to the execution terminals of each power station to guide their actual production. For example, the instructions may require a hydropower station to reduce its output during the midday period when photovoltaic power generation is high, and increase its output during the off-peak period, or require a pumped storage power station to start its pumping units during periods of severe photovoltaic power curtailment.

[0068] Step 105, wherein the conversion efficiency evaluation criterion is used to evaluate the conversion relationship between hydropower regulation losses and photovoltaic consumption increments under different dispatch modes.

[0069] In this embodiment, this limitation clarifies the physical essence of the evaluation criterion. The conversion relationship refers to the reduction in hydropower generation required for the system to absorb an additional kilowatt-hour of photovoltaic power. If the cost is too high, for example, if absorbing 1 kilowatt-hour of photovoltaic power results in a 0.5 kilowatt-hour loss of hydropower due to reduced water head or wasted water, it may be economically unfeasible. The conversion efficiency evaluation criterion guides the system to intelligently switch between different adjustment methods by quantifying the input-output ratio. Through this mechanism, this method can improve the overall energy utilization rate of the multi-energy complementary system and maximize the increase in power generation.

[0070] Example 2 details the process of constructing multiple preset scheduling models, describing the specific form, constraints, and dual peak-shifting physical mechanism of each mathematical model, providing a computable mathematical basis for the entire scheduling method.

[0071] Step 201, the multiple preset scheduling models specifically include the independent scheduling model of water and solar power, the complementary scheduling model of water and solar power, and the complementary scheduling model of water, solar power, and energy storage.

[0072] In this embodiment, to quantitatively assess complementary benefits and provide multiple operational options, the system pre-configures three standardized mathematical models. The three models share the same hydraulic and power network topology, but their optimization objectives and the range of control variables differ.

[0073] Step 202, in which the independent scheduling model of hydropower and solar power takes the optimal power generation of the hydropower station as the objective function.

[0074] In this embodiment, the independent hydropower-solar power dispatch model simulates the traditional operation mode where energy entities lack coordination. Under this mode, hydropower stations dispatch based solely on their own inflow conditions and equipment characteristics to maximize their own power generation revenue, without considering the photovoltaic (PV) grid integration requirements. PV power stations generate power at maximum capacity, but are directly curtailed when grid access is limited. Specifically, the objective function of this model can be expressed as maximizing the total power generation of all cascade hydropower stations within the dispatch period T. Let N be the number of hydropower stations and T be the number of dispatch periods (usually 24). The objective function is a maximization formula, that is, for each time period t and each hydropower station i, the product of its power generation and the duration of the time period is accumulated. Mathematically, this is expressed as:

[0075] Max(E ind )=∑(P hy,i,t *Δt);

[0076] Among them, E ind P represents the total power generation of the cascade hydropower system under independent dispatch mode. hy,i,t Let be the actual power generation of the i-th hydropower station in time period t, and Δt be the duration of each scheduling period.

[0077] In this model, the amount of electricity generated by photovoltaic power is limited by the remaining capacity of the transmission channel, i.e.:

[0078] P pv,grid,t =min(P pv_predict,t ,P channel -∑(P hy,i,t ));

[0079] Among them, P pv,grid,t P represents the actual grid-connected (connected to the grid) power of the photovoltaic power station during time period t. pv_predict,t P represents the predicted maximum power generation capacity of the photovoltaic power station within time period t, which is determined by solar radiation conditions. channel For the capacity of the power transmission channel, ∑(P) hy,i,t ) represents the channel power occupied by all hydropower stations within time period t.

[0080] Step 203: The hydro-solar complementary scheduling model takes the maximum total power generation of the system during the scheduling period as the objective function, and includes water balance constraints and power transmission channel constraints.

[0081] In this embodiment, the hydro-solar hybrid scheduling model introduces a cooperative mechanism, and its objective is no longer to optimize a single power source, but to maximize the total power generation of the entire combined system. The objective function becomes:

[0082] Max(E sys )=∑((∑(P hy,i,t )+∑(P pv,grid,t ))*Δt);

[0083] Among them, E sys This represents the total power generation of the combined system.

[0084] In this model, the hydropower station actively adjusts its output curve to make room for photovoltaic power.

[0085] The water balance constraint describes the dynamic change in reservoir storage, that is, the storage volume at the end of a certain period equals the storage volume at the end of the previous period plus the inflow minus the outflow. The mathematical formula is:

[0086] V i,t =V i,t-1 +(I i,t -Q out,i,t )*Δt;

[0087] Among them, V i,t For the i-th reservoir during the time period The final water storage capacity, V i,t-1 Let I be the water storage of the i-th reservoir at the end of the previous time period (t-1). i,t Let Q be the inflow rate of the i-th reservoir during time period t. out,i,t Let be the outflow from the i-th reservoir during time period t.

[0088] For cascade hydropower stations, it is also necessary to consider the hydraulic connection constraints between upstream and downstream, that is, the inflow of the downstream reservoir includes the evolved component of the outflow of the upstream reservoir.

[0089] The total output of the power transmission channel must not exceed the transmission limit P of the ultra-high voltage direct current channel. channel_max .Right now:

[0090] ∑(P hy,i,t )+∑(P pv,grid,t )<=P channel_max .

[0091] This constraint is the main physical reason for photovoltaic curtailment and forcing hydropower to stagger peak hours.

[0092] Step 204: The hydro-solar-storage complementary scheduling model adds energy balance constraints, water level constraints, and output constraints for pumped storage power stations to the existing hydro-solar-storage complementary scheduling model.

[0093] In this embodiment, the hydro-solar-storage complementary scheduling model is the most complete optimization form. Its objective function further includes the power generation of the pumped storage power station (positive value) and the power consumption for pumping (negative value).

[0094] The energy balance constraint describes the energy conversion process in the upper reservoir of a pumped-storage power station, taking into account the cycle efficiency. That is:

[0095] E ps,t =E ps,t-1+(P pump,t *η pump -P gen,t / η gen )*Δt;

[0096] Among them, E ps,t Let E be the current stored energy of the pumped storage power station at time t, and η represent the efficiency. ps,t-1 P represents the energy stored by the pumped storage power station at the end of the previous time period (t-1). pump,t η represents the pumping power (input power) of the pumped storage power station during time period t. pump For pumping efficiency (pump operating efficiency), P gen,t η represents the power generation (output power) of the pumped storage power station during time period t. gen For power generation efficiency.

[0097] Water level constraints and power output constraints apply not only to pumped storage but also to conventional hydropower.

[0098] Furthermore, the solution to the preset scheduling model corresponding to the target scheduling mode, including the execution of dual peak-shifting compensation, is also subject to preset safe operation constraints, which include at least: water level fluctuation constraints and downstream flow constraints.

