Energy consumption characteristic net rack planning method considering emerging industry

Abstract

The invention discloses a net rack planning method considering energy consumption characteristics of an emerging industry, and relates to the technical field of net rack planning, and the method comprises the steps: recognizing the energy consumption characteristics of the emerging industry, and constructing an energy consumption fluctuation prediction model; solving the energy consumption fluctuation prediction model through a load prediction algorithm, and dividing a power grid region; configuring an energy storage system based on the power grid region division result; and accessing renewable energy sources based on the configuration result of the energy storage system. According to the grid planning method considering the energy consumption characteristics of the emerging industry, the fluctuation prediction model considering the energy consumption characteristics of the emerging industry is constructed, the rule of power grid load fluctuation can be accurately captured, the defect that a traditional method cannot adapt to dynamic changes is overcome, and prediction accuracy and reliability are improved. The clustering algorithm is adopted to optimize power grid region division, the difference between energy consumption fluctuation and load distribution can be fully considered, it is ensured that power grid region division is more scientific and reasonable, and therefore the adjusting capacity and stability of a power grid are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of grid planning, and in particular to a grid planning method taking into account the energy consumption characteristics of emerging industries. Background Art

[0002] In the existing power grid planning, due to the rapid development of emerging industries, the load fluctuation of the power grid has become more and more complicated. Traditional power grid planning methods usually rely on historical load data and traditional forecasting models. Although they can reflect the trend of power grid load changes to a certain extent, they fail to fully consider the characteristics of emerging industries and the energy fluctuations they bring. These methods have certain limitations in the accuracy of handling load fluctuations and the ability to regulate the power grid. The power grid regional division method in the existing technology relies too much on static data, such as geographic information and load distribution, and ignores the dynamic characteristics of energy fluctuations, resulting in the inability of regional division to accurately reflect the actual load fluctuations.

[0003] In addition, current energy storage system configuration methods often lack precise optimization for fluctuations in energy demand in specific grid areas, resulting in unreasonable configuration of energy storage resources and inability to effectively cope with the impact of peak load fluctuations and fluctuations in renewable energy. The optimization methods for the access ratio of renewable energy are usually oversimplified and fail to consider the complex interaction between energy storage systems and load fluctuations, resulting in failure to maximize the efficiency of renewable energy use.

[0004] Therefore, existing technologies have certain defects in dealing with energy consumption fluctuations in emerging industries and grid load regulation, especially in the comprehensive consideration of grid regional division, energy storage system configuration and renewable energy access optimization. It is impossible to fully utilize the energy consumption fluctuation characteristics of emerging industries to improve the regulation capacity and stability of the grid. Summary of the invention

[0005] The present invention is proposed to solve the above-mentioned problems.

[0006] Therefore, the technical problem solved by the present invention is that the existing power grid planning method cannot effectively deal with the energy consumption fluctuation problems brought about by emerging industries, the unreasonable configuration of energy storage systems and the low efficiency of renewable energy access.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a grid planning method considering the energy consumption characteristics of emerging industries, comprising: Identify the energy consumption characteristics of emerging industries and build energy consumption fluctuation prediction models; Solve the energy fluctuation prediction model through load prediction algorithm and divide the power grid area; Based on the grid area division result, configure an energy storage system; Based on the configuration result of the energy storage system, renewable energy is connected.

[0008] As a preferred solution of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, wherein: the construction of the energy consumption fluctuation prediction model includes determining the category of new energy industries; Define energy consumption characteristic parameters for each industry category at different life cycle stages; Collecting energy consumption data of the new energy industry based on the energy consumption characteristic parameters; The Z-score standardization method was used to preprocess the collected energy consumption data; The preprocessed energy consumption data is used to build an energy consumption fluctuation prediction model.

[0009] As a preferred scheme of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, wherein: the construction of the energy consumption fluctuation prediction model includes using the time series decomposition method to decompose the energy consumption data into three parts: trend, seasonality and random fluctuation, and applying the VARMA multivariate autoregressive moving average model to establish a dynamic energy consumption fluctuation prediction model.

[0010] As a preferred solution of the grid planning method considering the energy characteristics of emerging industries described in the present invention, the time series decomposition method is expressed as:

[0011] in, Represents the original energy consumption data, at time The actual value of the moment, Represents the trend component, which is the medium- and long-term trend of data. represents the seasonal component, which is the periodic fluctuation in the data that repeats over time. Represents the random fluctuation component, i.e., the residual part, which is the random fluctuation in the data that cannot be explained by the trend and seasonal patterns; The VARMA multivariate autoregressive moving average model is expressed as,

[0012] in, Indicates the current time The random fluctuation component of represents the autoregressive coefficient, represents the moving average coefficient, represents the error term, represents the autoregressive order, Represents the moving average order.

