An electric field micro-siting optimization method and device, electronic equipment and storage medium

Abstract

This invention provides a method, apparatus, electronic device, and storage medium for microscopic site selection optimization in electric fields, relating to the field of new energy technology. The method includes: constructing a wake velocity model based on wake characteristic parameters; constructing a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory; constructing a wind direction and wind speed model based on wind direction and wind speed parameters according to the Weibull probability distribution; obtaining a power model based on initial wind speed, wind speed characteristics, wind direction characteristics, and the power curves of the power generation devices; constructing a state space based on the wind speed, wind direction, and location of the power generation devices; optimizing the state space through an Actor network and the power model, and constructing an action space with the goal of maximizing power generation; optimizing the state space set and action space set through a Critic network, and determining the action space result with the goal of maximizing total power generation, for constructing the site selection layout of the power generation devices.

Description

Technical Field

[0001] This invention relates to the field of new energy technology, and more specifically, to a method, apparatus, electronic device, and storage medium for optimizing microscopic electric field location. Background Technology

[0002] In a wind farm, the location of different power generation equipment determines the amount of power generated. In large-scale wind farms, the layout design of power generation equipment is a key factor affecting the economics of the wind farm project. An inappropriate wind farm layout design will result in wind power output that is lower than expected and will increase operation and maintenance costs.

[0003] In related technologies, there are no unified standards and specifications for micro-site selection. In actual wind farm projects, empirical layout methods are usually adopted. This method arranges wind turbines in the wind farm according to the dominant prevailing wind direction, using symmetrical forms such as matrix, row, and quincunx. It is mainly suitable for flat terrain and wind farms with one or more prevailing wind directions, and its application scope in engineering projects is limited. Summary of the Invention

[0004] The problem addressed by this invention is how to construct the site selection layout of power generation equipment.

[0005] To address the above problems, this invention provides a method, apparatus, electronic device, and storage medium for optimizing electric field micro-location.

[0006] In a first aspect, the present invention provides a method for optimizing microscopic location of electric fields, comprising:

[0007] Construct a wake velocity model based on wake characteristic parameters;

[0008] Based on the kinetic energy balance theory, a wake effect model between power generation devices is constructed based on the wake velocity model, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices;

[0009] Based on the Weibull probability distribution, a wind direction and wind speed model is constructed based on wind direction and wind speed parameters, wherein the wind direction and wind speed model is used to obtain the wind speed and wind direction characteristics of the power generation equipment;

[0010] A power model is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment;

[0011] The wind speed of the power generation equipment is determined based on the initial wind speed and the wind speed characteristics, the wind direction of the power generation equipment is determined based on the wind direction characteristics, and a state space is constructed based on the wind speed, the wind direction, and the location of the power generation equipment.

[0012] The state space is optimized using an Actor network and the power model, and an action space is constructed with the goal of maximizing power generation. The action space includes the optimized position of the power generation equipment.

[0013] The state space set and action space set are optimized by using a Critic network to determine the action space result with the goal of maximizing total power generation. The state space set and action space set are composed of the state space and action space of all the power generation equipment.

[0014] Optionally, constructing the wake velocity model based on wake characteristic parameters includes:

[0015] The wake velocity model is constructed based on the free-flow wind speed, turbine thrust coefficient, turbine rotor diameter, and wake diffusion constant. The wake velocity model is used to obtain the wake velocity at a given downstream distance.

[0016] Optionally, the step of constructing a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory includes:

[0017] The wake velocities of other power generation devices at the location of the target power generation device are obtained based on the wake velocity model.

[0018] Based on the distance between the target power generation equipment and the other power generation equipment along the wind direction, and the speed deficit ratio of the other power generation equipment on the target power generation equipment, a wake effect model between the target power generation equipment and the other power generation equipment is constructed.

[0019] Optionally, the step of constructing a wind direction and wind speed model based on wind direction and wind speed parameters according to the Weibull probability distribution includes:

[0020] Construct a probability distribution function based on the Weibull probability distribution;

[0021] The wind speed shape parameter is determined based on the range of wind speed variation, and the wind speed scale parameter is determined based on the average wind speed, wherein the wind speed parameter includes the wind speed shape parameter and the wind speed scale parameter;

[0022] Substitute the wind speed shape parameter and the wind speed scale parameter into the probability distribution function to determine the wind speed density function;

[0023] The wind speed is discretized to obtain the number of wind speed segments and the cut-in and cut-out wind speeds corresponding to each wind speed segment;

[0024] Substituting the cut-in wind speed and cut-out wind speed, wind direction shape parameter and wind direction scale parameter of the wind direction segment into the probability distribution function, the wind direction density function of the wind direction segment is determined, wherein the wind direction parameter includes the wind direction shape parameter and the wind direction scale parameter;

[0025] The wind speed density function and the wind direction density function are used as the wind direction and wind speed model.

