An edge-computing-based wind farm energy management method and system

Abstract

The application discloses a kind of based on edge computing's wind farm energy management method and system, it is related to wind power technology field, including: real-time acquisition wind turbine data, the data collected are preprocessed;Based on the data after pre-processing, calculate power generation power prediction value;Based on power generation power prediction value, calculate power grid load demand prediction value;Based on power grid load demand prediction value, calculate power grid load matching deviation;Based on power grid load matching deviation, generate wind turbine power output adjustment scheme.The wind farm energy management method based on edge computing provided in the application improves the real-time performance of data processing through the distributed architecture of edge computing node, especially in the case of complex wind farm operating environment and large amount of data, it can quickly respond to dynamically changing load demand;Based on hierarchical modeling prediction technology, not only improves the accuracy of wind power prediction and power grid load demand prediction, but also enhances the adaptability of model to complex weather conditions and nonlinear relationship.

Description

Technical Field

[0001] This invention relates to the field of wind power technology, specifically to a wind farm energy management method and system based on edge computing. Background Technology

[0002] Currently, with the widespread application of wind power technology, the proportion of wind farms in the power system is gradually increasing. However, due to the significant intermittent and fluctuating characteristics of wind power generation, its grid-connected operation poses a severe challenge to the stability and reliability of the power grid. Traditional wind farm energy management systems mainly rely on a centralized cloud processing model. This model typically includes data acquisition, processing, and optimization processes. Due to the large-scale data transmission and centralized computing involved, the system's real-time and dynamic response capabilities are insufficient, making it difficult to meet the rapidly changing demands of the power grid.

[0003] Existing wind power forecasting methods are typically based on static mathematical models or simple time series analysis. These methods struggle to adequately account for the dynamic changes in meteorological conditions and the complexity of wind farm operating environments. Especially when wind speed and direction change frequently, the forecasting error increases significantly. Furthermore, traditional load demand forecasting methods fail to fully consider the actual impact of wind power output on grid load demand, resulting in insufficient forecast accuracy and hindering reliable support for optimization decisions.

[0004] In terms of power regulation and scheduling optimization, existing methods mostly focus on simplified linear optimization models, neglecting the response speed limitations and power output constraints of wind turbines during operation. The resulting scheduling schemes may be unimplementable in practical applications due to failure to meet equipment physical limitations. Furthermore, traditional optimization methods do not adequately consider the dynamic interaction between wind farms and the power grid, failing to provide sufficient adaptability when load fluctuations and wind power fluctuations are significant.

[0005] Therefore, it is evident that traditional wind farm energy management technologies have the following shortcomings:

[0006] The data acquisition and processing efficiency is low, making it difficult to quickly respond to dynamic changes in load demand.

[0007] Existing power forecasting and load demand forecasting models perform poorly under complex weather conditions.

[0008] The optimization results were not fully considered due to the limitations of wind turbine operation and the characteristics of grid demand, and therefore some of them are not applicable in actual operation.

[0009] In summary, to address the aforementioned technical issues, an innovative wind farm energy management method needs to be proposed. This method should improve data processing efficiency, enhance prediction accuracy, and optimize algorithm applicability to meet the requirements of efficient operation of modern wind farms and grid stability. Summary of the Invention

[0010] In view of the above-mentioned problems, the present invention is proposed.

[0011] Therefore, the technical problem solved by this invention is that existing wind farm energy management methods suffer from insufficient real-time performance, low prediction accuracy, and poor feasibility of optimization schemes. In particular, in scenarios where dynamic load demand changes rapidly, it is difficult to effectively cope with the matching problem between wind power fluctuations and grid load demand, as well as the optimization problem of adjusting wind turbine power output in complex operating environments.

[0012] To address the aforementioned technical problems, this invention provides the following technical solution: a wind farm energy management method based on edge computing, comprising:

[0013] Real-time data collection from wind turbine generators; and preprocessing of the collected data.

[0014] Calculate the predicted power generation value based on the preprocessed data;

[0015] Calculate the grid load demand forecast based on the power generation forecast;

[0016] Calculate the grid load matching deviation based on the grid load demand forecast;

[0017] A power output adjustment scheme for wind turbines is generated based on the grid load matching deviation.

[0018] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the wind turbine data includes power generation data, grid load data, and meteorological data.

[0019] The preprocessing of the collected data includes normalizing the power generation data, represented as follows:

[0020]

[0021] Among them, P raw P represents the original power generation data of the wind turbine. min P represents the minimum value in the dataset. max P represents the maximum value in the dataset. norm This represents the normalized power generation data of the wind turbine generator;

[0022] The fluctuation amplitude of the power grid load data is calculated and expressed as follows:

[0023]

[0024] Among them, P grid,i This represents the i-th data point of the power grid load data. ΔP represents the average load of the power grid, N represents the total number of power grid load data, and ΔPgrid This indicates the amplitude of power grid load fluctuations;

[0025] Meteorological data is decomposed into wind speed and direction to generate vectorized meteorological feature data, represented as follows:

[0026]

[0027] Where v represents wind speed and θ represents wind direction. A two-dimensional vector representing wind speed and direction.