[0013] As a preferred solution of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, the energy consumption fluctuation prediction model is constructed, and the maximum likelihood estimation method is used to optimize the parameters. The optimal autoregressive coefficient and moving average coefficient are determined by calculating the log-likelihood function and maximizing the function, which is expressed as:

[0014] in, represents the log-likelihood function, represents the variance of the noise term of the model, Represents the random fluctuation component obtained from the time series decomposition, at the time point The value at Indicates the total length of the time series, indicating the total number of samples of the data; Use the quasi-Newton method to optimize the parameters in the maximum likelihood estimation to minimize the log-likelihood function, initialize the parameters, calculate the gradient and Hessian matrix, and use the update rule of the BFGS algorithm to adjust the parameters; The AIC information criterion is used to select the optimal autoregressive order and moving average order, expressed as,

[0015] in, represents the maximum value of the log-likelihood function, represents the number of parameters in the model, Contains parameters for the autoregressive coefficients, moving average coefficients, and the variance of the noise term.

[0016] As a preferred solution of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, wherein: the division of the power grid area includes calculating the energy demand fluctuation degree of different power grid areas based on the energy consumption data generated by the energy consumption fluctuation prediction model; According to the difference in energy demand fluctuation, combined with the geographical information and load distribution of the power grid, the K-means clustering algorithm is used to determine the division result of the power grid area by minimizing the variance of the load fluctuation in each area.

[0017] As a preferred solution of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, wherein: the configuration of the energy storage system includes optimizing the capacity and distribution of the energy storage system using a particle swarm optimization algorithm according to the energy demand fluctuation degree of each power grid area and the results of the energy fluctuation prediction model; The energy storage system configuration for each grid area is optimized by minimizing the difference between energy storage demand and grid load fluctuations.

[0018] As a preferred solution of the grid planning method considering the energy consumption characteristics of emerging industries described in the present invention, wherein: the access to renewable energy includes optimizing the timing and proportion of renewable energy access using a genetic algorithm based on the configuration parameters of the energy storage system and the results of the energy consumption fluctuation prediction model; Optimize the efficiency of renewable energy use by minimizing grid load fluctuations and maximizing the proportion of renewable energy access.

[0019] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the grid planning method considering the energy consumption characteristics of emerging industries as described above are implemented.

[0020] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the grid planning method considering the energy consumption characteristics of emerging industries as described above.

[0021] Beneficial effects of the present invention: The grid planning method considering the energy consumption characteristics of emerging industries provided by the present invention constructs a fluctuation prediction model considering the energy consumption characteristics of emerging industries. The present invention can accurately capture the law of grid load fluctuations, overcome the defect of traditional methods that cannot adapt to dynamic changes, and improve the accuracy and reliability of prediction. The K-means clustering algorithm is used to optimize the regional division of the power grid, which can fully consider the differences in energy consumption fluctuations and load distribution, ensure that the regional division of the power grid is more scientific and reasonable, and thus improve the regulation ability and stability of the power grid. The energy storage system is optimized and configured through the particle swarm optimization algorithm, which can achieve the best match between grid load fluctuations and energy storage needs, improve the utilization efficiency of the energy storage system, and reduce the waste of energy storage resources. The genetic algorithm is used to optimize the access timing and proportion of renewable energy, which can maximize the access proportion of renewable energy while ensuring that the grid load fluctuations are minimized, thereby improving the utilization efficiency of renewable energy and the proportion of green energy in the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0023] Figure 1 An overall flow chart of a grid planning method considering the energy consumption characteristics of emerging industries provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0025] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. Example

[0026] Reference Figure 1 , which is an embodiment of the present invention, provides a grid planning method considering the energy consumption characteristics of emerging industries, including: Identify the energy consumption characteristics of emerging industries and build an energy consumption fluctuation prediction model.

[0027] The energy consumption fluctuation prediction model is solved through the load forecasting algorithm and the power grid area is divided.

[0028] Based on the grid area division result, an energy storage system is configured.

[0029] Based on the configuration result of the energy storage system, renewable energy is connected.

[0030] Identify key new energy industry categories including solar power generation, wind power generation, electric vehicle manufacturing, energy storage systems, etc.

[0031] Specific energy consumption characteristic parameters are defined for each industry category at different life cycle stages (start-up, expansion, and maturity), including electricity load patterns, fluctuation amplitudes, and time period distribution, to guide subsequent data collection and model building.

[0032] By installing smart meters and data collection equipment, the energy consumption data of the new energy industries are collected in real time, including load demand, energy consumption fluctuations and peak electricity consumption periods, to ensure that the collected data is consistent with the energy consumption characteristic parameters.

[0033] The Z-score standardization method was used to clean and standardize the collected data to remove outliers and noise and ensure data quality and consistency.

[0034] The time series decomposition method is used to decompose the energy consumption data into three parts: trend, seasonality and random fluctuation. The multivariate autoregressive moving average model (VARMA) is used to establish a dynamic energy consumption fluctuation prediction model to capture the changing trend and fluctuation pattern of energy demand.

[0035] The energy consumption fluctuation prediction model is constructed by using a time series decomposition method to decompose the energy consumption data into three parts: trend, seasonality and random fluctuation, and a VARMA multivariate autoregressive moving average model is applied to establish a dynamic energy consumption fluctuation prediction model.

