Power grid load forecasting method based on wind and light uncertainty constraints and application thereof

Abstract

This application relates to the field of artificial intelligence technology, and in particular to a power grid load forecasting method based on wind and solar uncertainty constraints and its application. It uses bulldozer distance to quantify the uncertainty scenarios of wind, solar, and load forecasting errors, thereby setting up a sample set containing these uncertain scenarios to construct a forecasting model. This enables power grid load forecasting that considers the uncertainties of wind and solar power generation in wind and solar farms; the aim is to solve the problem of how to overcome the interference of wind and solar uncertainty phenomena in wind and solar power generation on power grid load forecasting.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a power grid load forecasting method based on wind and solar uncertainty constraints and its application. Background Technology

[0002] With the popularization of electric vehicles and the large-scale integration of wind and solar power generation, the grid load is showing increased volatility and intermittency. Traditional grid load-side dispatching methods are difficult to adapt to the reliable coordinated charge dispatching of electric vehicles and hydrogen energy storage in the new scenario.

[0003] Compared to traditional thermal or hydropower, wind and solar power rely on natural resources and cannot precisely control output power on demand like thermal or hydropower. This "wind and solar uncertainty" leads to poor robustness of traditional grid load-side dispatching methods in actual dispatching processes.

[0004] In view of this, this application proposes a grid load-side dispatching method based on electric vehicles and hydrogen energy storage, which aims to take into account the uncertainties of wind and solar power generation and achieve reliable dispatching of the grid load side. Summary of the Invention

[0005] The main purpose of this application is to provide a grid load forecasting method based on wind and solar uncertainty constraints, aiming to solve the problem of how to overcome the interference of wind and solar uncertainty phenomena in wind and solar power plants on grid load forecasting.

[0006] To achieve the above objectives, this application provides a power grid load forecasting method based on wind and solar uncertainty constraints, the method comprising:

[0007] S10, determine the optimal power distribution point of the wind and solar power generation model and the probability corresponding to the optimal power distribution point respectively. Based on the optimal power distribution point and the probability, simulate the uncertainty scenario of wind and solar power to construct a load prediction model. The optimal power distribution point is obtained by prediction based on the bulldozer distance.

[0008] S20, determine the power grid load forecast result based on the load forecast model.

[0009] Optionally, in S10, based on the optimal power distribution point and probability, a wind-solar uncertainty scenario is simulated to construct a load forecasting model, specifically including:

[0010] Optimal power distribution point Sum of probabilities Constructing discrete scenes A quantile point covering a specified high-probability interval is calculated. and ;

[0011] The interval formed by the quantile points The fluctuation range of the prediction error forms the boundary of the set of uncertainties in the load prediction model.

[0012] Optionally, the calculation expression for the optimal power distribution point is:

[0013]

[0014] In the formula, The optimal power distribution point; Historical output of wind and solar power generation models is used as a random variable. , a continuous probability density function; ,in The number of discrete points is represented by r; the order is represented by r.

[0015] Optimal power distribution point Corresponding probability The expression is:

[0016]

[0017] In the formula, Indicates the first The right adjacent point of each point Indicates the first The left adjacent point of each sub-point.

[0018] Optionally, the wind and solar power generation model includes a photovoltaic power model, which satisfies the following constraints:

[0019]

[0020] In the formula, Let i be the rated power of the photovoltaic power station at node i. Let be the actual light intensity at time t. For standard test conditions, the light intensity is... The power temperature coefficient of a photovoltaic cell. The temperature of the photovoltaic cell at time t, The standard test condition temperature.

[0021] Optionally, the wind and solar power generation model includes a wind power model, which satisfies the following constraints:

[0022]

[0023] In the formula, For nodes The rated power of the wind turbine unit, Let be the wind speed at time t. Fan cut-in wind speed The rated wind speed of the fan. Cut off the wind speed for the fan.

[0024] In addition, to achieve the above objectives, this application also provides a power grid load forecasting method based on wind and solar uncertainty constraints as described above, and its application in power plant site selection and capacity determination and power grid dispatch.