[0026] Optionally, obtaining the power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment includes:

[0027] The power curve of the power generation equipment is obtained by linear interpolation;

[0028] Based on the target wind direction and target wind speed of the target power generation equipment, the initial wind speed, the wind speed characteristics, and the wind direction characteristics are obtained according to the wake effect model and the wind direction and wind speed model.

[0029] The power model of the power generation equipment is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve.

[0030] Optionally, optimizing the state space using the Actor network and the power model to construct the action space with the goal of maximizing power generation includes:

[0031] A minimum distance constraint is constructed based on the Euclidean distance between the two power generation devices and a preset minimum spacing.

[0032] Establish boundary constraints based on the electric field space boundary.

[0033] Optionally, the step of optimizing the state space set and action space set through a Critic network to determine the action space result with the objective of maximizing total power generation includes:

[0034] Construct a value function based on the state space set and the action space set, wherein the value function is used to evaluate the expected reward of performing an action in a given state;

[0035] Using the maximum total power generation as a constraint on the expected return, the action space outcome is determined through the value function.

[0036] Secondly, the present invention also provides an electric field micro-location optimization device, comprising:

[0037] A standalone wake velocity module is used to construct a wake velocity model based on wake characteristic parameters;

[0038] The wake effect module is used to construct a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices.

[0039] The wind direction and speed module is used to construct a wind direction and speed model based on wind direction parameters and wind speed parameters according to the Weibull probability distribution, wherein the wind direction and speed model is used to obtain the wind speed characteristics and wind direction characteristics of the power generation equipment;

[0040] A power module is used to obtain a power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment.

[0041] A space construction module is used to determine the wind speed of the power generation equipment based on the initial wind speed and the wind speed characteristics, determine the wind direction of the power generation equipment based on the wind direction characteristics, and construct a state space based on the wind speed, wind direction, and position of the power generation equipment;

[0042] The Actor network module is used to optimize the state space through the Actor network and the power model, and construct the action space with the goal of maximizing power generation, wherein the action space includes the optimized position of the power generation equipment;

[0043] The Critic network module is used to optimize the state space set and action space set through the Critic network to determine the action space result with the goal of maximizing the total power generation. The state space set and the action space set are composed of the state space and the action space of all the power generation equipment.

[0044] Thirdly, the present invention provides an electronic device, including a memory and a processor;

[0045] The memory is used to store computer programs;

[0046] The processor is configured to implement the electric field micro-location optimization method as described in the first aspect when executing the computer program.

[0047] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the electric field micro-addressing optimization method as described in the first aspect.

[0048] The beneficial effects of the electric field micro-location optimization method of the present invention are:

[0049] A mapping relationship between wake characteristic parameters and wake velocity is established using wind speed data, quantifying the impact of each parameter on wake velocity. A wake effect model is used to describe the range and intensity of the wind speed reduction region formed downstream of a single wind turbine, providing a quantitative basis for assessing aerodynamic interference between adjacent turbines. Combining the wake velocity model and the kinetic energy balance principle, a wake effect model reflecting the impact of upstream power generation equipment on the downstream region is established. By introducing an energy conservation mechanism, the assessment of available wind resources for downstream turbines becomes more reasonable. This initial wind speed serves as the basic input for subsequent power prediction and state modeling, enhancing the adaptability of the entire modeling chain to complex wind field environments. A Weibull distribution is used for probabilistic modeling of wind speed, combined with wind direction frequency distribution, forming a joint wind direction and wind speed model. This model comprehensively characterizes the temporal distribution characteristics and directional patterns of wind resources, enabling layout design to not only consider wind energy magnitude but also match the dominant wind path, improving the overall wind energy capture efficiency. By combining initial wind speed, wind speed characteristics, and wind direction characteristics with the power curves of power generation equipment, the system predicts the output power under specific wind conditions, achieving an end-to-end mapping from environmental input to power generation response. This enables the estimation of power generation performance under different location configurations, allowing for layout adjustments guided by maximizing power generation and enhancing the economic orientation of site selection decisions. The Actor network receives the state information of each power generation device and outputs its location adjustment actions, generating an action space with the goal of maximizing total power generation. It can explore the optimal location adjustment direction while considering the interaction between itself and other power generation devices. The Critic network evaluates the value of all device state and action combinations, achieving multi-agent collaborative optimization. It is used to balance individual and overall interests at the global level, avoiding local optimum traps. The final output action space represents the power generation equipment layout and site selection scheme that maximizes total power generation under current wind resource conditions. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the electric field micro-location optimization method according to an embodiment of the present invention;

[0051] Figure 2 This is a flowchart of the electric field micro-location optimization method according to an embodiment of the present invention;

[0052] Figure 3 This is an example diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0053] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0054] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0055] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0056] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0057] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0058] This embodiment provides a method, apparatus, electronic device, and storage medium for optimizing microscopic electric field location.