[0099] Water level fluctuation constraint: Limit the water level fluctuation of hydropower stations and pumped storage power stations within the dispatch cycle to not exceed the preset upper limit of fluctuation;

[0100] Discharge flow constraint: Limiting the discharge flow of hydropower stations to between the upper and lower limits of the allowable value in order to ensure the safety of downstream ecology and navigation.

[0101] Specifically, water level fluctuation constraints limit the rise and fall of reservoir water levels per unit time to prevent reservoir bank landslides or ecological damage. The formula is expressed as:

[0102] |(Z t -Z t-1 )|<=ΔZ _max ;

[0103] Among them, Z t Z represents the water level of the reservoir at the end of time period t (usually referring to the water level in front of the dam or the water level in the upper reservoir). t-1 Let ΔZ be the water level of the reservoir at the end of the previous time period (t-1). _max This is the maximum allowable daily (or time-period) variation in water level.

[0104] The discharge flow constraint limits the discharge flow of hydropower stations to between the lower limit of ecological flow and the upper limit of flood control flow, that is:

[0105] Q eco <=Q out_t <=Qflood ;

[0106] Among them, Q out_t Q represents the total discharge of the hydropower station during time period t. eco As the lower limit of ecological flow, Q flood This is the upper limit for flood control flow.

[0107] The above constraints ensure the engineering feasibility and security of the scheduling scheme.

[0108] Step 205: Solve the preset scheduling model corresponding to the target scheduling mode, including performing dual peak-shifting compensation: based on the photovoltaic predicted output data, control the cascade hydropower stations to change the output time schedule allocation so that the peak of hydropower generation avoids the peak of photovoltaic output; based on the comparison results of the photovoltaic predicted output data and the capacity of the power transmission channel, when the period of photovoltaic curtailment is identified, control the pumped storage power station to operate in pumping mode to absorb the photovoltaic curtailment.

[0109] In this embodiment, this step describes the physical regulation mechanism within the system. Dual peak shifting refers to the superposition of two layers of regulation. The first layer is peak shifting for conventional hydropower. Conventional hydropower stations reduce their output during midday, shifting the electricity that would normally be generated at midday to the morning and evening peak hours, thus freeing up space for photovoltaic power generation during midday. The second layer is valley filling and peak shifting for pumped storage. When the regulation capacity of conventional hydropower is exhausted (e.g., limited by natural inflow and minimum discharge flow constraints) while photovoltaic power still has surplus electricity, the pumped storage power station starts its pumps, using the excess photovoltaic power to pump water from the lower reservoir to the upper reservoir, effectively creating a load and absorbing the surplus photovoltaic power. The synergistic effect of these two mechanisms maximizes the utilization rate of renewable energy.

[0110] Step 206: The hydropower regulation loss involved in the conversion efficiency assessment criteria is quantified as the hydropower-compensated photovoltaic power loss rate. This rate is defined as the reduction in hydropower generation caused by the absorption of a unit of photovoltaic curtailment, calculated as the ratio of the reduction in hydropower generation under the complementary dispatch mode to the increase in photovoltaic grid-connected power generation under the independent dispatch mode. Specifically, the reduction in hydropower generation is obtained by comparing the hydropower generation under the complementary dispatch mode and the independent dispatch mode, while the increase in photovoltaic grid-connected power generation is obtained by comparing the photovoltaic grid-connected power generation under the complementary dispatch mode and the independent dispatch mode.

[0111] In this embodiment, this indicator is defined to quantify the cost of hydropower regulation. Let ΔE hy ΔE represents the reduction (positive value) in hydropower generation under the complementary mode compared to the independent mode. pv Let k be the increase in photovoltaic power generation connected to the grid. Then, the hydropower compensation rate for photovoltaic power loss is k. hy for:

[0112] k hy=ΔE hy / ΔE pv .

[0113] This metric intuitively reflects exchange efficiency. For example, k hy =0.1 means that for every 0.1 kWh of hydropower sacrificed, 1 kWh of photovoltaic power can be gained in return, which is obviously very cost-effective; while if k hy A value of 0.8 indicates that the exchange cost is very high. This quantitative indicator forms the basis of the decision-making algorithm in subsequent embodiments.

[0114] Step 207: The photovoltaic consumption increment assessed by the preset conversion efficiency evaluation criteria is determined by calculating the system complementary power generation. The system complementary power generation is defined as the increment of the total system power generation obtained by solving the preset scheduling model corresponding to the target scheduling mode, relative to the total system power generation obtained by solving the independent water-solar scheduling model.

[0115] In this embodiment, the system complementarily increases the power generation ΔE. sys It is the ultimate indicator for evaluating dispatch effectiveness, directly reflecting the additional electricity generated by the entire energy base after adopting complementary dispatch, and serves as the basis for assessing economic and social benefits. The formula is:

[0116] ΔE sys =E complementary -E independent ;

[0117] Among them, E complementary E represents the total system power generation under complementary dispatch mode. independent This represents the total power generation of the system under the independent scheduling mode of hydropower and solar power.

[0118] This increase actually comes from the reduction in photovoltaic curtailment minus the conversion losses of hydropower and pumped storage.

[0119] Example 3 elaborates on the basic decision-making scheme and provides a static decision-making logic that is easy to calculate and implement in engineering. As the basic implementation method of the present invention, it is suitable for simple scheduling scenarios with high requirements for computational efficiency or lack of real-time status monitoring data.

[0120] The target scheduling mode for the current period is determined from multiple preset scheduling models, including the application of static decision logic based on preset fixed thresholds: determining whether the power loss rate of hydropower compensating for photovoltaic power is lower than the preset fixed threshold; if so, the target scheduling mode is determined to be the hydro-solar complementary scheduling model; if not, or if it is determined that there is photovoltaic power curtailment in the current period, the target scheduling mode is determined to be the hydro-solar-storage complementary scheduling model.

[0121] In this embodiment, this step involves selecting and applying a preset fixed threshold. The preset fixed threshold physically corresponds to the overall cycle efficiency loss rate of the pumped-storage power station. Typically, the pump-to-power cycle efficiency of a pumped-storage power station is approximately 0.75, or 75%, meaning that for every 1 kWh of electricity pumped and stored, only 0.75 kWh is generated back, resulting in an energy loss of 0.25 kWh. Therefore, in a specific implementation of this embodiment, the preset fixed threshold is set to 0.25.

[0122] In practical implementation, the system first calculates the power loss rate k of hydropower-compensated photovoltaic power under the current operating conditions according to the formula. hy Furthermore, the following judgment logic is executed:

[0123] If k hy The value is less than 0.25, meaning that the power generation loss incurred by hydropower to absorb a unit of curtailed photovoltaic power is less than 0.25 kWh. In this case, using hydropower for regulation is superior to pumped storage in terms of energy conversion efficiency (the latter has a fixed loss of 0.25 kWh). Therefore, the system determines that the hydro-solar complementary mode is optimal, utilizing only the flexibility of hydropower for peak-shaving regulation without activating pumped storage.