[0025] Optionally, the power plant site selection and capacity determination include the following constraints:

[0026] (1) Location and volume constraints:

[0027] Minimize annual comprehensive cost For the goal:

[0028]

[0029] In the formula, The investment cost for hydrogen energy storage, For operating costs, For network loss costs;

[0030] in:

[0031]

[0032]

[0033]

[0034]

[0035] In the formula, This represents the set of candidate nodes, i.e., the set of power grid nodes within the planning area where hydrogen energy storage systems can be deployed; This represents the maximum number of hydrogen energy storage sites that can be built, which is an integer constraint upper limit of the upper-level planning. Represents a 0-1 decision variable, indicating whether or not a node is present. Constructing a hydrogen energy storage system, including =1 indicates construction. =0 means no construction; Representing 0-1 auxiliary variables, used to characterize nodes. The power cost item is active. Representing 0-1 auxiliary variables, used to characterize nodes. The energy capacity cost item is in active status; and They are nodes The rated power capacity and rated energy capacity of the hydrogen energy storage system at the location; and These are the minimum and maximum rated energy capacities of a single hydrogen energy storage system, respectively. , and These represent the minimum rated power and node of a single hydrogen energy storage system, respectively. The rated power at the location and the maximum rated power of a single system.

[0036] Optionally, power grid dispatch includes the following constraints:

[0037] (1) The goal is to minimize operating costs under the worst-case scenario:

[0038]

[0039] In the formula, This represents the unit cost of electricity at time t; This represents the net active power purchased by the system from the upstream power grid at time t. A negative value indicates that the system sells electricity to the upstream power grid. This represents the unit network loss cost factor at time t; This represents the total active power loss of the system at time t; This represents the penalty coefficient for system power imbalance or exceeding limits, used to enhance the robustness of the scheme; The absolute value term represents the system power imbalance or the extent to which safety constraints are exceeded at time t, and ensures that both positive and negative deviations are penalized. Indicates the total number of time periods in the scheduling cycle;

[0040] (2) Power balance constraint

[0041]

[0042] In the formula, and , representing the total active power actually injected into the system by the photovoltaic and wind farms at time t, respectively, are uncertain variables whose values ​​vary within the uncertainty set u; This represents the total discharge power of the hydrogen energy storage system at time t. This represents the total charging power of the hydrogen energy storage system at time t. This represents the conventional stationary active load at time t, excluding electric vehicles. This represents the total charging load power of the electric vehicle cluster at time t. This represents the set of uncertainties related to wind and solar power output and electric vehicle load.

[0043] (3) Equipment SOC operating constraints

[0044]

[0045] In the formula, It represents the average state of charge of the hydrogen energy storage system at time t, that is, the percentage of the current stored energy relative to the rated energy capacity; and These represent the minimum permissible state of charge for a hydrogen energy storage system and the minimum permissible state of charge for a hydrogen energy storage system, respectively.

[0046] (4) Robust Equivalent Transformation Constraints

[0047]

[0048]

[0049] In the formula, This represents the predicted value of photovoltaic output at time t. It represents the upper bound of the prediction error of photovoltaic power output at time t, that is, the maximum possible magnitude of the deviation of the actual output from the predicted value; Budgetary parameters representing the uncertainty of photovoltaics, where, ∈[0,1], used to adjust the robustness and conservatism; =1 indicates that the worst-case total error fluctuation is considered. =0 indicates that uncertainty is ignored; This represents the point-predicted value of the electric vehicle charging load at time t. This represents the maximum fluctuation deviation of the electric vehicle charging power at time t relative to the predicted value.

[0050] In addition, to achieve the above objectives, this application also provides a computer system, the computer system comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, it implements the steps of the power grid load forecasting method based on wind and solar uncertainty constraints as described in any of the preceding claims.

[0051] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the power grid load forecasting method based on wind and solar uncertainty constraints as described in any of the preceding claims.

[0052] This application has at least the following beneficial effects:

[0053] 1. The uncertainty scenario of wind and solar load prediction error is quantified by bulldozer distance, and a sample set containing the uncertainty scenario is set to build a prediction model, thereby realizing grid load prediction that takes into account the uncertainty phenomenon of wind and solar power generation.

[0054] 2. Construct a two-layer planning model for optimized scheduling and location-based capacity determination, and introduce robust constraints to enhance the system's anti-interference capability. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating the power grid load forecasting method based on wind and solar uncertainty constraints involved in the embodiments of this application; Figure 2 This is a physical architecture diagram of the power grid load forecasting method based on wind and solar uncertainty constraints involved in the embodiments of this application; Figure 3 This is a schematic diagram of the modeling of the wind and solar power generation model involved in the embodiments of this application; Figure 4 This is a schematic diagram of the hardware operating environment of the computer system involved in the embodiments of this application.