[0059] like Figure 1 As shown, an embodiment of the present invention provides an electric field micro-location optimization method, comprising:

[0060] Step S100: Construct a wake velocity model based on wake characteristic parameters.

[0061] By using wind speed data, a mapping relationship between wake characteristic parameters and wake velocity is established, quantifying the impact of each parameter on the wake velocity. The wake velocity output by the model can be used to assess the degree of wake interference between power generation equipment, providing data support for the layout design of wind farm power generation equipment and reducing the impact of wake losses on overall power generation efficiency.

[0062] Step S200: Based on the kinetic energy balance theory, a wake effect model between power generation devices is constructed based on the wake velocity model, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices.

[0063] Based on the kinetic energy balance theory, the kinetic energy relationship between upstream and downstream power generation equipment in the wake region is analyzed. By quantifying the relationship between wind speed attenuation and energy redistribution in the wake velocity model, a kinetic energy change equation for the wake region is established to reflect the dynamic characteristics of the interaction between power generation equipment. The output of the wake velocity model is combined with the kinetic energy balance equation to determine the wind speed disturbance characteristics of the wake region. The variation of wind speed in the wake region with spatial location is described.

[0064] Using a wake effect model, the initial wind speed of upstream power generation equipment is deduced from the measured wind speed of downstream equipment. By incorporating kinetic energy balance theory, the wake effect model can more realistically reflect the impact of wake interference between power generation equipment on wind speed, reducing the initial wind speed deviation caused by neglecting energy redistribution in traditional methods, and providing reliable input for power generation equipment power prediction. A dynamic compensation mechanism enables the model to adapt to complex operating conditions such as sudden wind direction changes and fluctuations in turbulence intensity, ensuring the applicability of the wake effect model under different conditions and reducing prediction errors caused by environmental changes.

[0065] Step S300: Based on the Weibull probability distribution, construct a wind direction and wind speed model based on wind direction and wind speed parameters, wherein the wind direction and wind speed model is used to obtain the wind speed and wind direction characteristics of the power generation equipment.

[0066] Statistical modeling of wind speed data is performed using the Weibull distribution function to calculate shape and scale parameters. This distribution reflects the randomness and skewness of wind speed and is suitable for wind energy resource analysis under different conditions. A joint wind direction-wind speed model is constructed by combining the Weibull distribution of wind speed with the frequency distribution of wind direction. Using the model's output, the probability distribution of wind speed and the frequency of wind direction for each wind direction are used to generate wind speed and wind direction characteristics for power generation equipment.

[0067] Step S400: Obtain a power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment.

[0068] The output power of a power generation device is predicted by integrating initial wind speed, wind speed characteristics, wind direction characteristics, and the power curve of the device. Power curve adaptation is then performed to extract parameters such as cut-in wind speed, rated wind speed, and cut-out wind speed to determine the expected output power of the device under specific wind field conditions.

[0069] Step S500: Determine the wind speed of the power generation equipment based on the initial wind speed and the wind speed characteristics; determine the wind direction of the power generation equipment based on the wind direction characteristics; and construct a state space based on the wind speed, wind direction, and location of the power generation equipment.

[0070] Based on the initial wind speed and wind speed characteristics, the wind speed of the current power generation equipment under the influence of other power generation equipment is determined. According to the wind direction and wind speed characteristics, the frequency of occurrence of the power generation equipment in each wind direction and the wind speed of the power generation equipment are determined. The wind speed, wind direction and position are jointly constructed to form the state space, which is the basis for the algorithm to make decisions and learn. It is used to input into the Actor network, and the Actor network optimizes the action space, that is, optimizes the position of a single power generation equipment based on the power obtained from the state space.

[0071] Step S600: Optimize the state space using the Actor network and the power model to construct an action space with the goal of maximizing power generation, wherein the action space includes the optimized position of the power generation equipment.

[0072] The Actor network is responsible for selecting the optimal action (position adjustment) for each agent (i.e., each wind turbine) to maximize the total power generation of the wind farm. The input to the Actor network is the state of each turbine, including its position, wind speed, and wind direction. The output of the Actor network is the action of each turbine, i.e., the position adjustment result.

[0073] Step S700: Optimize the state space set and action space set through the Critic network, and determine the action space result with the goal of maximizing the total power generation. The state space set and the action space set are composed of the state space and action space of all the power generation equipment.