[0124] Conversely, if k hy A value >=0.25 indicates that the regulation cost of hydropower is too high, such as requiring significant deviation from the optimal water level or resulting in water wastage, with energy losses exceeding the fixed losses of pumped storage. In this case, or when the system detects that even with full hydropower regulation capacity, there is still unabsorbed photovoltaic curtailment (i.e., curtailment greater than 0), the system determines that introducing pumped storage is necessary and economical. The target scheduling mode is determined to be a hydro-solar-storage complementary scheduling model, controlling the pumped storage units to start pumping to fill the photovoltaic absorption gap. Although the above static strategy is simple, it can effectively define the division of labor between hydropower and pumped storage in typical scenarios such as the flood season.

[0125] The static decision-making logic based on a fixed threshold described above is only one specific implementation of the conversion efficiency evaluation criterion. In other optional implementations, the evaluation criterion may also employ self-learning rules based on historical data, inference rules based on fuzzy logic, or adaptive rules based on multi-factor dynamic correction as described in subsequent embodiments. This invention does not limit the specific form of the evaluation criterion; as long as the criterion can characterize the conversion relationship between hydropower regulation losses and photovoltaic absorption increments, it falls within the protection scope of this invention.

[0126] Example 4 describes in detail the three-stage scheduling process of pre-judgment-layered optimization-coordinated adjustment. In view of the problem that traditional methods directly solve large-scale mixed integer nonlinear programming models, which are time-consuming and difficult to converge, this example proposes a funnel-shaped screening process that can improve computational efficiency. Through progressive decision-making, the optimal mode is quickly locked.

[0127] Step 401: Based on the photovoltaic power output forecast data and initial state data, determine the target scheduling mode for the current time period from multiple preset scheduling models, such as... Figure 2 As shown, the specific process includes a three-stage progressive scheduling flow: pre-judgment, hierarchical optimization, and collaborative adjustment.

[0128] Phase 1: Based on the photovoltaic power output forecast data and the hydropower operation status under the independent hydropower dispatch mode obtained by solving the independent hydropower dispatch model, calculate the estimated hydropower marginal loss rate;

[0129] Phase 2: Calculate the dynamic loss rate threshold for the current period based on the initial state data, and determine the target scheduling mode based on the comparison between the estimated hydropower marginal loss rate and the dynamic loss rate threshold.

[0130] Phase 3: Solve the preset scheduling model corresponding to the determined target scheduling mode to obtain the output allocation results of each power station, namely the hydropower output value sequence, photovoltaic grid-connected power sequence and pumping / power generation sequence of pumped storage power stations for each time period within the scheduling cycle, and encapsulate the above sequences into day-ahead scheduling instructions.

[0131] In this embodiment, the design logic of the three-stage progressive scheduling process is as follows: first estimate, then decide, and finally refine. In the first stage, the system does not directly run a complex complementary optimization model. Instead, based on the existing independent scheduling results of hydropower and solar power (which is a baseline data with minimal computational cost), it quickly calculates the potential cost of hydropower regulation, i.e., the estimated marginal loss rate of hydropower, using the algebraic formulas detailed in subsequent embodiments. This process typically takes milliseconds.

[0132] In Phase Two, the system introduces a dynamic decision-making mechanism. Unlike a fixed threshold, the dynamic loss rate threshold changes in real time based on the system's state of charge and grid pressure; its specific calculation method will be described in subsequent embodiments.

[0133] Step 402, Phase Two, determines the target scheduling mode, specifically including executing the following logic:

[0134] Determine whether the estimated marginal loss rate of hydropower is less than the dynamic loss rate threshold k. th And simultaneously determine the total predicted amount of photovoltaic curtailment E cur Is it less than the available pumped storage capacity E of the pumped storage power station? ps_avail If both judgment conditions are met, the target scheduling mode is determined to be the hydro-solar complementary scheduling model; if either judgment condition is not met, the target scheduling mode is determined to be the hydro-solar-storage complementary scheduling model.

[0135] This judgment logic reflects the dual constraints of economic efficiency and physical feasibility. The first condition (k) hy<k th To determine whether it is economically worthwhile, the second condition (E) cur <E ps_avail ) Determine if it is physically feasible. Only when hydropower regulation is both cheap and can physically absorb the curtailment of photovoltaic power should the hydro-solar complementary mode be chosen; otherwise, if either condition is not met, such as hydropower being too expensive, or too much curtailment of solar power exceeding the regulation capacity of the hydropower station, pumped storage must be activated.

[0136] Step 403, following stage three, also includes a collaborative adjustment step: Based on the solution results of the preset scheduling model corresponding to the target scheduling mode, calculate the actual hydropower marginal loss rate; determine whether the deviation between the actual hydropower marginal loss rate and the estimated hydropower marginal loss rate exceeds a preset allowable deviation threshold; if so, update the estimated hydropower marginal loss rate using the actual hydropower marginal loss rate, and return to stage two to redetermine the target scheduling mode, such as... Figure 4 As shown.

[0137] In one specific implementation of this embodiment, the tolerance threshold is set to 0.05, meaning that mode re-judgment is triggered when the difference between the estimated loss rate and the actual loss rate exceeds 5 percentage points. The selection of this threshold is based on the following: a threshold that is too small will lead to frequent iterations and increase the computational burden; a threshold that is too large may cause mode selection bias, affecting the final benefit. Engineering practice shows that values ​​within the range of 0.03 to 0.08 can achieve a good balance.

[0138] In this embodiment, the collaborative adjustment step constructs a closed-loop feedback error correction mechanism. Since the loss rate in stage one is estimated, it may deviate from the actual loss rate obtained after fine-tuning in stage three. For example, the estimated k... hy The value was 0.2, which led the system to choose the water-solar complementary mode; however, after optimization, it was found that the actual k value was 0.2. hy A deviation as high as 0.4 might trigger complex ramping constraints, leading to a significant drop in efficiency. In this case, if the allowable deviation threshold is set to 0.05, the deviation between 0.2 and 0.4 exceeds the limit. The system will then feed back 0.4 to Phase Two. In the new assessment, 0.4 might exceed the dynamic threshold (e.g., 0.3), triggering a mode switch to select the hydro-solar-storage complementary mode and resolving the problem. This mechanism ensures that simplified predictions do not lead to errors in the final decision, enhancing the system's robustness.

[0139] Example 5 provides a rapid estimation method for the marginal loss rate of hydropower, focusing on solving the technical problem in Stage 1 of how to calculate the loss rate without solving equations. This example details the specific implementation process of combining offline data mining with online rapid calculation.