[0056] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art.

[0058] First Embodiment

[0059] Reference Figure 1 This embodiment provides a grid load forecasting method based on wind and solar uncertainty constraints. The method is applied to wind and solar power plants, which include pre-set wind and solar power generation models.

[0060] For example, refer to Figure 2 The diagram shows the overall architecture of a wind and solar power plant. The wind and solar power plant controls wind power generation and photovoltaic power generation through a pre-set wind and solar power generation model. Wind power generation supplies energy to coal-fired units, and photovoltaic power generation and energy storage system provide bidirectional energy supply. Wind power generation, photovoltaic power generation, coal-fired units and energy storage system are all distributed to the load side of users or the power grid through energy dispatching system.

[0061] The method includes the following steps:

[0062] S10, determine the optimal power distribution point of the wind and solar power generation model and the probability corresponding to the optimal power distribution point respectively. Based on the optimal power distribution point and the probability, simulate the uncertainty scenario of wind and solar power to construct a load prediction model. The optimal power distribution point is obtained by prediction based on the bulldozer distance.

[0063] In this embodiment, the input parameter for the wind and solar power plant is illumination. and wind speed Significant uncertainties exist, leading to a decrease in its photovoltaic output power. With wind power output It is also a random variable, and in order to carry out subsequent robust optimization scheduling and planning, it is necessary to mathematically model this uncertainty.

[0064] In this embodiment, we propose to approximate the prediction error of wind and solar power output or its normalized output as a continuous random variable, whose statistical characteristics can be described by a continuous probability density function (PDF).

[0065] Let the random variable of wind power or solar power output be... Its probability distribution is obtained from historical output data or statistical analysis of prediction errors, and its continuous probability density function is: Because continuous distributions are directly processed in optimization calculations. It is extremely complex. Therefore, this embodiment introduces the Wasserstein distance (also known as the bulldozer distance) for optimization. By minimizing the Wasserstein distance between the discrete distribution and the continuous distribution, the optimal discrete scenario is found to approximate the continuous distribution.

[0066] It's important to note that the bulldozer distance is a method for measuring the difference between two probability distributions. Its core idea is quite intuitive: imagine you need to move a pile of sand (the first probability distribution) and reshape it into a different, specified shape (the second probability distribution). In this process, you need to move a certain amount of "soil" (probability mass), and each piece of soil needs to be moved a certain distance. The Wasserstein distance measures the minimum "work" required to complete this "reshaping" project; this work is equal to the sum of the volume of soil moved multiplied by the distance traveled.

[0067] Compared to traditional metrics like KL divergence and JS divergence, Wasserstein distance has a key advantage: it effectively handles the case where two distributions have no overlap. When two distributions are completely non-overlapping, KL divergence becomes infinitely large and meaningless, while JS divergence becomes a constant, no longer providing information about the proximity of the distributions. Wasserstein distance, however, even in this situation, still provides a smoothly varying, meaningful value that reflects the true "distance" between the two distributions. This makes it extremely useful in machine learning, especially in training generative adversarial networks (GANs), as it provides a more stable and effective learning signal, thus mitigating problems such as model training failures or poor diversity in generated samples.

[0068] The following explains how to predict the optimal power distribution point and probability based on the bulldozer distance:

[0069] Assuming variables The continuous probability density function is , hoping to Approximation of a discrete scene with discrete points The optimal scene generation method based on Wasserstein distance then obtains the optimal point division. :

[0070]

[0071] with points Corresponding probability Calculate using the following formula:

[0072]

[0073] In the formula, The optimal power distribution point; Historical output of wind and solar power generation models is used as a random variable. , a continuous probability density function; ,in The number of discrete points is represented by r; the order is represented by r. Indicates the first The right adjacent point of each point Indicates the first The left adjacent point of each sub-point.

[0074] Obtain the optimal division point and the corresponding probability Next, we simulated uncertainties in wind and solar power scenarios to build a load forecasting model, specifically including:

[0075] S11, the optimal power distribution point Sum of probabilities Constructing discrete scenes A quantile point covering a specified high-probability interval is calculated. and ;

[0076] S12, the interval formed by the quantile points The fluctuation range of the prediction error forms the boundary of the set of uncertainties in the load prediction model.

[0077] In some alternative implementations, the boundary of the box-shaped uncertainty set specifically includes the box-shaped uncertainty set and the elliptical uncertainty set:

[0078] Define the predicted values ​​and fluctuation boundaries of wind and solar power output, and characterize the random fluctuation range of wind and solar power output to provide a basis for uncertainty description for subsequent robust scheduling.