[0028] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the predicted power generation value is expressed as follows:

[0029] P base =α1·v 3 +α2·v 2 +α3·v+α4

[0030] P weather =P base +β1·v·cos(θ)+β2·v·sin(θ)

[0031] P wind,pred =P weather +β3·P norm

[0032] Among them, P base The values ​​represent estimated power generation, α1, α2, α3, and α4 represent the fitting coefficients for the wind turbine power curve, β1 and β2 represent the correction coefficients for the wind speed component, and P represents the estimated power generation value. wind,pred β3 represents the final predicted power generation, and β3 represents the correction factor for historical power generation.

[0033] The coefficients for calculating the predicted power generation value are optimized by minimizing the mean square error, and are expressed as follows:

[0034]

[0035] The optimized model outputs P wind,pred This is the predicted value for power generation.

[0036] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the predicted grid load demand is expressed as follows:

[0037] P grid,base =γ1·ΔP grid +γ2·P grid,hist

[0038] P grid,corrected =P grid,base +γ3·Pwind,pred

[0039] P grid,pred =P grid,corrected +γ4·(ΔP grid ) 2 +γ5·(ΔP grid ·P wind,pred )

[0040] Where γ1 and γ2 represent the regression coefficients of fluctuation amplitude and historical load, P grid,hist γ represents historical grid load data, and γ3 represents the correction coefficient for the impact of wind power on grid load demand (ΔP). grid ) 2 The term representing the square of the load fluctuation amplitude is used to capture the nonlinear effects of load fluctuations, ΔP. grid ·P wind,pred γ4 and γ5 represent the interaction term between load fluctuation and wind power, respectively, and represent the regression coefficients of the nonlinear supplement.

[0041] The calculation coefficients for the power grid load demand forecast are optimized by minimizing the mean square error, and are expressed as follows:

[0042]

[0043] Optimized P grid,pred This represents the predicted value of the grid load demand.

[0044] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the grid load matching deviation is expressed as:

[0045] ΔP grid =P grid,actual -P grid,pred

[0046] Among them, P grid,actual Represented as actual grid load;

[0047] The proposed wind turbine power output adjustment scheme includes defining an optimization objective function and setting constraints.

[0048] The objective function is solved iteratively using the particle swarm optimization algorithm to generate adjustment schemes, and the optimization results are sent to the wind turbine.

[0049] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the objective function is expressed as follows:

[0050]

[0051] Where J represents the objective function, λ1 and λ2 represent the weight coefficients, and Pwind,adjusted,i P represents the adjusted power output of the i-th wind turbine. wind,current,i This represents the current actual output power of the i-th wind turbine.

[0052] As a preferred embodiment of the wind farm energy management method based on edge computing described in this invention, the constraints include power output limits, response speed limits, and total power constraints.

[0053] The constraint condition is expressed as follows:

[0054] P wind,min,i ≤P wind,adjusted,i ≤P wind,,max,i

[0055] |P wind，adjusted，i -P wind，current，i |≤R maxi ·T

[0056]

[0057] Among them, P wind,min,i、 P wind,max,i R represents the minimum and maximum power limits of the i-th wind turbine. max,i Let T represent the maximum adjustment rate of the i-th wind turbine unit, and T represent the adjustment time period.

[0058] Another objective of this invention is to provide a wind farm energy management system based on edge computing. By constructing a wind farm energy management system based on edge computing, this invention solves the problems of low data processing efficiency, insufficient power prediction accuracy, and lack of constraint adaptability in existing wind farm data management systems, effectively improving the real-time performance, accuracy, and applicability of wind farm energy management.

[0059] To address the aforementioned technical problems, this invention provides the following technical solution: a wind farm energy management system based on edge computing, comprising: a data acquisition module, an energy prediction module, a status monitoring module, an energy efficiency assessment module, and a power adjustment module; the data acquisition module is used to acquire wind turbine data in real time and preprocess the acquired data; the energy prediction module is used to calculate the predicted power generation value based on the preprocessed data; the status monitoring module is used to calculate the predicted grid load demand value based on the predicted power generation value; the energy efficiency assessment module is used to calculate the grid load matching deviation based on the predicted grid load demand value; and the power adjustment module is used to generate a wind turbine power output adjustment scheme based on the grid load matching deviation.

[0060] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the wind farm energy management method based on edge computing as described above.

[0061] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the wind farm energy management method based on edge computing as described above.