[0140] Step 501, in Stage 1, calculate the estimated marginal loss rate of hydropower, such as... Figure 3As shown, the specific steps include: determining the predicted amount of photovoltaic curtailment based on photovoltaic power output data and the capacity of power transmission channels; calculating the output adjustment loss coefficient based on the hydropower operating status and the pre-stored optimal hydropower operating efficiency curve, whereby the output adjustment loss coefficient characterizes the degree to which the actual hydropower operating efficiency deviates from the optimal operating efficiency; and calculating the estimated hydropower marginal loss rate based on the predicted amount of photovoltaic curtailment, the output adjustment loss coefficient, and the output data under the independent hydropower dispatch mode, without needing to run the hydropower-photovoltaic complementary dispatch model.

[0141] In this embodiment, the incremental rate principle is used for approximate calculation, and the predicted amount of photovoltaic curtailment E is calculated. cur,t It is a definite input quantity that can be calculated using a formula:

[0142] E cur,t =max(0,P pv,t +P hy,ind,t -P channel );

[0143] In the formula, P hy,ind,t P represents the power output (in ten thousand kW) during time period t when hydropower is independently dispatched. pv,t For the ideal photovoltaic output (ten thousand kW) during period t, P channel The capacity of the power transmission channel (in ten thousand kW).

[0144] The predicted amount of photovoltaic curtailment represents the amount of photovoltaic power exceeding the channel capacity, assuming that hydropower maintains its independent operation and output remains unchanged. This portion of power is the target amount that hydropower needs to give way to.

[0145] Step 502, the output adjustment loss coefficient is calculated by comparing the current actual operating efficiency of hydropower with the pre-constructed optimal operating efficiency of hydropower; the optimal operating efficiency of hydropower is pre-constructed through the following steps: obtaining historical operating data of hydropower stations, constructing optimal efficiency curves under different head and flow conditions through data fitting, and pre-storing them in the scheduling system.

[0146] In this embodiment, this step is an offline preparation process. The system extracts the power output P corresponding to different power generation flows Q at various water heads H by mining SCADA (Supervisory Control and Data Acquisition) data from the hydropower station over the past few years, and fits an efficiency function η=f(H,Q). For each water head H, the optimal flow flow Q that maximizes efficiency η is found. opt And optimal efficiency η opt The above data is stored in a database as a multidimensional table or fitted formula. During online calculation, the output adjustment loss coefficient α... loss,t The calculation formula is:

[0147] α loss,t =1-(η curr,t / η opt,t).

[0148] Where, η curr,t η is the efficiency corresponding to the current planned output of the hydropower station. opt,t To the current head H t The highest possible efficiency that a water turbine can achieve.

[0149] This coefficient intuitively reflects the proportion of efficiency loss that may occur if a hydropower station deviates from its current operating conditions to cooperate with photovoltaic regulation.

[0150] Step 503: Based on the predicted amount of photovoltaic curtailment, the output adjustment loss coefficient, and the output data under the independent hydropower dispatch mode, calculate the estimated hydropower marginal loss rate. Specifically, calculate the difference between the output under the independent hydropower dispatch mode and the preset minimum allowable hydropower output, and integrate or accumulate the difference over time in combination with the output adjustment loss coefficient to obtain the estimated reduction in hydropower generation; divide the estimated reduction in hydropower generation by the sum of the predicted photovoltaic curtailment amounts to obtain the estimated hydropower marginal loss rate.

[0151] In this embodiment, a specific estimation formula is provided, which estimates the reduction in hydropower generation ΔE. hy_est Instead of solving through an optimization model, it is calculated using the following approximate formula:

[0152] ΔE hy_est =∑ t [(P hy,ind,t -P hy,min,t )*α loss,t *Δt];

[0153] Among them, P hy,ind,t For the planned power output of the hydropower station in independent dispatch mode during time period t, P hy,min,t The minimum allowable output (in ten thousand kW) for hydropower during time period t; α loss,t This is the output adjustment loss coefficient for hydropower during time period t.

[0154] The physical meaning of the formula is: To make room for photovoltaic power generation, the maximum adjustment range of a hydropower station is the current independent output P. hy,ind With minimum safe output P hy_min The difference. Multiply the adjustment range by the output adjustment loss coefficient α. loss This approximates the energy loss during the regulation process. The estimated marginal loss rate of hydropower, k... hy_est The calculation is as follows:

[0155] k hy_est =ΔE hy_est / ∑ t (E cur,t );

[0156] Among them, Ecur,t This represents the predicted amount of solar power curtailment.

[0157] For example, in a certain time period t, the amount of curtailed photovoltaic power is 100,000 kWh. The independent output of hydropower is 500,000 kW, with a minimum output of 100,000 kW and a downward adjustment range of 400,000 kW. Looking up the table, the output adjustment loss coefficient under the current hydropower head is 0.05, which is a 5% efficiency loss. Therefore, the estimated hydropower loss for this time period is 400,000 kW * 0.05 * 1h = 20,000 kWh. If curtailment only occurs during this time period throughout the day, the estimated loss rate k is... hy =2 / 10 = 0.2. The system can then proceed to stage two for decision-making based on this value of 0.2. This method avoids cumbersome nonlinear programming iterations and shortens the decision-making time.

[0158] For example, a complete example of prediction calculation is given below.

[0159] Assume the following parameters for a certain scheduling day: the scheduling cycle is 24 time periods, each lasting 1 hour; the power transmission capacity is 8 million kW; the output curves of cascade hydropower under independent scheduling mode are: 4.8 million kW from 0-6, 5 million kW from 7-18, and 4.8 million kW from 19-24; the predicted output curves of photovoltaic power are: 0-6, 0; 7-9, 2 million kW; 10-14, 6 million kW; 15-17, 3 million kW; and 18-24, 0; the minimum allowable output of hydropower is 1 million kW.

[0160] Step 1: Determine the predicted amount of photovoltaic curtailment. During the period from 10:00 to 14:00, the total system output is 500 + 600 = 11 million kW, which exceeds the channel capacity by 8 million kW. The amount of photovoltaic curtailment is (1100 - 800) × 5h = 15 million kWh.

[0161] Step two, calculate the output adjustment loss coefficient. Query the pre-stored optimal efficiency curve; the optimal efficiency of hydropower under the current head is 0.92, and the actual operating efficiency is 0.88. Therefore, α... loss =1-0.88 / 0.92=0.043.

[0162] Step 3: Calculate the estimated reduction in hydropower generation. The adjustable range of hydropower is 500-100=400 million kW, and the estimated loss is 400×0.043×5h=860,000 kWh.

[0163] Step four: Calculate the estimated marginal loss rate of hydropower. hy_est =86 / 1500=0.057. This value is much smaller than the pumped storage baseline loss rate of 0.25, indicating that the hydro-solar hybrid mode can meet the requirements.