[0079] (1) Box-type uncertainty set:

[0080]

[0081] In the formula, This represents the actual active power output of a wind farm or photovoltaic power station, which is an uncertain quantity that we cannot accurately predict or attempt to describe. This represents the predicted wind and solar power output at a specific point in time; it is the center of the set. This represents the prediction error, which is the deviation between the actual output and the predicted value. It represents the lower bound of the prediction error (usually a negative value or zero), indicating the maximum possible difference between the output and the predicted value. This represents the upper bound of the prediction error, indicating the maximum possible increase in power output compared to the predicted value. This fluctuation boundary information can typically be obtained from the error statistics between historical prediction data and actual power output data. In particular, the box-type uncertainty set, with its simple interval-form structure and linear tractability, is primarily used in the short-term dispatch layer. It extends the optimal discrete scenario generated based on Wasserstein distance into a well-defined power fluctuation boundary and forms a linear constraint through robust equivalence transformation, efficiently ensuring the feasibility of the power grid in real-time operation when facing extreme single fluctuations.

[0082] (2) Elliptic uncertainty set:

[0083]

[0084] In the formula, This represents the set of candidate nodes, i.e., the set of power grid nodes (buses) within the planning area where hydrogen energy storage systems can be deployed. This represents the maximum allowed number of hydrogen energy storage sites, which is an integer upper limit of the upper-level planning constraints. Represents a 0-1 decision variable, indicating whether or not a node is present. Construction of hydrogen energy storage system ( =1 indicates construction. =0 means no construction). Representing 0-1 auxiliary variables, used to characterize nodes. The activation status of the power cost item. Representing 0-1 auxiliary variables, used to characterize nodes. The energy capacity cost item is active. and They are nodes The rated power capacity and rated energy capacity of the hydrogen energy storage system at the location, and These are the minimum and maximum rated energy capacities of a single hydrogen energy storage system, respectively. , and These represent the minimum rated power and node of a single hydrogen energy storage system, respectively. The rated power at the location and the maximum rated power of a single system.

[0085] In particular, the elliptic uncertainty set focuses on risk assessment at the medium- to long-term planning level. It characterizes the spatiotemporal correlation of wind and solar power output through the covariance matrix, addressing the over-conservatism problem that may arise from the box set. In the planning stage, it is used to assess the robustness of hydrogen energy storage configurations in the face of complex joint fluctuations, achieving synergy from micro-level operational defense to macro-level risk-economic trade-offs.

[0086] Furthermore, and optionally, based on the uncertainty set defined in the above formula, the fluctuation range of wind and solar power output is transformed into a robust equivalence constraint of the scheduling model, and the optimization problem of the "worst-case scenario" is transformed into a solvable deterministic constraint:

[0087]

[0088] In the formula, Indicates time The actual power output of the wind and solar energy is an uncertain variable. Indicates time The predicted wind and solar power output is the center of the interval; , They represent time. The lower and upper bounds of the prediction error. They define the maximum negative and positive magnitudes by which the output may deviate from the predicted value, and are usually determined based on historical error data; This indicates the prediction error.

[0089] This constraint ensures the scheduling scheme is feasible across the entire fluctuation range of wind and solar power output. Interval optimization is employed to handle wind and solar uncertainties, ensuring the scheduling scheme's feasibility in the worst-case scenario. The goal of interval optimization is to find an economical and robust scheduling scheme. It does not pursue "optimality" in all scenarios, but prioritizes ensuring "feasibility in the worst-case scenario" before optimizing economics. The core advantage of this method is that it only requires knowledge of the predicted wind and solar power output and its possible fluctuation range (i.e., the upper and lower bounds of the interval), without needing difficult-to-obtain probability distribution information. To ensure the system does not become unbalanced in the worst-case scenario, the key power balance constraint needs to hold throughout the entire uncertainty interval.

[0090] S20, determine the power grid load forecast result based on the load forecast model.

[0091] After the processing in step S10, a load forecasting model that takes into account the uncertainty of wind and solar power is constructed. Based on the load forecasting model, the grid load of the wind and solar power plant is predicted, and the grid load forecasting results are obtained.

[0092] The focus of this embodiment is to describe how to construct a load forecasting model that takes into account the uncertainties of wind and solar power; the specific forecasting results are not described in this embodiment.