[0062] The beneficial effects of this invention are as follows: The wind farm energy management method based on edge computing provided by this invention significantly improves the real-time performance of data processing through the distributed architecture of edge computing nodes, especially in situations where wind farm operating environments are complex and data volumes are massive, enabling rapid response to dynamically changing load demands. The prediction technology based on hierarchical modeling not only improves the accuracy of wind power prediction and grid load demand prediction but also enhances the model's adaptability to complex meteorological conditions and nonlinear relationships. The optimization scheme achieves the optimal trade-off between load demand and wind turbine operating costs through a dynamic weight adjustment mechanism and a balanced objective function, while constraints ensure the feasibility of the optimization results. Overall, this invention provides efficient and flexible technical support for the intelligent operation of wind farms and the stability of the power grid, and is particularly suitable for scenarios with severe load fluctuations or complex wind farm operating conditions. Attached Figure Description

[0063] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is an overall flowchart of a wind farm energy management method based on edge computing, provided as an embodiment of the present invention.

[0065] Figure 2 This is an overall structural diagram of a wind farm energy management system based on edge computing, provided for the second embodiment of the present invention. Detailed Implementation

[0066] 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. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0067] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0068] Example 1

[0069] Reference Figure 1 As an embodiment of the present invention, a wind farm energy management method based on edge computing is provided, comprising:

[0070] Real-time data collection from wind turbine generators; and preprocessing of the collected data.

[0071] Calculate the predicted power generation value based on the preprocessed data;

[0072] Calculate the grid load demand forecast based on the power generation forecast;

[0073] Calculate the grid load matching deviation based on the grid load demand forecast;

[0074] A power output adjustment scheme for wind turbines is generated based on the grid load matching deviation.

[0075] The wind turbine data includes power generation data, grid load data, and meteorological data;

[0076] The preprocessing of the collected data includes normalizing the power generation data, represented as follows:

[0077]

[0078] Among them, P raw P represents the original power generation data of the wind turbine. min P represents the minimum value in the dataset. max P represents the maximum value in the dataset. norm This represents the normalized power generation data of the wind turbine generator;

[0079] The fluctuation amplitude of the power grid load data is calculated and expressed as follows:

[0080]

[0081] Among them, P grid,i This represents the i-th data point of the power grid load data. ΔP represents the average load of the power grid, N represents the total number of power grid load data, and ΔP grid This indicates the amplitude of power grid load fluctuations;

[0082] Meteorological data is decomposed into wind speed and direction to generate vectorized meteorological feature data, represented as follows:

[0083]

[0084] Where v represents wind speed in m / s, and θ represents wind direction in degrees (measured clockwise from due north). A two-dimensional vector representing wind speed and direction, containing the eastward (x-axis) and northward (y-axis) components of wind speed.

[0085] It should be noted that the preprocessing of the collected data is not only a formatting process for the raw data, but also improves the quality of subsequent model input data and the applicability of the model by extracting features and standardizing different types of data. Regarding the normalization method for power generation data, by standardizing the data range, the potential impact of inconsistent data distribution between different wind turbine units is eliminated, providing a unified input for subsequent model training and prediction.

[0086] Furthermore, the calculation of the fluctuation amplitude of power grid load data, by calculating the mean square error of the fluctuation data, can dynamically capture the characteristics of power grid load changes. Especially in scenarios with high-frequency load fluctuations, this calculation method can provide a clear characterization of the fluctuation amplitude, enhancing the adaptability of the load forecasting model to sudden demand.

[0087] Furthermore, vectorizing meteorological feature data decomposes wind speed and direction data into two-dimensional components, reducing the nonlinear relationships in the original data and making the feature dimensions of the model input more adaptable to wind power prediction scenarios. This processing method can significantly reduce the training complexity of the model, while exhibiting higher prediction accuracy in scenarios with rapidly changing wind speeds or multiple wind directions.

[0088] Specifically, the data preprocessing process is tailored to the sampling frequency and noise characteristics of different data sources, using edge computing nodes for real-time denoising and interpolation to ensure the real-time nature and continuity of the input data. This approach is particularly suitable for the distributed data acquisition needs of remote wind farms.

[0089] The predicted power generation value is expressed as follows:

[0090] P base =α1·v 3 +α2·v 2 +α3·v+α4

[0091] P weather =P base +β1·v·cos(θ)+β2·v·sin(θ)

[0092] P wind,pred =Pweather +β3·P norm

[0093] Among them, P base The values ​​represent estimated power generation, α1, α2, α3, and α4 represent the fitting coefficients for the wind turbine power curve, β1 and β2 represent the correction coefficients for the wind speed component, and P represents the estimated power generation value. wind,pred β3 represents the final predicted power generation, and β3 represents the correction factor for historical power generation.

[0094] The coefficients for calculating the predicted power generation value are optimized by minimizing the mean square error, and are expressed as follows:

[0095]

[0096] The optimized model outputs P wind,pred This is the predicted value for power generation.

[0097] It should be noted that the power generation prediction model uses the wind turbine power curve as its basis and captures the nonlinear relationship between wind speed and power generation through physical modeling, thereby realizing a direct response of wind speed changes to power output. The model's fitting parameters are trained using historical wind turbine operating data, enabling the model to maintain high accuracy when adapting to various wind turbine characteristics.