[0074] The Critic network is responsible for evaluating the value of the combination of actions of each agent in the current state, that is, predicting the total power generation that these actions will bring. The Actor network is used to select the optimal actions, and the Critic network is used to evaluate the value of these actions. The input to the Critic network includes the states and actions of all turbines. Through this collaborative work, the MADDPG algorithm can effectively optimize the location of turbines in a wind farm to maximize power generation.

[0075] In this embodiment, a mapping relationship between wake characteristic parameters and wake velocity is established using wind speed data to quantify the impact of each parameter on the wake velocity. This is used to describe the range and intensity of the wind speed reduction area formed downstream of a single wind turbine, providing a quantitative basis for assessing aerodynamic interference between adjacent turbines. Combining the wake velocity model and the kinetic energy balance principle, a wake effect model reflecting the impact of upstream power generation equipment on the downstream area is established. By introducing an energy conservation mechanism, the assessment of available wind resources for downstream turbines becomes more reasonable. This initial wind speed serves as the basic input for subsequent power prediction and state modeling, enhancing the adaptability of the entire modeling chain to complex wind field environments. A Weibull distribution is used for probabilistic modeling of wind speed, combined with the wind direction frequency distribution, to form a joint wind direction and wind speed model. This model comprehensively characterizes the temporal distribution characteristics and directional patterns of wind resources, enabling the layout design to not only consider wind energy magnitude but also match the dominant wind path, improving the overall wind energy capture efficiency. By combining initial wind speed, wind speed characteristics, and wind direction characteristics with the power curves of power generation equipment, the system predicts the output power under specific wind conditions, achieving an end-to-end mapping from environmental input to power generation response. This enables the estimation of power generation performance under different location configurations, allowing for layout adjustments guided by maximizing power generation and enhancing the economic orientation of site selection decisions. The Actor network receives the state information of each power generation device and outputs its location adjustment actions, generating an action space with the goal of maximizing total power generation. It can explore the optimal location adjustment direction while considering the interaction between itself and other power generation devices. The Critic network evaluates the value of all device state and action combinations, achieving multi-agent collaborative optimization. It is used to balance individual and overall interests at the global level, avoiding local optimum traps. The final output action space represents the power generation equipment layout and site selection scheme that maximizes total power generation under current wind resource conditions.

[0076] Optionally, constructing the wake velocity model based on wake characteristic parameters includes:

[0077] The wake velocity model is constructed based on the free-flow wind speed, turbine thrust coefficient, turbine rotor diameter, and wake diffusion constant. The wake velocity model is used to obtain the wake velocity at a given downstream distance.

[0078] The wake velocity model is expressed as:

[0079] ,

[0080] in, This represents the wake velocity at a given downstream distance d. Indicates free-flow wind speed. Indicates the turbine thrust coefficient. Indicates the diameter of the turbine rotor. This represents the wake diffusion constant.

[0081] Optionally, the step of constructing a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory includes:

[0082] The wake velocities of other power generation devices at the location of the target power generation device are obtained based on the wake velocity model.

[0083] Based on the distance between the target power generation equipment and the other power generation equipment along the wind direction, and the speed deficit ratio of the other power generation equipment on the target power generation equipment, a wake effect model between the target power generation equipment and the other power generation equipment is constructed.

[0084] The wake effect model is expressed as:

[0085] ,

[0086] in, Indicates the first The initial wind speed of the generator unit, where N represents the total number of generator units in the power plant. Indicates the first The distance between the k-th power generation unit and the k-th power generation unit along the wind direction. This indicates that the k-th power generation device is in the... The wake velocity at the location of the generator unit. For the k-th power generation unit in the th... The speed deficit ratio of the generator equipment.

[0087] The speed deficit ratio is expressed as:

[0088] ,

[0089] in, Indicates the swept area of ​​the turbine rotor. This represents the overlap area between the wake of the k-th turbine and the rotor of the i-th turbine.

[0090] The wake effect model provides a crucial environmental feedback mechanism, enabling the algorithm to adjust the turbine positions of power generation equipment based on the impact of the wake effect, thereby improving the overall performance of the wind farm. The MADDPG algorithm allows multiple agents (turbines) to interact and collaborate within the wind farm model. Agent decisions are based not only on their own state but also on the actions of other agents and the overall state of the wind farm, simulating the wake effect and spatial distribution among turbines.