[0164] Example 6 describes the adaptive calculation and decision-making of the dynamic loss rate threshold, elaborates on the specific calculation logic of the dynamic loss rate threshold in stage two, and demonstrates how to use multi-dimensional state perception technology to enable the dispatching system to adjust the tolerance for hydropower regulation losses in real time according to the charge state of pumped storage power stations and the peak-shaving demand of the power grid, so as to achieve better adaptive control than a fixed threshold.

[0165] Step 601: The dynamic loss rate threshold is obtained by applying a dual-factor correction to the pumped storage benchmark loss rate. The calculation formula is the product of the pumped storage benchmark loss rate, the capacity availability coefficient, and the peak shaving urgency coefficient. The pumped storage benchmark loss rate is determined based on the comprehensive cycle efficiency of the pumped storage power station.

[0166] In this embodiment, the core calculation formula is defined as follows:

[0167] k th,t =k ps *Φ cap,t *Φ dem,t ;

[0168] Where, k th,t Φ is the dynamic loss rate threshold for time period t. cap,t Φ is the capacity availability factor. dem,t k is the peak-shaving urgency factor. ps The pumping base loss rate has the same physical meaning as the static threshold in Example 3, and depends on the inherent energy conversion efficiency of the device.

[0169] Pumped storage benchmark loss rate k ps :

[0170] k ps =1-η ps ;

[0171] In the formula: η ps This refers to the overall efficiency of pumped storage.

[0172] For example, when the overall efficiency of pumped storage is 75%, the benchmark loss rate k ps The value is set to 0.25. This benchmark value represents the decision-making threshold under ideal conditions (sufficient storage capacity and full photovoltaic power generation). Two time-varying correction coefficients Φ are introduced. cap,t and Φ dem,t The system maps physical constraints and grid demand to economic indicators, achieving dynamic floating.

[0173] It should be noted that the pumped storage baseline loss rate k ps The value of k is related to the equipment performance of the pumped storage power station. For pumped storage power stations with a comprehensive efficiency in the range of 70%-80%, k ps The value range of k is 0.20-0.30. psWhen k is too large, the system tends to use hydropower regulation more often, and the frequency of pumped storage start-up decreases; when k ps When the value is too small, the system tends to start the pumped storage earlier. In engineering practice, adjustments can be made based on the actual equipment efficiency and operational strategy preferences.

[0174] Step 602: The capacity availability coefficient represents the available regulating capacity status of the pumped storage power station at the current moment, and its value is negatively correlated with the current storage capacity. The specific calculation method is: the difference between the maximum storage capacity of the pumped storage power station and the current storage capacity, divided by the difference between the maximum storage capacity and the minimum storage capacity of the pumped storage power station.

[0175] In this embodiment, the capacity availability factor Φ cap,t This reflects the remaining water pumping capacity of a pumped-storage hydroelectric power station. The calculation formula is:

[0176] Φ cap,t =(E ps_max -E ps,t ) / (E ps_max -E ps_min );

[0177] Among them, E ps_max E ps_min These represent the maximum and minimum storage capacities (in ten thousand kWh) of pumped storage power stations, respectively; E ps,t The current stored energy (in ten thousand kWh) of the pumped storage power station at time t.

[0178] Its physical logic is as follows: when the upper reservoir is close to full storage (E ps,t Approaching E ps_max When the molecule approaches 0, Φ becomes... cap,t It approaches 0. At this point, according to the formula in step 601, the dynamic threshold k... th,t It will also become smaller. A smaller threshold indicates that the system has a lower tolerance for hydropower loss, forcing the judgment condition k to become less sensitive. hy <k th It's difficult for this to hold true (unless hydropower loss is almost zero), so the system tends not to activate pumped storage (because there's nowhere left to pump water), or it forces the system to prioritize hydropower regulation. Conversely, when the reservoir is empty, this coefficient approaches 1, and the system reverts to normal decision-making logic. This mechanism transforms the hard constraint of reservoir capacity into a soft threshold signal, avoiding the risk of overflow from pumped storage power stations.

[0179] Step 603: The peak shaving urgency coefficient represents the pressure state of the power grid accepting photovoltaic power output during the current period. The calculation method is as follows: divide the portion of the photovoltaic predicted output during the current period that exceeds the capacity of the power transmission channel by the maximum pumping power of the pumped storage power station, and take the minimum value of the obtained ratio and the value 1 as the peak shaving urgency coefficient.

[0180] In this embodiment, the peak shaving urgency coefficient Φdem,t Used to reflect the degree of photovoltaic overload. The calculation formula is:

[0181] Φ dem,t =Min(1,(P pv,t -P channel ) / P ps_max );

[0182] In the formula, P pv,t Ideal photovoltaic output (10,000 kW) for period t; P channel Power transmission capacity (ten thousand kW); P ps_max This is the maximum pumping power of the pumped storage system (in ten thousand kW).

[0183] Its physical logic is: when the photovoltaic output P pv,t Much greater than the channel capacity P channel When the amount of abandoned electricity increases, it indicates that there is a large amount of power that must be consumed. As the amount of abandoned electricity increases, this coefficient gradually increases and tends to 1. When this coefficient is large, the dynamic threshold k... th,t Maintaining a high level makes k hy <k th This is more likely to be true, allowing hydropower to operate at a higher loss rate, in which case absorbing solar power becomes the top priority. Conversely, if solar power does not exceed the channel capacity, the coefficient is 0 (the formula implicitly includes Max(0,...) handling, or the logic has excluded scenarios without curtailment), the threshold is 0, and the system will naturally shut down the pumped storage function. The above design ensures that when the grid's peak-shaving pressure is high, the system can automatically relax economic constraints, prioritizing safety and absorption.

[0184] Example 7: Verify the effectiveness of the method proposed in this invention in practical engineering applications, especially verify the performance of the dual peak-shifting compensation mechanism in different hydrological seasons.

[0185] In this embodiment, a hydro-solar-storage hybrid energy base in a certain river basin was selected for verification. The system parameters were set as follows: cascade hydropower installed capacity of 4.82 million kW, photovoltaic installed capacity of 17.78 million kW, pumped storage installed capacity of 3 million kW, and external transmission capacity of 8 million kW.

[0186] Scenario 1: A typical day during the flood season. At this time, water inflow is abundant, but hydropower regulation capacity is limited, requiring the maintenance of high water levels and full power generation. After adopting the adaptive scheduling method of this invention, the system exhibits distinct three-stage characteristics.