[0098] Furthermore, the meteorological feature correction component in the model dynamically adjusts the base power generation estimate by introducing wind speed and wind direction components. Under complex meteorological conditions (such as frequent wind direction changes or sudden increases in wind speed), it can more accurately reflect the power output change trend. This correction method solves the problem of insufficient sensitivity to wind direction changes in traditional power prediction methods by enhancing the model's adaptability to the dynamic characteristics of meteorological data.

[0099] Furthermore, by incorporating the correction part of historical power generation data and using normalized historical power generation as an input factor, the model can compensate for uncontrollable errors in meteorological data (such as sensor drift or sampling noise), thereby improving the stability and robustness of the overall prediction results.

[0100] Specifically, the training method optimizes the model coefficients by minimizing the mean squared error, and uses an iterative optimization strategy to dynamically adjust each coefficient, ensuring that the fitting error gradually decreases as the amount of training data increases. This training method is particularly suitable for scenarios involving multiple wind turbines working together to generate electricity, and can optimize the output differences between different units.

[0101] The predicted value of the power grid load demand is expressed as follows:

[0102] P grid,base =γ1·ΔP grid +γ2·P grid,hist

[0103] P grid,corrected =P grid,base +γ3·P wind,pred

[0104] P grid,pred =P grid,corrected +γ4·(ΔP grid ) 2 +γ5·(ΔP grid ·P wind,pred )

[0105] Where γ1 and γ2 represent the regression coefficients of fluctuation amplitude and historical load, P grid,hist γ represents historical grid load data, and γ3 represents the correction coefficient for the impact of wind power on grid load demand (ΔP). grid ) 2 The term representing the square of the load fluctuation amplitude is used to capture the nonlinear effects of load fluctuations, ΔP. grid ·P wind,pred γ4 and γ5 represent the interaction term between load fluctuation and wind power, respectively, and represent the regression coefficients of the nonlinear supplement.

[0106] The calculation coefficients for the power grid load demand forecast are optimized by minimizing the mean square error, and are expressed as follows:

[0107]

[0108] Optimized P grid,pred This represents the predicted value of the grid load demand.

[0109] It should be noted that the power grid load demand forecasting model consists of three parts: basic calculation, correction calculation, and nonlinear supplementation. The basic calculation part captures the main trends in power grid demand through regression analysis of load fluctuation amplitude and historical load data, providing highly correlated basic forecast values ​​for the model.

[0110] Furthermore, the correction calculation part combines the dynamic impact of wind power forecast values ​​on grid load demand. By incorporating wind power forecast values ​​into the load demand calculation formula, the model's sensitivity to actual wind power changes is effectively improved. Especially in grids with a high proportion of wind power integration, this correction term can significantly improve the matching degree between the forecast results and the actual values.

[0111] Furthermore, the nonlinear supplementary part, through the introduction of squared and cross terms, can dynamically capture the complex nonlinear relationships in grid load demand. These nonlinear characteristics not only originate from the load fluctuation amplitude itself, but also interact with the fluctuation characteristics of wind power output. The supplementary terms in the model achieve an accurate fit to these complex relationships.

[0112] Specifically, the training method for optimizing the parameters of the load prediction model through mean squared error adopts a multi-level stepwise optimization strategy. After the initial regression calculation, the weights of the supplementary terms are further adjusted through cross-validation to ensure the robustness of the prediction model under high-fluctuation load scenarios.

[0113] The power grid load matching deviation is expressed as:

[0114] ΔP grid =P grid,actual -P grid,pred

[0115] Among them, P grid,actual Represented as actual grid load;

[0116] The proposed wind turbine power output adjustment scheme includes defining an optimization objective function and setting constraints.

[0117] The objective function is solved iteratively using the particle swarm optimization algorithm to generate adjustment schemes, and the optimization results are sent to the wind turbine.

[0118] It should be noted that the calculation of the grid load matching deviation directly reflects the difference between the actual grid load and the predicted load, providing a clear optimization direction and quantitative standard for the subsequent optimization objective function.

[0119] Furthermore, the matching deviation calculation results can dynamically update the parameters of the power grid load demand prediction model in real-time feedback, allowing the load matching deviation to continuously converge to an optimal range, thereby achieving continuous model optimization. This dynamic update method is particularly suitable for scenarios where load demand changes rapidly.

[0120] Furthermore, in the process of generating wind turbine power output adjustment schemes, the grid load matching deviation is introduced into the objective function to ensure that the optimization scheme takes minimizing the load deviation as the core objective, while also taking into account the operating limitations of the wind turbines, thereby improving the feasibility of the adjustment scheme.

[0121] Specifically, the calculation results of the matching deviation can also be used to evaluate the overall operating efficiency of the wind farm. By comparing with historical matching deviation data, key indicators of the wind farm's operating status can be provided, offering guidance for subsequent scheduling optimization.

[0122] The objective function is expressed as follows:

[0123]

[0124] Where J represents the optimization objective function, λ1 and λ2 represent weighting coefficients used to balance load matching and adjust costs; P wind,adjusted,i P represents the adjusted power output of the i-th wind turbine. wind,current,iThis represents the current actual output power of the i-th wind turbine.