[0093] In the technical solution provided in this embodiment, the uncertainty scenario of wind and solar load prediction error is quantified by bulldozer distance, thereby setting up a sample set containing the uncertainty scenario to construct a prediction model, thereby realizing grid load prediction that takes into account the uncertainty phenomenon of wind and solar power generation.

[0094] Second Embodiment

[0095] Based on the uncertainty modeling of the first embodiment, referring to Figure 3 This embodiment provides a modeling method for wind and solar power generation. Specifically, considering the active power output characteristics of photovoltaic and wind power and the reactive power regulation capability of the inverter, corresponding power modes are established for different types of distributed power sources.

[0096] (1) Photovoltaic power model

[0097]

[0098] In the formula, Let i be the rated power of the photovoltaic power station at node i. Let be the actual light intensity at time t. For standard test conditions, the light intensity is... The power temperature coefficient of a photovoltaic cell. The temperature of the photovoltaic cell at time t, The standard test condition temperature.

[0099] This model comprehensively considers the effects of light intensity and temperature on the output power of photovoltaic cells.

[0100] (2) Wind power model

[0101] Wind power generation is closely related to wind speed and is generally described using a piecewise function.

[0102]

[0103] In the formula, For nodes The rated power of the wind turbine unit, Let be the wind speed at time t. Fan cut-in wind speed The rated wind speed of the fan. Cut off the wind speed for the fan.

[0104] This piecewise function model accurately calculates the output power of wind turbines in different wind speed ranges based on the relationship between wind speed and wind power.

[0105] As one implementation scheme, this embodiment provides a power grid load forecasting method based on wind and solar uncertainty constraints as described above, and its application in power plant site selection and capacity determination and power grid dispatch.

[0106] Third Embodiment

[0107] Based on any of the foregoing embodiments, in this embodiment, the power grid dispatch includes an electric vehicle charging station model, which includes a dynamic equation for charging power:

[0108]

[0109] In the formula, for Total charging power at all times; For the number of electric vehicles; For the first The vehicle's state of charge factor; This is the rated charging power.

[0110] Specifically, in some alternative implementations, grid dispatch includes a distributed energy storage system (DESS), wherein the operating constraint equations of the DESS include:

[0111] (1) Charging and discharging power limitation

[0112]

[0113]

[0114] In the formula, Hydrogen energy storage system The charging and discharging power at any given time; Define the maximum charging and discharging power of the hydrogen energy storage system. Clearly define the charging and discharging power range of the hydrogen energy storage system at any given time to prevent overcharging and over-discharging, and ensure equipment lifespan and safety.

[0115] (2) Energy state update equation

[0116]

[0117] In the formula, For hydrogen energy storage systems in Energy storage at any given moment; The energy stored at time t-1; , These are the charging efficiency coefficient and the discharging efficiency coefficient, respectively. The time interval is represented by . This equation reflects the energy conversion relationship of the hydrogen energy storage system during the charging and discharging process, taking into account charging and discharging efficiency, and accurately calculating the energy storage capacity at different times.

[0118] (3) State of charge (SOC) constraint

[0119]

[0120]

[0121] In the formula, This refers to the rated energy capacity of the hydrogen energy storage system. , These are the minimum and maximum state of charge (SOC) limits, respectively. By setting upper and lower limits for SOC, this equation avoids deep charging and discharging of the hydrogen energy storage system, extends equipment lifespan, and ensures system operational reliability.

[0122] Fourth embodiment

[0123] Based on any of the foregoing embodiments, in this embodiment, a two-layer planning model of optimized scheduling and site selection / capacity determination is constructed to achieve power plant site selection / capacity determination at the upper layer and power grid scheduling at the lower layer. Specifically:

[0124] The power plant site selection and capacity determination include the following constraints:

[0125] (1) Location and volume constraints:

[0126] Minimize annual comprehensive cost For the goal:

[0127]

[0128] In the formula, The investment cost for hydrogen energy storage, For operating costs, For network loss costs;

[0129] in:

[0130]

[0131]

[0132]

[0133]

[0134] In the formula, This represents the set of candidate nodes, i.e., the set of power grid nodes (buses) within the planning area where hydrogen energy storage systems can be deployed. This represents the maximum allowed number of hydrogen energy storage sites, which is an integer upper limit of the upper-level planning constraints. Represents a 0-1 decision variable, indicating whether or not a node is present. Construction of hydrogen energy storage system ( =1 indicates construction. =0 means no construction). Representing 0-1 auxiliary variables, used to characterize nodes. The activation status of the power cost item. Representing 0-1 auxiliary variables, used to characterize nodes. The energy capacity cost item is active. and They are nodes The rated power capacity and rated energy capacity of the hydrogen energy storage system at the location, and These are the minimum and maximum rated energy capacities of a single hydrogen energy storage system, respectively. , and These represent the minimum rated power and node of a single hydrogen energy storage system, respectively. The rated power at the location and the maximum rated power of a single system.