[0091] Optionally, the step of constructing a wind direction and wind speed model based on wind direction and wind speed parameters according to the Weibull probability distribution includes:

[0092] Construct a probability distribution function based on the Weibull probability distribution;

[0093] The wind speed shape parameter is determined based on the range of wind speed variation, and the wind speed scale parameter is determined based on the average wind speed, wherein the wind speed parameter includes the wind speed shape parameter and the wind speed scale parameter;

[0094] Substitute the wind speed shape parameter and the wind speed scale parameter into the probability distribution function to determine the wind speed density function;

[0095] The wind speed is discretized to obtain the number of wind speed segments and the cut-in and cut-out wind speeds corresponding to each wind speed segment;

[0096] Substituting the cut-in wind speed and cut-out wind speed, wind direction shape parameter and wind direction scale parameter of the wind direction segment into the probability distribution function, the wind direction density function of the wind direction segment is determined, wherein the wind direction parameter includes the wind direction shape parameter and the wind direction scale parameter;

[0097] The wind speed density function and the wind direction density function are used as the wind direction and wind speed model.

[0098] Parameters such as wind speed and direction for each wind turbine are input as state information into the MADDPG algorithm. The agent determines its actions based on this state information, adjusting the turbine's position to optimize the wind farm's performance. A rose diagram is used to describe the characteristics of wind direction changes. In the map, the wind direction is evenly divided into... Segment. The length of each segment. This indicates the annual wind frequency within that direction. Clearly, .

[0099] The Weibull probability distribution is used to describe the wind speed distribution characteristics. The wind speed density function is expressed as:

[0100] ,

[0101] in, Represents the wind speed density function. Indicates wind speed. Indicates wind speed shape parameters, Represents wind speed scale parameters. Wind speed shape parameters. Related to the range of wind speed variation, wind speed scale parameters It is related to the average wind speed.

[0102] The wind speed distribution, shape, and scale parameters may differ in each wind direction segment. For ease of calculation, in this embodiment of the invention, the wind speed range in each direction is discretized at intervals of 1 m / s. This indicates the cut-in wind speed of the turbine in the power generation equipment. This indicates the cut-out wind speed, and the number of wind speed segments. Represented as:

[0103] ,

[0104] No. Wind direction and speed and The annual frequency between them is expressed as:

[0105] ,

[0106] in, , , , and Let represent the wind direction shape parameter and wind direction scale parameter for the j-th wind direction, respectively.

[0107] Optionally, obtaining the power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment includes:

[0108] The power curve of the power generation equipment is obtained by linear interpolation;

[0109] Based on the target wind direction and target wind speed of the target power generation equipment, the initial wind speed, the wind speed characteristics, and the wind direction characteristics are obtained according to the wake effect model and the wind direction and wind speed model.

[0110] The power model of the power generation equipment is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve.

[0111] In the wake effect model and wind direction and speed model mentioned above, only the wind speed in one direction is considered. In fact, the state space used to construct the state space is a function of the turbine position, wind direction and wind speed.

[0112] No. The first generator unit The first wind direction The wind speed is expressed as:

[0113] ,

[0114] in, Indicates the first The first generator unit The first wind direction Section wind speed, Indicates the first The first wind direction The free-flow wind speed of the section, where N represents the number of power generation devices in the power plant. Indicates the first The k-th power generation device under the wind direction is in the... Speed ​​deficit ratio on the typhoon generator Indicates the first The kth generator and the kth generator are along the first Distance in one wind direction, Indicates the first Under the wind direction, the kth generating unit is in the... The wake velocity at the location of the power generation equipment.

[0115] Wind speed is The power generation of the power generation equipment can be expressed as:

[0116] ,

[0117] Where P(u) represents the power curve of a given water turbine, This indicates the cut-in wind speed of the turbine in the power generation equipment. This indicates the cut-off wind speed. Because... Since it cannot be represented by a single function, linear interpolation is used to evaluate it.

[0118] The total power of a wind farm is expressed as:

[0119] ,

[0120] in, Indicates total power. N represents the total number of wind sections, and N represents the number of power generation devices in the power plant. Indicates the wind speed segment. Indicates the first Power in the wind direction segment, Indicates the first Wind direction section Power in the wind speed range, Indicates the first The first generator unit The first wind direction Section wind speed.

[0121] The calculation of power generation provides a reward signal for the MADDPG algorithm. The agent increases power generation by performing actions (adjusting the position of the power generation equipment), and the algorithm rewards or punishes the agent based on the change in power generation, thereby guiding the agent to learn the optimal strategy.

[0122] Optionally, optimizing the state space using the Actor network and the power model to construct the action space with the goal of maximizing power generation includes:

[0123] A minimum distance constraint is constructed based on the Euclidean distance between the two power generation devices and a preset minimum spacing.

[0124] Establish boundary constraints based on the electric field space boundary.

[0125] set up The micro-location optimization problem is to determine the minimum allowable distance between turbines in a power generation system. Maximizing the power generation of the turbine is expressed as: .