[0187] During periods of low photovoltaic output, the hydropower loss rate k hy The system is relatively low, automatically selecting a hydro-solar complementary mode, which can absorb solar power simply by fine-tuning the hydropower output; during the peak solar power generation period at midday, the amount of curtailed solar power increases dramatically. hyIf the water level spikes and exceeds the dynamic threshold, the system automatically switches to a hydro-solar-storage complementary mode, with the pumped-storage unit pumping water at full power; once the pumped-storage unit is full, the capacity availability factor Φ cap As the threshold decreases, the system again relies on deep peak shaving of hydropower (even with significant losses) to cover the losses. Results show that compared to traditional methods, the photovoltaic absorption rate is significantly improved, and the risk of hydropower wastage due to forced peak shaving is reduced.

[0188] Scenario 2: Typical day during the dry season. At this time, water inflow is low, and although the theoretically large hydropower regulation capacity is limited by the minimum discharge flow and low water level, the actual regulation capacity is limited. In this scenario, the proportion of curtailed photovoltaic power is high, and the peak-shaving urgency coefficient Φ is high. dem The dynamic threshold remained consistently high, hovering around 1.0. Consequently, the system operated with pumped storage almost continuously for compensation until the pumped storage station was fully charged or emptied. Verification results demonstrate that this method can automatically identify operating conditions during the dry season, maximizing the utilization of the pumped storage's circulation capacity and ensuring the stability of power supply during this period.

[0189] According to one aspect of this application, the dual peak-shifting compensation mechanism of hydro-solar-storage can also be as follows:

[0190] Under limited channel or load constraints, during the midday hours when photovoltaic (PV) output is concentrated, if hydropower is independently dispatched to seize the consumption space, PV will generate a large amount of curtailed electricity. If the peak output periods of hydropower and PV are staggered, adjusting their own output time allocation methods can effectively reduce curtailment. The process of achieving staggered peak and valley values ​​of hydropower output and PV output through hydropower regulation to improve the utilization rate of renewable energy is defined as hydro-PV peak shifting compensation.

[0191] With the addition of pumped storage, the curtailment of photovoltaic power is further reduced. Pumped storage power stations, through their flexible charging and discharging capabilities, pump water for energy storage during peak photovoltaic power generation periods at midday and discharge it at other times. This collaborative mode is known as dual peak-shaving compensation for hydropower, photovoltaic power, and pumped storage. It is defined as: a strategy that comprehensively considers the complementary characteristics of hydropower, photovoltaic power, and pumped storage, utilizing the peak-shaving capacity of hydropower and the peak-shifting capacity of pumped storage to jointly change the power output sequence, so that peak power generation actively avoids peak photovoltaic power generation, thereby systematically improving the utilization rate of renewable energy. The increased power generation from hydropower and photovoltaic power depends on the peak-shaving capacity of hydropower and the absorption capacity of pumped storage during periods of photovoltaic curtailment.

[0192] According to one aspect of this application, quantitatively assessing the increased power generation from complementary dispatch can also include:

[0193] When hydropower output is abundant and there is no additional storage capacity, or when photovoltaic (PV) output is abundant, hydropower and PV compete for grid connection due to complex constraints such as transmission channels and reservoir scheduling. When PV power cannot be transmitted through the transmission channels, curtailment occurs. Hydropower can reduce PV curtailment by adjusting its output timing to make room for transmission channels, but this may result in its own power generation loss. Pumped storage, on the other hand, uses the curtailed PV power to pump water, reducing PV curtailment while ensuring hydropower output. Multi-energy complementary short-term scheduling can effectively increase the total system power generation. The increased power generation from system complementarity is defined as the increment in total power generation relative to the independent hydropower-PV scheduling mode.

[0194] ;

[0195] ;

[0196] In the formula, The total number of scheduling periods is considered in this paper, which mainly focuses on day-ahead scheduling, i.e., 24 hours. , These represent the increased power generation (in ten thousand kWh) of the hydro-solar system and the hydro-solar-storage system during the dispatch period. , The figures represent the increased power generation (in ten thousand kWh) during time period t for hydro-solar hybrid (superscript (hpb)) and hydro-solar-storage hybrid dispatch (superscript (psb)). , , The figures represent the total system power generation (in ten thousand kWh) under the following modes: hydro-solar-storage complementary mode (superscript (cps)), hydro-solar complementary mode (superscript (coo)), and hydro-solar independent mode (superscript (ind)) during time period t.

[0197] According to one aspect of this application, the water-photovoltaic exchange efficiency index is specifically defined and calculated as follows:

[0198] In the hydro-solar complementary scheduling model, hydropower serves as the primary power source, and complementarity is achieved by allowing hydropower to free up space for solar power, thus promoting solar power consumption. The fundamental premise of this hydro-solar complementarity is that hydropower has no wasted water under independent scheduling scenarios and possesses the potential to compensate for solar power waste. During complementary scheduling, to maximize the total system power generation, hydropower will deviate from its optimal scheduling mode when operating independently. It will operate with minimal or no power loss, staggering its output schedule with solar power, using peak-shaving compensation to maximize the absorption of wasted solar power. The increase comes from the increased solar power consumption, while the potential loss comes from the loss of hydropower generation. The total power generation is increased through complementarity. Therefore, the total increase in power generation generated by short-term hydro-solar complementary scheduling is essentially the change in hydropower generation. Changes in actual grid-connected photovoltaic power generation The sum of these is defined as the hydro-solar power conversion efficiency. :

[0199] ;

[0200] In the formula, The larger the value, the more photovoltaic energy can be replaced by a unit of hydropower. The maximum replacement rate means that hydropower can replace photovoltaic power output with almost no loss. The minimum replacement rate means that a unit output of hydropower can only replace a unit output of photovoltaic power. This replacement efficiency is the reciprocal of... This refers to the amount of electricity lost by hydropower in order to replace a unit of photovoltaic power, i.e., the hydropower loss coefficient.

[0201] According to one aspect of this application, the evaluation index for peak-shifting compensation effect of hydro-solar-storage is specifically defined and calculated as follows:

[0202] The hydro-solar hybrid effect takes into account the variations in hydropower output. Three factors are defined as compensation indicators for peak-shaving: the maximum hydropower output peak-shaving depth, the maximum peak-shaving duration, and the ideal photovoltaic power generation. The hydro-solar-storage hybrid effect requires additional consideration of the pumped storage effect, specifically the maximum pumping amplitude and pumping duration under pumped mode. These factors are described below:

[0203] Maximum output peak shift depth (10,000 kW) refers to the maximum reduction in hydropower output during the peak period of photovoltaic power generation when dispatched relatively independently.

[0204] ;

[0205] ;

[0206] Maximum off-peak duration (h) refers to the duration of peak-shifting compensation scheduling for hydropower.

[0207] ;

[0208] ;

[0209] in, This is a counting function that measures the number of time periods that meet the conditions.