[0135] In this embodiment, the set of decision variables obtained from solving the upper-level planning problem... The location, rated power, and capacity of the hydrogen energy storage system will be used as fixed parameters input into the lower-level real-time scheduling model, forming the physical resource boundary available for scheduling.

[0136] For the lower-level grid dispatch, in this embodiment, the lower-level real-time dispatch model uses a dispatch period of 15 minutes or 1 hour. Based on the hydrogen energy storage system configuration, ultra-short-term wind and solar power forecasts, and electric vehicle load forecasts determined by the upper-level planning, it continuously solves for the optimal dispatch plan. Its core is to formulate adjustment instructions for grid interaction power, hydrogen energy storage system charging and discharging power, and electric vehicle charging load. The dispatch objective is to minimize the total operating cost (including electricity purchase cost and grid loss cost) under the worst-case uncertainty scenario, while satisfying all equipment operating constraints and system power balance. Dispatch instructions are issued to the controllers of each site, achieving peak shaving and valley filling, smoothing fluctuations, and improving the absorption of new energy sources by adjusting the power of electrolyzers / fuel cells and guiding the orderly charging of electric vehicles. This dispatch process forms a closed loop: the dispatch center generates and issues a set of dispatch instructions based on the model solution results, including the grid power purchase plan, the hydrogen energy storage system charging and discharging power setpoints, and electric vehicle charging guidance signals. After receiving the instructions, each site controller executes local control and feeds back the actual operating status (such as SOC and actual charging power) to the dispatch center. The scheduling center combines the latest ultra-short-term forecast data and executes the above optimizations on a rolling basis in the next cycle to achieve dynamic closed-loop scheduling.

[0137] Based on the facility locations and capacities determined by the upper-level planning, the goal is to minimize operating costs under the worst-case scenario:

[0138]

[0139] In the formula, This represents the unit cost of electricity at time t; This represents the net active power purchased by the system from the upstream power grid at time t. A negative value indicates that the system sells electricity to the upstream power grid. This represents the unit network loss cost factor at time t; This represents the total active power loss of the system at time t; This represents the penalty coefficient for system power imbalance or exceeding limits, used to enhance the robustness of the scheme; The absolute value term represents the system power imbalance or the extent to which safety constraints are exceeded at time t, and ensures that both positive and negative deviations are penalized. This indicates the total number of time periods in the scheduling cycle.

[0140] The following constraints must be met:

[0141] (1) Power balance constraint

[0142]

[0143] In the formula, This represents the unit cost of electricity at time t; This represents the net active power purchased by the system from the upstream power grid at time t. A negative value indicates that the system sells electricity to the upstream power grid. This represents the unit network loss cost factor at time t; This represents the total active power loss of the system at time t; This represents the penalty coefficient for system power imbalance or exceeding limits, used to enhance the robustness of the scheme; The absolute value term represents the system power imbalance or the extent to which safety constraints are exceeded at time t, and ensures that both positive and negative deviations are penalized. This indicates the total number of time periods in the scheduling cycle.

[0144] (2) Equipment operation constraints (taking SOC as an example)

[0145]

[0146] In the formula, It represents the average state of charge of the hydrogen energy storage system at time t, that is, the percentage of the current stored energy relative to the rated energy capacity; and These represent the minimum permissible state of charge (SPC) and the minimum permissible state of charge (SPC) of a hydrogen energy storage system, respectively.

[0147] (3) Robust Equivalent Transformation Constraints

[0148]

[0149]

[0150] In the formula, This represents the predicted value of photovoltaic output at time t. It represents the upper bound of the prediction error of photovoltaic power output at time t, that is, the maximum possible magnitude of the deviation of the actual output from the predicted value; Budget parameters representing photovoltaic uncertainty ( ∈[0,1]), used to adjust the robustness and conservatism. =1 indicates that the worst-case total error fluctuation is considered. =0 indicates that uncertainty is ignored; This represents the point-predicted value of the electric vehicle charging load at time t. This represents the maximum fluctuation deviation of the electric vehicle charging power at time t relative to the predicted value.