[0125] It should be noted that the optimization objective function dynamically adjusts the execution direction of the optimization scheme by balancing the weight relationship between the grid load matching deviation and the wind turbine adjustment cost, thereby achieving the optimal decision under different operating scenarios.

[0126] Furthermore, the matching deviation term in the objective function is dynamically adjusted via a weighting coefficient λ1 to prioritize the minimization of deviation, making it suitable for scenarios with drastic fluctuations in grid load. For example, when grid load demand increases significantly, this term can quickly drive wind turbines to adjust their output power to meet the demand.

[0127] Furthermore, the design of the adjustment cost term incorporates the squared difference between the actual output of the wind turbine and the adjusted power, reflecting the energy consumption and operating costs of the adjustment scheme. By dynamically adjusting the weighting coefficient λ2, a balance can be found between stabilizing grid operation and reducing dispatch costs.

[0128] Specifically, the overall design of the objective function, while ensuring that grid demands are met, improves the economy and flexibility of wind turbines through multi-objective optimization, thus supporting the coordinated operation of the grid and wind farms. The constraints include power output limits, response speed limits, and total power constraints.

[0129] The constraints include:

[0130] Power output limit:

[0131] P wind,min,i ≤P wind,adjusted,i ≤P wind,max,i

[0132] Among them, P wind,min,i、 P wind,max,i This represents the minimum and maximum power limits of the i-th wind turbine.

[0133] Total power constraint:

[0134]

[0135] Ensure that the total power of the wind turbines after adjustment meets the grid load requirements.

[0136] Response speed limit:

[0137] |P wind,adjusted,i -P wind,current,i |≤R max,i ·T

[0138] Among them, R max,i Let T represent the maximum adjustment rate of the i-th wind turbine unit, and T represent the adjustment time period.

[0139] It should be noted that the design of the constraints combines power output limits, response speed limits, and total power constraints to ensure the technical feasibility and safety of the adjustment scheme, while meeting the actual operation requirements of wind farms.

[0140] Furthermore, power output limiting restricts the adjustment range of each wind turbine to prevent exceeding the equipment's operating capacity, especially under extreme conditions of high or low wind speeds, thus protecting the physical safety of the wind turbine's equipment.

[0141] Furthermore, the response speed limit combines the device's maximum response rate and adjustment time cycle to ensure that the adjustment scheme can respond promptly to rapidly changing load demands, while preventing power fluctuations caused by excessively rapid adjustments.

[0142] Specifically, the total power constraint ensures the overall stability of the power grid operation by dynamically balancing the relationship between the total power output of wind turbines and the grid demand, while improving the power output efficiency of wind farms.

[0143] It should also be noted that the particle swarm optimization algorithm uses particle encoding, where each particle represents a wind turbine power adjustment scheme, denoted as,

[0144] P wind,adjusted =[P wind,adjusted,1 ,P wind,adjusted,2 ,…,P wind,adjusted,N ]

[0145] The dimension of each particle is the number of wind turbine units, N.

[0146] Particle initialization: Randomly initialize the position of each particle (power output scheme), represented as follows:

[0147] P wind,adjusted,i ∈[P wind,min,i ,P wind,max,i ]

[0148] The velocity of each particle is initialized, expressed as:

[0149] v i,t =rand·(P wind,max,i -P wind,min,i )

[0150] Ensure that the initial particles satisfy the power constraints and the total power constraints.

[0151] Velocity update (in conjunction with constraints): When updating the particle velocity, corrections are made based on the total power constraint, expressed as follows:

[0152] v i,t+1 =w·v i,t+c1·rand1·(p best,i -x i,t )+c2·rand2·(g best -x i,t )

[0153] Limiting the updated speed is represented as follows:

[0154] v i,t+1 =min(max(v) i,t+1 ,-R max,i ·T),R max,i ·T)

[0155] Position update (combined with power constraints): Update the particle's position (power output scheme), represented as follows:

[0156] x i,t+1 =x i,t +v i,t+1

[0157] The updated position is trimmed to ensure that the power output limit is met, as shown below.

[0158] P wind,min,i ≤x i,t+1 ≤P wind,max,i

[0159] The fitness function is used to calculate the fitness of the current particle, and is expressed as follows:

[0160]

[0161] For particles that do not meet the total power constraint, an additional penalty term is added, denoted as,

[0162]

[0163] Update the best position, update the historical best position p for each particle. best,i Update the global optimal position g. best .

[0164] Termination condition: When the objective function J converges or reaches the maximum number of iterations, the iteration stops, and the global optimal position g is output. best .

[0165] Optimal solution g best This refers to the wind turbine power output adjustment scheme, expressed as follows:

[0166] P wind,adjusted =[P wind,adjusted,1 ,P wind,adjusted,2 ,…,P wind,adjusted,N ]

[0167] The optimization scheme is implemented by sending adjustment commands to each wind turbine through edge computing nodes.