[0126] There are positional constraints between the turbines in a power generation system, as shown below:

[0127] ,

[0128] C = N(N-1) / 2,

[0129] in, This represents the positional constraint between the i-th and k-th generators. Let represent the coordinates of the i-th and k-th generating units respectively, C represent the number of constraints, N represent the total number of generating units in the power plant, and L represent the side length of the power plant. If the power plant is rectangular or has other shapes, then corresponding constraints can be applied. Restrictions should be imposed.

[0130] The physical constraints in the wind farm model are incorporated into the MADDPG algorithm as constraints on the agent's actions. This ensures that the agent adheres to the actual physical limitations while searching for the optimal solution. Therefore, wind farm micro-situation is a constrained optimization problem with many constraints.

[0131] Optionally, the step of optimizing the state space set and action space set through a Critic network to determine the action space result with the objective of maximizing total power generation includes:

[0132] Construct a value function based on the state space set and the action space set, wherein the value function is used to evaluate the expected reward of performing an action in a given state;

[0133] Using the maximum total power generation as a constraint on the expected return, the action space outcome is determined through the value function.

[0134] Define the state space and action space:

[0135] like Figure 2 As shown, wind farm modeling defines the environment of the optimization problem, including parameters such as turbine location, wind speed, and wind direction. These parameters constitute the state space of the agent (turbine) in the MADDPG algorithm, which is the foundation for the algorithm's decision-making and learning. State Space It includes environmental information such as the location of all turbines in the wind farm, wind speed, and wind direction. For each turbine... Its state It can be represented as:

[0136] ,

[0137] in, Indicates turbine Location coordinates, Indicates turbine The wind speed at the location. Indicates turbine The wind direction at your location.

[0138] Action space A contains the movement actions that each turbine can take. For each turbine... Its actions It can be represented as:

[0139] ,

[0140] in, Indicates turbine exist Distance of movement in direction Indicates turbine exist The distance traveled in the direction. The entire wind farm state space. It is the set of all turbine states: Action space A is the set of all turbine actions: .

[0141] The MADDPG algorithm uses an Actor network to select optimal actions and a Critic network to evaluate the value of these actions. Through this collaborative work, the MADDPG algorithm can effectively optimize the location of turbines in a wind farm to maximize power generation.

[0142] The goal of an Actor network is to learn a policy. It chooses to be in a given state The following action To maximize the action value evaluated by the Critic network, the update formula for the Actor network is expressed as:

[0143] ,

[0144] in, Representation Strategy The performance metric, E, represents the mathematical expectation. This indicates the Critic network's information about actions. gradient, Indicates the state of the Actor network. The action to choose.

[0145] The goal of a Critic network is to learn a value function. It evaluates the state given Next action The expected return, the update formula for the Critic network, is expressed as:

[0146] ,

[0147] in, Indicates the execution of an action The instant reward obtained afterward This represents the preset discount factor used to calculate the present value of future returns. This is the next state of the target Critic network evaluation. And the next action The value of R is determined by constructing a value function Q based on the total power generation R:

[0148] ,

[0149] Where R represents the total power generation, Let N represent the power generation of the i-th power generation device, and N represent the total number of power generation devices in the power plant.

[0150] The loss function of the Critic network is expressed as:

[0151] ,

[0152] Where L(·) represents the loss function, and E represents the expected value. This represents the parameters of the Critic network, including the weights and biases of all layers.

[0153] To stabilize the training process, the MADDPG algorithm uses an objective function. The parameters of the objective function are updated using the following formula:

[0154] ,

[0155] in: The parameters represent the objective function. This represents the parameters of the current network. This represents the soft update coefficient of the target network update.

[0156] The optimization is performed using a Critic network. The action space set corresponding to the optimal value function obtained in each optimization is retained. Further iterative optimization is then performed on this basis until the maximum number of iterations is reached or the preset iteration condition is met. That is, if Q does not change significantly after several iterations, the iteration ends and the action space result is determined.

[0157] An embodiment of the present invention provides an electric field micro-location optimization device, comprising:

[0158] A standalone wake velocity module is used to construct a wake velocity model based on wake characteristic parameters;

[0159] The wake effect module is used to construct a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices.

[0160] The wind direction and speed module is used to construct a wind direction and speed model based on wind direction parameters and wind speed parameters according to the Weibull probability distribution, wherein the wind direction and speed model is used to obtain the wind speed characteristics and wind direction characteristics of the power generation equipment;

[0161] A power module is used to obtain a power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment.