[0151] In this embodiment, the lower-level real-time scheduling model operates under the physical constraints determined by the upper-level planning. The worst-case scenario operating cost obtained by the lower-level model is fed back to the upper-level model to evaluate the economy and robustness of the site selection and capacity sizing scheme. Through this bidirectional iteration, efficient coordination between long-term planning and short-term operation is achieved.

[0152] As one implementation scheme, Figure 4 This is a schematic diagram of the hardware operating environment of the computer system involved in the embodiments of this application.

[0153] like Figure 4 As shown, the computer system may include: a processor 1001, such as a CPU; a memory 1005; a user interface 1003; a network interface 1004; and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0154] Those skilled in the art will understand that Figure 4 The computer system architecture shown does not constitute a limitation on the computer system and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0155] like Figure 4 As shown, the memory 1005, as a storage medium, may include an operating system, a network communication module, a user interface module, and computer programs. The operating system is a program that manages and controls the hardware and software resources of the computer system, as well as the operation of the computer programs and other software or programs.

[0156] exist Figure 4 In the computer system shown, the user interface 1003 is mainly used to connect to the terminal and communicate data with the terminal; the network interface 1004 is mainly used to communicate data with the backend server; and the processor 1001 can be used to call the computer program stored in the memory 1005.

[0157] In this embodiment, the computer system includes: a memory 1005, a processor 1001, and a computer program stored in the memory and executable on the processor, wherein:

[0158] When processor 1001 calls a computer program stored in memory 1005, it performs the following operations:

[0159] S10, determine the optimal power distribution point of the wind and solar power generation model and the probability corresponding to the optimal power distribution point respectively. Based on the optimal power distribution point and the probability, simulate the uncertainty scenario of wind and solar power to construct a load prediction model. The optimal power distribution point is obtained by prediction based on the bulldozer distance.

[0160] S20, determine the power grid load forecast result based on the load forecast model.

[0161] Furthermore, those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in a computer system to implement the process steps of the embodiments of the above methods.

[0162] Therefore, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the various steps of the power grid load forecasting method based on wind and solar uncertainty constraints as described in the above embodiments.

[0163] The computer-readable storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0164] It should be noted that, since the storage medium provided in the embodiments of this application is the storage medium used to implement the methods of the embodiments of this application, those skilled in the art can understand the specific structure and variations of the storage medium based on the methods described in the embodiments of this application, and therefore will not be repeated here. All storage media used in the methods of the embodiments of this application fall within the scope of protection of this application.

[0165] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0167] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0169] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0170] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A power grid load forecasting method based on wind and solar uncertainty constraints, characterized in that, Applied to wind and solar power plants, wherein the wind and solar power plants include pre-set wind and solar power generation models, the method includes the following steps: S10, determine the optimal power distribution point of the wind and solar power generation model and the probability corresponding to the optimal power distribution point respectively. Based on the optimal power distribution point and the probability, simulate the uncertainty scenario of wind and solar power to construct a load prediction model. The optimal power distribution point is obtained by prediction based on the bulldozer distance. S20, determine the power grid load forecast result based on the load forecast model.

2. The power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 1, characterized in that, In S10, based on the optimal power distribution point and probability, a load forecasting model is constructed by simulating uncertainties in wind and solar power scenarios, specifically including: Optimal power distribution point Sum of probabilities Constructing discrete scenes A quantile point covering a specified high-probability interval is calculated. and ; The interval formed by the quantile points The fluctuation range of the prediction error forms the boundary of the set of uncertainties in the load prediction model.

3. The power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 2, characterized in that, The calculation expression for the optimal power distribution point is: ； In the formula, This is the optimal power distribution point; Historical output of wind and solar power generation models is used as a random variable. , a continuous probability density function; ,in The number of discrete points is represented by r; the order is represented by r. Optimal power distribution point Corresponding probability The expression is: ； In the formula, Indicates the first The right adjacent point of each point Indicates the first The left adjacent point of each sub-point.

4. The power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 1, characterized in that, The wind and solar power generation model includes a photovoltaic power model, which satisfies the following constraints: ； In the formula, Let i be the rated power of the photovoltaic power station at node i. Let be the actual light intensity at time t. For standard test conditions, the light intensity is... The power temperature coefficient of a photovoltaic cell. The temperature of the photovoltaic cell at time t, The standard test condition temperature.