[0168] Monitor the adjusted output power P wind,actual and load data P grid,actual Optimize repeatedly based on the new deviation.

[0169] Specifically, the application of the particle swarm optimization algorithm in this invention has been customized to meet the needs of wind turbine power adjustment. First, in the particle initialization phase, each particle is assigned a position representing the wind turbine's power adjustment scheme. During initialization, the particle's position range is determined by the minimum and maximum power output range of each wind turbine, while also satisfying the matching condition between the total power output of the wind farm and the grid load demand. If the initial position of a particle does not meet the total power constraint, the particle's position ratio is adjusted through normalization to ensure overall power balance. Next, during the velocity update process, the particle's velocity is dynamically adjusted not only based on the gradient change of the objective function but also in conjunction with the current grid load deviation. If the load deviation is large, particles are preferentially guided to move towards the direction with the larger power adjustment gap to accelerate the optimization process. Furthermore, during iteration, the historical optimal position and the global optimal position of the particles are updated periodically, and a dynamic convergence strategy is used to determine whether the optimization termination condition has been met, ensuring efficient convergence of the optimization process.

[0170] In the processes of power generation forecasting and grid load demand forecasting, this invention employs dynamic parameter optimization and weight adjustment mechanisms to ensure the model's adaptability and prediction accuracy. For the power generation forecasting model, its parameters are continuously updated through training on historical data, enabling the model to adapt to the characteristics of different wind turbines and meteorological conditions. The training process uses a stepwise optimization strategy, adjusting the parameter weights based on the prediction error in each iteration, thereby continuously narrowing the gap between actual power generation and the predicted value. In grid load demand forecasting, the weight coefficients of the nonlinear supplementary component are dynamically adjusted according to the real-time characteristics of load fluctuations. For example, when load fluctuations are significant, the weight of the correlation coefficient with the fluctuations is increased, allowing the model to more accurately capture the impact of fluctuations on demand. Simultaneously, the weights of the load correction component are adaptively updated with changes in wind power output, ensuring that the prediction results reflect real-time changes in the actual demand of the grid.

[0171] This invention leverages a distributed architecture of edge computing nodes to achieve efficient processing and real-time optimization of wind farm data. Edge nodes employ multi-threading technology to process different types of data in parallel, such as power generation data, grid load data, and meteorological data. Each type of data is collected, preprocessed, and stored by an independent thread to improve system response speed. Furthermore, edge nodes utilize local storage and caching technologies to maintain basic wind farm optimization functions even in the event of network outages or communication delays. For large-scale wind farms, multiple nodes collaborate through a distributed computing framework. Each node independently executes data collection tasks, while the master node integrates optimization results and distributes them uniformly to the wind turbines. Additionally, edge nodes employ a real-time feedback mechanism to compare actual operating conditions with model predictions, dynamically adjusting model parameters and optimization objectives to cope with complex meteorological conditions and load demand changes. This architecture effectively improves the stability and real-time response capabilities of wind farm operation.

[0172] The constraints set in this invention not only limit the search space of the optimization algorithm but also ensure that the generated power adjustment scheme can be implemented in actual operation. The power output constraint defines the adjustment range based on the wind turbine's operating capacity, preventing the output power from exceeding the equipment's physical limitations. The response speed constraint is based on the wind turbine's maximum adjustment rate, ensuring that the adjustment process does not cause mechanical damage to the equipment while meeting the time requirements of changes in grid demand. The total power constraint dynamically balances the total output of the wind farm with the actual demand of the grid, avoiding insufficient or overloaded overall power output of the wind farm. Furthermore, these constraints take effect in real time in each iteration of the particle swarm optimization algorithm. Specifically, after the particle position and velocity are updated, the algorithm checks whether any particles violate the constraints. If a violation occurs, the algorithm adjusts the particle's position and velocity by pruning or introducing a penalty mechanism, forcing it back to the feasible solution space. This dynamic constraint mechanism not only improves the practicality of the optimization scheme but also significantly reduces the generation of infeasible solutions.

[0173] Example 2

[0174] Reference Figure 2 As one embodiment of the present invention, a wind farm energy management system based on edge computing is provided, comprising:

[0175] The system includes a data acquisition module 100, an energy prediction module 200, a condition monitoring module 300, an energy efficiency assessment module 400, and a power adjustment module 500.

[0176] The data acquisition module 100 is used to acquire wind turbine data in real time and preprocess the acquired data;

[0177] The energy prediction module 200 is used to calculate the predicted power generation value based on the preprocessed data;

[0178] The status monitoring module 300 is used to calculate the grid load demand forecast based on the power generation forecast value;

[0179] The energy efficiency assessment module 400 is used to calculate the grid load matching deviation based on the grid load demand forecast value;

[0180] The power adjustment module 500 is used to generate a power output adjustment scheme for wind turbines based on the grid load matching deviation.