[0210] Ideal power generation of photovoltaic (10,000 kWh) refers to the total ideal output of photovoltaic power within a day, that is, the sum of photovoltaic curtailment and actual grid-connected power.

[0211] ;

[0212] Maximum pumping power consumption (10,000 kW) refers to the maximum daily pumping power consumption of a pumped storage power station in pumping mode during the hydro-solar-storage complementary dispatch process.

[0213] ;

[0214] Pumping duration (h) refers to the duration of pumping mode operation of the pumped storage power station.

[0215] .

[0216] According to one aspect of this application, the dual peak-shifting compensation assessment criteria may also be as follows:

[0217] In the hydropower-solar peak-shifting compensation mechanism, compared to the independent dispatch mode, hydropower will change its output process by increasing hydropower curtailment or reducing its own power generation, thereby improving the photovoltaic absorption capacity and increasing the total power generation of the system. Therefore, less hydropower loss can replace more photovoltaic increment. The amount of electricity lost by hydropower to replace a unit of photovoltaic power is called the hydropower loss coefficient. The coefficient increases dynamically with the increase of ideal photovoltaic power generation.

[0218] If only pumped storage is used to compensate for and regulate photovoltaic energy, pumped storage, as a common energy storage method, involves storing surplus photovoltaic power during periods of photovoltaic redundancy and generating electricity during periods of no or low sunlight. Although this can fully utilize surplus photovoltaic power and alter the power output time distribution process, there is energy loss during the transition between pumping and generating states, i.e., the pumped storage power station loss rate μ:

[0219] μ = 1 - η;

[0220] In the formula: η is the comprehensive cycle efficiency of the pumped storage power station, which refers to the ratio of the power generation of the power station to the pumping power within a certain period of time. It is used to measure the loss in the energy conversion process and is generally between 0.65 and 0.75.

[0221] Therefore, different compensation methods are accompanied by different energy losses, and there exists a compensation order:

[0222] (1) Hydropower loss rate The loss rate μ is much lower than that caused by using pumped storage for compensation: without using pumped storage for compensation and regulation, the impact of hydropower regulation on its own power generation is low, and it can fully compensate for and absorb the curtailed photovoltaic power.

[0223] (2) Hydropower loss rate If the loss rate μ is higher than that resulting from pumped storage compensation, or if there is photovoltaic curtailment, pumped storage compensation should be activated to jointly compensate photovoltaic output with hydropower during peak periods. With pumped storage added to the joint compensation, the hydropower loss rate will be reduced.

[0224] (3) Hydropower and pumped storage work together to compensate and regulate, and reach the upper limit of pumped storage capacity: At this time, hydropower will deepen its own compensation range, increase the peak shifting depth and peak shifting duration, and the hydropower loss rate will further increase to the upper limit. At this time, the photovoltaic curtailment of the system is difficult to be fully absorbed.

[0225] According to one aspect of this application, multiple preset scheduling models are constructed, including a water-solar independent scheduling model, a water-solar complementary scheduling model, and a water-solar-storage complementary scheduling model, as detailed below:

[0226] 1) Construct an independent scheduling model for water and solar power.

[0227] When hydropower and solar power are independently optimized and dispatched, hydropower is optimized and dispatched with its own power generation as the target, while the output of photovoltaic power is independently arranged and transmitted under the limited load and channel conditions.

[0228] ;

[0229] ;

[0230] ;

[0231] In the formula: This represents the total power generation of the system during the dispatch period under independent dispatch mode (in ten thousand kWh). , These represent the average power output of the i-th hydropower station in the hydro-solar independent mode and the actual grid-connected power output (in ten thousand kW) of the k-th photovoltaic power station in the cascade hydropower station during time period t. and The power generation flow (m³) of the i-th hydropower station in time period t are respectively. 3 / s) and head (m); denoted as the hydropower output coefficient; N represents the number of hydropower stations; and J represents the number of photovoltaic power stations.

[0232] 2) Construct a water-solar complementary scheduling model.

[0233] When hydropower and solar power are dispatched in a complementary manner, the overall power generation of the system during the dispatch period is maximized as the objective function, with the unified dispatch of hydropower and solar power energy.

[0234] ;

[0235] .

[0236] 3) Construct a water-solar-storage complementary scheduling model.

[0237] When hydropower, solar power, and pumped storage are used in a complementary dispatch system, the objective function is still to maximize the overall power generation of the system during the dispatch period.

[0238] ;

[0239] ;

[0240] Power generation conditions: ;

[0241] Pumping conditions: ;

[0242] In the formula: , Let be the real-time average output (in ten thousand kW) of the j-th pumped storage power station for power generation and pumping; Q is the number of pumped storage power stations. , Let be the power generation and pumping output coefficient of the j-th pumped storage power station; , These represent the power generation flow rate and pumping flow rate (m³) of the j-th pumped storage power station during time period t, respectively, in power generation and pumping modes. 3 / s); , For the j-th pumped storage power station, Average power generation head and average pumping head (m) during the time period.

[0243] This application introduces a dynamic loss rate threshold adaptive calculation mechanism. By constructing a capacity availability coefficient (sensing the real-time charge state of the pumped storage) and a peak-shaving urgency coefficient (sensing the overload pressure of the grid photovoltaic system), a fixed pumped storage benchmark efficiency is double-corrected in real time. This approach transforms the decision-making logic from static comparison optimization to dynamic perception: when the pumped storage capacity is tight, the activation threshold is automatically lowered to prevent overflow; when the grid peak-shaving pressure is high, the threshold is automatically relaxed to prioritize consumption, achieving precise adaptation of the complementary strategy to real-time operating conditions. This solves the problem of existing scheduling strategies being rigid and unable to adapt to dynamic operating conditions.

[0244] This application proposes a three-stage progressive scheduling process: pre-judgment, hierarchical optimization, and coordinated adjustment. Utilizing a rapid pre-estimation method based on output adjustment loss coefficients, the cost of hydropower regulation is predicted in milliseconds without running complex optimization models, locking in the target mode before optimization. Hierarchical processing reduces the computational dimensionality of nonlinear programming, improving the timeliness of day-ahead scheduling while ensuring decision-making accuracy. It solves the problems of low efficiency and slow convergence in traditional full-factor models.