5. The power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 1 or 4, characterized in that, The wind and solar power generation model includes a wind power model, which satisfies the following constraints: ； In the formula, For nodes The rated power of the wind turbine unit, Let be the wind speed at time t. Fan cut-in wind speed The rated wind speed of the fan. Cut off the wind speed for the fan.

6. An application of a power grid load forecasting method based on wind and solar uncertainty constraints as described in any one of claims 1 to 5 in power plant site selection and capacity determination and power grid dispatching.

7. The application of the power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 6 in power plant site selection and capacity determination and power grid dispatching, characterized in that, The power plant site selection and capacity determination include the following constraints: (1) Location and volume constraints: Minimize annual comprehensive cost For the goal: ； In the formula, The investment cost for hydrogen energy storage, For operating costs, For network loss costs; in: ； ； ； ； In the formula, This represents the set of candidate nodes, i.e., the set of power grid nodes within the planning area where hydrogen energy storage systems can be deployed; This represents the maximum number of hydrogen energy storage sites that can be built, which is an integer constraint upper limit of the upper-level planning. Represents a 0-1 decision variable, indicating whether or not a node is present. Constructing a hydrogen energy storage system, including =1 indicates construction. =0 means no construction; Representing 0-1 auxiliary variables, used to characterize nodes. The power cost item is active. Representing 0-1 auxiliary variables, used to characterize nodes. The energy capacity cost item is in active status; and They are nodes The rated power capacity and rated energy capacity of the hydrogen energy storage system at the location; and These are the minimum and maximum rated energy capacities of a single hydrogen energy storage system, respectively. , and These represent the minimum rated power and node of a single hydrogen energy storage system, respectively. The rated power at the location and the maximum rated power of a single system.

8. The power grid load forecasting method based on wind and solar uncertainty constraints as described in claim 7, applied in power plant site selection and capacity determination and power grid dispatching, characterized in that, Power grid dispatch includes the following constraints: (1) The goal is to minimize operating costs under the worst-case scenario: ； In the formula, This represents the unit cost of electricity at time t; This represents the net active power purchased by the system from the upstream power grid at time t. A negative value indicates that the system sells electricity to the upstream power grid. This represents the unit network loss cost factor at time t; This represents the total active power loss of the system at time t; This represents the penalty coefficient for system power imbalance or exceeding limits, used to enhance the robustness of the scheme; The absolute value term represents the system power imbalance or the extent to which safety constraints are exceeded at time t, and ensures that both positive and negative deviations are penalized. Indicates the total number of time periods in the scheduling cycle; (2) Power balance constraint: ； In the formula, and , representing the total active power actually injected into the system by the photovoltaic and wind farms at time t, respectively, are uncertain variables whose values ​​vary within the uncertainty set u; This represents the total discharge power of the hydrogen energy storage system at time t. This represents the total charging power of the hydrogen energy storage system at time t. This represents the conventional stationary active load at time t, excluding electric vehicles. This represents the total charging load power of the electric vehicle cluster at time t. This represents the set of uncertainties related to wind and solar power output and electric vehicle load. (3) Equipment SOC operating constraints: ； In the formula, It represents the average state of charge of the hydrogen energy storage system at time t, that is, the percentage of the current stored energy relative to the rated energy capacity; and These represent the minimum permissible state of charge for a hydrogen energy storage system and the minimum permissible state of charge for a hydrogen energy storage system, respectively. (4) Robust equivalence transformation constraint: ； ； In the formula, This represents the predicted value of photovoltaic output at time t. It represents the upper bound of the prediction error of photovoltaic power output at time t, that is, the maximum possible magnitude of the deviation of the actual output from the predicted value; Budgetary parameters representing the uncertainty of photovoltaics, where, ∈[0,1], used to adjust the robustness and conservatism; =1 indicates that the worst-case total error fluctuation is considered. =0 indicates that uncertainty is ignored; This represents the point-predicted value of the electric vehicle charging load at time t. This represents the maximum fluctuation deviation of the electric vehicle charging power at time t relative to the predicted value.

9. A computer system, characterized in that, The computer system includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the power grid load forecasting method based on wind and solar uncertainty constraints as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the power grid load forecasting method based on wind and solar uncertainty constraints as described in any one of claims 1 to 5.

Images