[0162] A space construction module is used to determine the wind speed of the power generation equipment based on the initial wind speed and the wind speed characteristics, determine the wind direction of the power generation equipment based on the wind direction characteristics, and construct a state space based on the wind speed, wind direction, and position of the power generation equipment;

[0163] The Actor network module is used to optimize the state space through the Actor network and the power model, and construct the action space with the goal of maximizing power generation, wherein the action space includes the optimized position of the power generation equipment;

[0164] The Critic network module is used to optimize the state space set and action space set through the Critic network to determine the action space result with the goal of maximizing the total power generation. The state space set and the action space set are composed of the state space and the action space of all the power generation equipment.

[0165] like Figure 3 As shown, an electronic device 300 provided in this embodiment of the invention includes a memory 310 and a processor 320; the memory 310 is used to store a computer program; the processor 320 is used to implement the electric field micro-addressing optimization method as described above when the computer program is executed.

[0166] Alternatively, an electronic device 300 includes a memory 310 and a processor 320 coupled to the memory 310; the memory 310 is configured to store a computer program; and the processor 320 is configured to perform the following operations when the computer program is executed:

[0167] Construct a wake velocity model based on wake characteristic parameters;

[0168] Based on the kinetic energy balance theory, a wake effect model between power generation devices is constructed based on the wake velocity model, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices;

[0169] Based on the Weibull probability distribution, a wind direction and wind speed model is constructed based on wind direction and wind speed parameters, wherein the wind direction and wind speed model is used to obtain the wind speed and wind direction characteristics of the power generation equipment;

[0170] A power model is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment;

[0171] The wind speed of the power generation equipment is determined based on the initial wind speed and the wind speed characteristics, the wind direction of the power generation equipment is determined based on the wind direction characteristics, and a state space is constructed based on the wind speed, the wind direction, and the location of the power generation equipment.

[0172] The state space is optimized using an Actor network and the power model, and an action space is constructed with the goal of maximizing power generation. The action space includes the optimized position of the power generation equipment.

[0173] The state space set and action space set are optimized by using a Critic network to determine the action space result with the goal of maximizing total power generation. The state space set and action space set are composed of the state space and action space of all the power generation equipment.

[0174] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the electric field micro-addressing optimization method as described above.

[0175] Alternatively, a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the following operations:

[0176] Construct a wake velocity model based on wake characteristic parameters;

[0177] Based on the kinetic energy balance theory, a wake effect model between power generation devices is constructed based on the wake velocity model, wherein the wake effect model is used to obtain the initial wind speed of the power generation devices;

[0178] Based on the Weibull probability distribution, a wind direction and wind speed model is constructed based on wind direction and wind speed parameters, wherein the wind direction and wind speed model is used to obtain the wind speed and wind direction characteristics of the power generation equipment;

[0179] A power model is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment;

[0180] The wind speed of the power generation equipment is determined based on the initial wind speed and the wind speed characteristics, the wind direction of the power generation equipment is determined based on the wind direction characteristics, and a state space is constructed based on the wind speed, the wind direction, and the location of the power generation equipment.

[0181] The state space is optimized using an Actor network and the power model, and an action space is constructed with the goal of maximizing power generation. The action space includes the optimized position of the power generation equipment.

[0182] The state space set and action space set are optimized by using a Critic network to determine the action space result with the goal of maximizing total power generation. The state space set and action space set are composed of the state space and action space of all the power generation equipment.

[0183] The present invention will now be described an electronic device 300 that can serve as a server or client of the present invention, which is an example of a hardware device that can be applied to various aspects of the present invention. Electronic device 300 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 300 can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0184] Electronic device 300 includes a computing unit that can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) or a computer program loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The computing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0185] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention according to actual needs. Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units.

[0186] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for optimizing microscopic electric field location, characterized in that, include: Construct a wake velocity model based on wake characteristic parameters; Based on the kinetic energy balance theory, a wake effect model between power generation devices is constructed based on the wake velocity model. The wake velocity of other power generation devices at the location of the target power generation device is obtained based on the wake velocity model. Based on the distance between the target power generation device and the other power generation devices along the wind direction and the velocity deficit ratio of the other power generation devices on the target power generation device, the wake effect model between the target power generation device and the other power generation devices is constructed. The wake effect model is used to obtain the initial wind speed of the power generation device. Based on the Weibull probability distribution, a wind direction and wind speed model is constructed based on wind direction and wind speed parameters, wherein the wind direction and wind speed model is used to obtain the wind speed and wind direction characteristics of the power generation equipment; A power model is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment; The wind speed of the power generation equipment is determined based on the initial wind speed and the wind speed characteristics, the wind direction of the power generation equipment is determined based on the wind direction characteristics, and a state space is constructed based on the wind speed, the wind direction, and the location of the power generation equipment. The state space is optimized using an Actor network and the power model, and an action space is constructed with the goal of maximizing power generation. The action space includes the optimized position of the power generation equipment. The state space set and action space set are optimized by using a Critic network to determine the action space result with the goal of maximizing total power generation. The state space set and action space set are composed of the state space and action space of all the power generation equipment.