[0181] Example 3

[0182] One embodiment of the present invention differs from the previous two embodiments in that:

[0183] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0184] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0185] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0186] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0187] Example 4

[0188] As one embodiment of the present invention, a wind farm energy management method based on edge computing is provided. To verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0189] This experiment, based on the actual operating environment of a wind farm, simulates the performance of different energy management methods in a typical wind farm. The experimental scenario is a wind farm with 24 wind turbines, characterized by complex terrain and variable weather conditions. Data is collected once per second. Collected data includes real-time wind turbine power generation, grid load demand and its fluctuation characteristics, and meteorological data (wind speed and direction). The experiment lasts 60 minutes, primarily targeting scenarios with large wind speed fluctuations and frequent changes in grid load demand, to verify the real-time performance and dynamic response capabilities of the proposed method. All data acquisition devices are actually deployed sensors. Data from each wind turbine is collected via edge computing nodes and stored in a local storage system, while grid load and meteorological data are used as global references. The traditional method compared in this experiment is a centralized cloud management model, with a centralized server at the core for data processing and scheduling scheme generation.

[0190] The experimental process of this invention involves an edge computing node at each wind turbine in a wind farm, responsible for real-time collection of local power generation, grid load data, and meteorological data. After collection, the edge node immediately normalizes the power generation data to remove dimensional differences and calculates the fluctuation amplitude of the grid load data to extract load fluctuation characteristics. Simultaneously, wind speed and direction data are vectorized into two-dimensional meteorological features to reduce the impact of nonlinear inputs on the prediction model. After data preprocessing, power generation and load demand predictions are completed using the local computing resources of the edge nodes. The prediction model uses historical power generation data, vectorized meteorological features, and load fluctuation characteristics for real-time iterative optimization, outputting predicted power generation and load demand values. Based on the prediction results and real-time grid load deviations, the edge nodes generate power adjustment schemes using a customized particle swarm optimization algorithm, considering power output limits, response rates, and total power constraints. The final optimized scheme is directly sent from the edge nodes to the wind turbines for execution, adjusting the wind turbines' power output. The entire process is completed within the edge nodes, without transmitting data to the cloud, thus ensuring real-time processing and rapid response capabilities of the adjustment scheme.

[0191] Traditional methods employ a centralized management model. Real-time power generation data collected by each wind turbine is transmitted to a cloud server via the network. Grid load demand data and meteorological data are also uniformly transmitted to the cloud for centralized processing. After all data is normalized and basic feature extracted in the cloud, it is fed into a static power generation prediction model for forecasting. Power prediction is based on simple regression analysis of historical power generation and wind speed data, with the predicted values ​​serving as input for load demand prediction. Load demand prediction uses a linear regression model to analyze the historical variation characteristics of grid load, ignoring the dynamic impact of real-time wind power changes on load demand. After forecasting, the cloud server uses a linear optimization algorithm to generate a power adjustment scheme. However, because the algorithm does not fully consider the physical operating limitations of the wind turbines, the generated scheme requires manual review and adjustment to meet actual operating requirements. The final adjustment scheme is then distributed from the cloud to the wind turbines for execution. The centralization of data transmission and cloud computing leads to long processing and response times, especially in scenarios with frequent load fluctuations, significantly limiting the applicability of the optimization scheme. Experimental results are shown in Table 1.

[0192] Table 1 Comparison of Experimental Data

[0193] index Traditional methods Method of the present invention Data preprocessing time (seconds) 180 25 Mean Square Error (MSE) of Power Generation Forecast <![CDATA[6.5MW 2 ]]> <![CDATA[2.1MW 2 ]]> Mean Squared Error (MSE) of Load Demand Forecast <![CDATA[3.5MW 2 ]]> <![CDATA[0.99MW 2 ]]> Adjust the scheme generation time (seconds). 45 8 Adaptability (deviation) to load fluctuations ±6MW ±0.5MW

[0194] This invention significantly reduces data transmission latency during the data preprocessing stage through the distributed deployment of edge computing nodes. Traditional methods rely on centralized cloud computing, requiring data to be transmitted to the cloud for processing, resulting in lengthy preprocessing times. In contrast, this invention directly performs data standardization, feature extraction, and vectorization at the edge nodes, requiring only global aggregation of key indicators, thus greatly improving processing efficiency.

[0195] In the power generation prediction stage, the method of this invention employs a dynamic nonlinear model, using vectorized meteorological characteristics and historical power generation data as key input variables. Combined with a real-time optimization parameter adjustment mechanism, it significantly reduces prediction errors. Traditional static mathematical models are insufficiently adaptable to complex meteorological conditions and fail to effectively capture the dynamic impact of wind speed fluctuations on power generation. The nonlinear modeling of this invention solves this problem through a dynamic correction strategy.

[0196] In load demand forecasting, this invention significantly improves the responsiveness of forecast results to dynamic changes in grid demand by extracting load fluctuation characteristics, correcting for wind power fluctuations, and introducing nonlinear supplementary terms. Traditional methods rely solely on linear regression models for historical trend analysis, failing to reflect the impact of real-time wind power fluctuations on load demand, leading to forecast results that deviate from actual demand. The method of this invention is more suitable for complex load fluctuation scenarios, achieving a forecast accuracy far superior to traditional methods.