[0245] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A multi-energy complementary and coordinated scheduling method based on dual peak-shifting compensation for hydro-solar-storage, characterized in that, include: Based on the pre-configured physical parameters of each power station, a variety of pre-set scheduling models are constructed, including independent water-solar scheduling model, water-solar complementary scheduling model, and water-solar-storage complementary scheduling model. Real-time collection of photovoltaic power output forecast data, hydropower inflow data, and initial status data of pumped storage power stations for the current time period; Based on photovoltaic power output forecast data and initial state data, and using preset conversion efficiency evaluation criteria, the target scheduling mode for the current period is determined from multiple preset scheduling models. Based on the photovoltaic power output forecast data and hydropower inflow data, the preset scheduling model corresponding to the target scheduling mode is solved, and the day-ahead scheduling instructions containing the power output plans of each power station are generated. Among them, the conversion efficiency evaluation criteria are used to evaluate the conversion relationship between hydropower regulation losses and photovoltaic consumption increments under different dispatch modes; The incremental photovoltaic power consumption assessed by the preset conversion efficiency evaluation criteria is determined by calculating the complementary power generation of the system. The system complementary power generation increase is defined as: the increase in the total power generation of the system obtained by solving the preset scheduling model corresponding to the target scheduling mode, relative to the total power generation of the system obtained by solving the independent scheduling model of water and solar power. Solving for the preset scheduling model corresponding to the target scheduling mode includes performing dual peak-shifting compensation: Based on photovoltaic power generation forecast data, the power generation schedule of cascade hydropower stations is adjusted to avoid the peak of photovoltaic power generation. During periods when photovoltaic power curtailment is identified, the pumped storage power station is controlled to operate in pumping mode to absorb the curtailed photovoltaic power. The hydropower regulation loss involved in the conversion efficiency assessment criteria is quantified as the hydropower compensation photovoltaic power loss rate; The hydropower compensation for photovoltaic power loss rate is defined as: the reduction in hydropower generation caused by the consumption of a unit of photovoltaic curtailment, and is calculated as the ratio of the reduction in hydropower generation in the complementary dispatch mode to the increase in photovoltaic grid-connected power generation in the independent dispatch mode. The target scheduling mode for the current time period is determined from multiple preset scheduling models, including the application of static decision logic based on preset fixed thresholds: Determine whether the power loss rate of hydropower compensating for photovoltaic power is lower than a preset fixed threshold; If so, the target scheduling mode will be determined as the water-solar complementary scheduling model; If not, or if it is determined that there is photovoltaic curtailment in the current period, then the target scheduling mode will be determined as the hydro-solar-storage complementary scheduling model.

2. The method according to claim 1, characterized in that, Multiple pre-defined scheduling models include independent water-solar scheduling model, water-solar complementary scheduling model, and water-solar-storage complementary scheduling model; Among them, the independent scheduling model of hydropower and solar power takes the optimal power generation of the hydropower station as the objective function; The hydro-solar complementary dispatch model takes the maximum total power generation of the system during the dispatch period as the objective function, and includes water balance constraints and power transmission channel constraints. The hydro-solar-storage complementary scheduling model adds energy balance constraints, water level constraints, and output constraints for pumped storage power stations to the existing hydro-solar complementary scheduling model.

3. The method according to claim 1, characterized in that, Based on photovoltaic power output forecast data and initial state data, the target scheduling mode for the current time period is determined from multiple preset scheduling models, including a three-stage progressive scheduling process of pre-judgment, hierarchical optimization, and coordinated adjustment. Phase 1: Based on the photovoltaic power output forecast data and the hydropower operation status under the independent hydropower dispatch mode obtained by solving the independent hydropower dispatch model, calculate the estimated hydropower marginal loss rate; Phase 2: Calculate the dynamic loss rate threshold for the current period based on the initial state data, and determine the target scheduling mode based on the comparison between the estimated hydropower marginal loss rate and the dynamic loss rate threshold. Phase 3: Solve the preset scheduling model corresponding to the target scheduling mode to obtain the output allocation results of each power station.

4. The method according to claim 3, characterized in that, Calculate the estimated marginal loss rate of hydropower, including: The predicted amount of photovoltaic curtailment is determined based on the photovoltaic power output data and the capacity of the power transmission channels. Based on the hydropower operation status and the pre-stored optimal hydropower operation efficiency curve, the output adjustment loss coefficient is calculated, which characterizes the degree to which the actual hydropower operation efficiency deviates from the optimal operation efficiency. Based on the predicted amount of photovoltaic curtailment, the power output adjustment loss coefficient, and the power output data under the independent hydropower dispatch mode, the estimated marginal loss rate of hydropower is calculated without running the hydropower-solar complementary dispatch model.

5. The method according to claim 4, characterized in that, Determine the target scheduling mode, including executing the following logic: Determine whether the estimated marginal loss rate of hydropower is less than the dynamic loss rate threshold, and at the same time determine whether the total predicted amount of photovoltaic curtailment is less than the available pumped storage capacity of the pumped storage power station. If both judgment conditions are met simultaneously, the target scheduling mode is determined to be the hydro-solar complementary scheduling model; otherwise, the target scheduling mode is determined to be the hydro-solar-storage complementary scheduling model.

6. The method according to claim 5, characterized in that, include: The dynamic loss rate threshold is obtained by applying a dual-factor correction to the pumped storage benchmark loss rate. The calculation formula is the product of the pumped storage benchmark loss rate, the capacity availability coefficient, and the peak shaving urgency coefficient. The pumped storage benchmark loss rate is determined based on the comprehensive cycle efficiency of the pumped storage power station. The capacity availability factor characterizes the available regulating capacity status of a pumped storage power station at the current moment. Its value is negatively correlated with the current stored energy. It is calculated as follows: the difference between the maximum stored energy of the pumped storage power station and the current stored energy, divided by the difference between the maximum stored energy of the pumped storage power station and the minimum stored energy. The peak shaving urgency factor characterizes the pressure state of the power grid in accepting photovoltaic power output during the current period. It is calculated by dividing the portion of the predicted photovoltaic power output during the current period that exceeds the capacity of the power transmission channel by the maximum pumping power of the pumped storage power station, and taking the minimum value of the obtained ratio and the value of 1 as the peak shaving urgency factor.

7. The method according to claim 3, characterized in that, Phase three also includes steps for implementing coordinated adjustments: Based on the solution results of the preset scheduling model corresponding to the target scheduling mode, the actual hydropower marginal loss rate is calculated. Determine whether the deviation between the actual and estimated marginal loss rates of hydropower exceeds a preset allowable deviation threshold. If so, update the estimated hydropower marginal loss rate using the actual hydropower marginal loss rate, and return to execution phase two to redetermine the target scheduling mode.

8. The method according to claim 4, characterized in that, Based on the predicted amount of photovoltaic curtailment, the power output adjustment loss coefficient, and the power output data under the independent hydropower dispatch mode, the estimated marginal loss rate of hydropower is calculated using the following method: Calculate the difference between the output under the independent hydropower dispatch mode and the preset minimum allowable hydropower output, and combine it with the output adjustment loss coefficient to perform time-period integration or accumulation to obtain the estimated reduction in hydropower generation. The estimated marginal loss rate of hydropower is obtained by dividing the estimated reduction in hydropower generation by the sum of the predicted amount of solar power curtailment.

Images