2. The electric field micro-location optimization method according to claim 1, characterized in that, The step of constructing the wake velocity model based on wake characteristic parameters includes: The wake velocity model is constructed based on the free-flow wind speed, turbine thrust coefficient, turbine rotor diameter, and wake diffusion constant. The wake velocity model is used to obtain the wake velocity at a given downstream distance.

3. The electric field micro-location optimization method according to claim 1, characterized in that, The construction of the wind direction and wind speed model based on the Weibull probability distribution and wind direction and wind speed parameters includes: Construct a probability distribution function based on the Weibull probability distribution; The wind speed shape parameter is determined based on the range of wind speed variation, and the wind speed scale parameter is determined based on the average wind speed, wherein the wind speed parameter includes the wind speed shape parameter and the wind speed scale parameter; Substitute the wind speed shape parameter and the wind speed scale parameter into the probability distribution function to determine the wind speed density function; The wind speed is discretized to obtain the number of wind speed segments and the cut-in and cut-out wind speeds corresponding to each wind speed segment; Substituting the cut-in wind speed and cut-out wind speed, wind direction shape parameter and wind direction scale parameter of the wind direction segment into the probability distribution function, the wind direction density function of the wind direction segment is determined, wherein the wind direction parameter includes the wind direction shape parameter and the wind direction scale parameter; The wind speed density function and the wind direction density function are used as the wind direction and wind speed model.

4. The electric field micro-location optimization method according to claim 1, characterized in that, The process of obtaining the power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment includes: The power curve of the power generation equipment is obtained by linear interpolation; Based on the target wind direction and target wind speed of the target power generation equipment, the initial wind speed, the wind speed characteristics, and the wind direction characteristics are obtained according to the wake effect model and the wind direction and wind speed model. The power model of the power generation equipment is obtained based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve.

5. The electric field micro-location optimization method according to claim 1, characterized in that, The step of optimizing the state space using the Actor network and the power model to construct the action space with the goal of maximizing power generation includes: A minimum distance constraint is constructed based on the Euclidean distance between the two power generation devices and a preset minimum spacing. Establish boundary constraints based on the electric field space boundary.

6. The electric field micro-location optimization method according to claim 1, characterized in that, The process of optimizing the state space set and action space set using a Critic network to determine the action space result with the objective of maximizing total power generation includes: Construct a value function based on the state space set and the action space set, wherein the value function is used to evaluate the expected reward of performing an action in a given state; Using the maximum total power generation as a constraint on the expected return, the action space outcome is determined through the value function.

7. An electric field micro-location optimization device, characterized in that, include: A standalone wake velocity module is used to construct a wake velocity model based on wake characteristic parameters; The wake effect module is used to construct a wake effect model between power generation devices based on the wake velocity model according to the kinetic energy balance theory. It obtains the wake velocity of other power generation devices at the location of the target power generation device based on the wake velocity model. Based on the distance along the wind direction between the target power generation device and the other power generation devices, and the velocity deficit ratio of the other power generation devices on the target power generation device, it constructs the wake effect model between the target power generation device and the other power generation devices. The wake effect model is used to obtain the initial wind speed of the power generation device. The wind direction and speed module is used to construct a wind direction and speed model based on wind direction parameters and wind speed parameters according to the Weibull probability distribution, wherein the wind direction and speed model is used to obtain the wind speed characteristics and wind direction characteristics of the power generation equipment; A power module is used to obtain a power model based on the initial wind speed, the wind speed characteristics, the wind direction characteristics, and the power curve of the power generation equipment. A space construction module is used to determine the wind speed of the power generation equipment based on the initial wind speed and the wind speed characteristics, determine the wind direction of the power generation equipment based on the wind direction characteristics, and construct a state space based on the wind speed, wind direction, and position of the power generation equipment; The Actor network module is used to optimize the state space through the Actor network and the power model, and construct the action space with the goal of maximizing power generation, wherein the action space includes the optimized position of the power generation equipment; The Critic network module is used to optimize the state space set and action space set through the Critic network to determine the action space result with the goal of maximizing the total power generation. The state space set and the action space set are composed of the state space and the action space of all the power generation equipment.

8. An electronic device, characterized in that, Including memory and processor; The memory is used to store computer programs; The processor is configured to implement the electric field micro-location optimization method as described in any one of claims 1-6 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the electric field micro-location optimization method as described in any one of claims 1-6.

Images