[0197] In generating power adjustment schemes, this invention employs a customized particle swarm optimization algorithm, combining power output constraints and response speed constraints to optimize wind turbine power adjustment schemes in real time. Compared to traditional linear optimization algorithms, the method of this invention is more adaptable to changes in dynamic constraints, and the generated adjustment schemes exhibit stronger adaptability and stability to grid load fluctuations, while significantly reducing optimization time. This improvement benefits from the parallel search characteristics of the particle swarm optimization algorithm and the real-time processing capabilities of edge nodes.

[0198] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A wind farm energy management method based on edge computing, characterized in that, include: Real-time data collection from wind turbine generators; and preprocessing of the collected data. Calculate the predicted power generation value based on the preprocessed data; Calculate the grid load demand forecast based on the power generation forecast; Calculate the grid load matching deviation based on the grid load demand forecast; A power output adjustment scheme for wind turbines is generated based on grid load matching deviation. The wind turbine data includes power generation data, grid load data, and meteorological data; The preprocessing of the collected data includes normalizing the power generation data, represented as follows: ， in, This represents the original power generation data of the wind turbine. This represents the minimum value in the dataset. This represents the maximum value in the dataset. This represents the normalized power generation data of the wind turbine generator; The fluctuation amplitude of the power grid load data is calculated and expressed as follows: ， ， in, The first data representing the grid load data Data points, This represents the average load on the power grid. This indicates the total number of grid load data. This indicates the amplitude of power grid load fluctuations; Meteorological data is decomposed into wind speed and direction to generate vectorized meteorological feature data, represented as follows: ， in, Indicates wind speed. Indicates wind direction. A two-dimensional vector representing wind speed and direction; The predicted power generation value is expressed as follows: ， ， ， in, This represents the estimated power generation value. , , , This represents the fitting coefficient of the wind turbine power curve. , This represents the correction factor for the wind speed component. This represents the final predicted power generation capacity. A correction factor representing historical power generation; The coefficients for calculating the predicted power generation value are optimized by minimizing the mean square error, and are expressed as follows: ， Optimized model output This is the predicted value for power generation.

2. The wind farm energy management method based on edge computing as described in claim 1, characterized in that: The predicted value of the power grid load demand is expressed as follows: ， ， ， in, , The regression coefficients represent the fluctuation amplitude and historical load. This represents historical power grid load data. This represents the correction factor for the impact of wind power on grid load demand. The term representing the square of the load fluctuation amplitude is used to capture the nonlinear effects of load fluctuations. This represents the interaction term between load fluctuations and wind power output. , Represents the regression coefficients for nonlinear supplementation; The calculation coefficients for the power grid load demand forecast are optimized by minimizing the mean square error, and are expressed as follows: ， Optimized This represents the predicted value of the grid load demand.

3. The wind farm energy management method based on edge computing as described in claim 2, characterized in that: The power grid load matching deviation is expressed as: ， in, Represented as actual grid load; The proposed wind turbine power output adjustment scheme includes defining an optimization objective function and setting constraints. The objective function is solved iteratively using the particle swarm optimization algorithm to generate adjustment schemes, and the optimization results are sent to the wind turbine.

4. The wind farm energy management method based on edge computing as described in claim 3, characterized in that: The objective function is expressed as follows: ， in, This represents the objective function to be optimized. , Indicates the weighting coefficient. Indicates the first Adjusted power output of the typhoon generator set Indicates the first The actual output power of the typhoon generator is currently... This indicates the total number of wind turbines in the wind farm. This indicates the wind turbine unit number.

5. The wind farm energy management method based on edge computing as described in claim 4, characterized in that: The constraints include power output limits, response speed limits, and total power constraints. The constraint condition is expressed as follows: ， ， ， in, , Indicates the first Minimum and maximum power limits for typhoon generators Indicates the first The maximum adjustment rate of the typhoon generator unit. This indicates an adjustment to the time period.

6. A system employing the wind farm energy management method based on edge computing as described in any one of claims 1 to 5, characterized in that, include: The system includes a data acquisition module (100), an energy prediction module (200), a condition monitoring module (300), an energy efficiency assessment module (400), and a power adjustment module (500). The data acquisition module (100) is used to acquire wind turbine data in real time and preprocess the acquired data; The energy prediction module (200) is used to calculate the predicted power generation value based on the preprocessed data; The status monitoring module (300) is used to calculate the grid load demand forecast based on the power generation forecast; The energy efficiency assessment module (400) is used to calculate the grid load matching deviation based on the grid load demand forecast; The power adjustment module (500) is used to generate a power output adjustment scheme for wind turbines based on the grid load matching deviation.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the wind farm energy management method based on edge computing as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the wind farm energy management method based on edge computing as described in any one of claims 1 to 5.

Images