[0036] The time series decomposition method is expressed as,

[0037] in, Represents the original energy consumption data, at time The actual value of the moment, Represents the trend component, which is the medium- and long-term trend of data. Represents seasonal components, which are periodic fluctuations in the data that occur repeatedly over time (such as annual cycles, weekly cycles, etc.). Represents the random fluctuation component, i.e., the residual part, which is the random fluctuation in the data that cannot be explained by the trend and seasonal patterns; The VARMA multivariate autoregressive moving average model is expressed as,

[0038] in, Indicates the current time The random fluctuation component of represents the autoregressive coefficient, represents the moving average coefficient, represents the error term, represents the autoregressive order, Represents the moving average order.

[0039] The energy consumption fluctuation prediction model is constructed by using the maximum likelihood estimation method to optimize parameters, and the optimal autoregressive coefficient and moving average coefficient are determined by calculating the log-likelihood function and maximizing the function, which is expressed as:

[0040] in, represents the log-likelihood function, represents the variance of the noise term of the model, Represents the random fluctuation component obtained from the time series decomposition, at the time point The value at Indicates the total length of the time series, indicating the total number of samples of the data; Use the quasi-Newton method to optimize the parameters in the maximum likelihood estimation to minimize the log-likelihood function, initialize the parameters, calculate the gradient and Hessian matrix, and use the update rule of the BFGS algorithm to adjust the parameters; The AIC information criterion is used to select the optimal autoregressive order and moving average order, expressed as,

[0041] in, represents the maximum value of the log-likelihood function, represents the number of parameters in the model, , including parameters of the autoregressive coefficient, moving average coefficient, and noise term variance.

[0042] The grid planning method proposed in this embodiment takes into account the energy consumption characteristics of emerging industries, especially for the characteristics of the new energy industry, and adopts the Akaike Information Criterion (AIC) and the Quasi-Newton Method (BFGS Algorithm) to optimize the VARMA (Vector Autoregressive Moving Average) model to achieve accurate prediction of power demand fluctuations in the new energy industry and optimization of grid planning.

[0043] The energy consumption data of the new energy industry has significant seasonality, long-term growth trends and irregular volatility. Traditional linear regression or time series models often cannot accurately capture these complex characteristics. Therefore, the present invention decomposes the energy consumption data into three parts: trend, seasonality and random fluctuations through the time series decomposition method, so as to process different components respectively. The trend component after time series decomposition reflects the energy consumption growth pattern of the new energy industry on a longer time scale, the seasonal component reveals the energy consumption fluctuations that change with the seasons, and the random fluctuation component reflects unpredictable sudden factors, such as climate change or equipment failure.

[0044] The introduction of the VARMA model is to capture historical dependence and random fluctuations while being able to model the interaction between multiple factors at the same time. In the new energy industry, energy fluctuations are not only affected by a single factor, but are the result of the interaction of multiple factors. These factors include climate conditions, changes in market demand, policy regulation, etc. The multivariate processing capabilities of the VARMA model enable it to effectively model these interactions, thereby improving the accuracy of predictions.

[0045] In order to select the optimal VARMA model order, the present invention adopts the Akaike Information Criterion (AIC). AIC can find a balance between model complexity and prediction ability, and avoid overfitting through penalty terms. In the prediction of new energy power grids, the application of AIC helps determine the optimal autoregressive order and moving average order of the VARMA model, so that the model can more accurately reflect the changing trend of energy demand.

[0046] Once the order of the model is determined, the present invention uses the quasi-Newton method (BFGS algorithm) to optimize the coefficients in the VARMA model. The BFGS algorithm can quickly find the optimal solution in the multidimensional parameter space through an efficient iterative update process. This is crucial for large-scale energy consumption data in the new energy industry, because these data are often high-dimensional and change rapidly. Through the efficient optimization of the BFGS algorithm, the parameters of the model can converge quickly, thereby ensuring the real-time and accuracy of the prediction.

[0047] In the maximum likelihood estimation process, the Akaike Information Criterion (AIC) is not only used to select the order of the VARMA model, but also plays an important role in the maximum likelihood estimation. The introduction of AIC can ensure that the optimal balance between the complexity and fit of the model is achieved when estimating parameters. Specifically, AIC helps avoid errors caused by overfitting of the model during the maximum likelihood estimation process by penalizing the complexity of the model, thereby improving the predictive ability of the model.

[0048] A significant advantage of this method is that it can adapt to the energy fluctuation characteristics in the new energy power grid. The fluctuation of new energy power generation is strong, and traditional load forecasting methods often cannot effectively handle such fluctuations. However, through multi-stage modeling, the present invention can not only capture the long-term growth trend, but also accurately simulate short-term seasonal fluctuations and random fluctuations, thereby providing a scientific basis for load scheduling and resource allocation of the power grid.

[0049] In addition, the application of the BFGS algorithm can greatly improve the optimization calculation efficiency. Compared with the traditional gradient descent method, the BFGS algorithm can more effectively use gradient information in the calculation process, avoid calculation bottlenecks, and quickly obtain the optimal solution when facing large-scale data. This enables the present invention to maintain high real-time performance and accuracy when facing high-frequency and dynamically changing energy consumption data in new energy power grids.

[0050] In summary, the present invention optimizes the VARMA model by combining the Akaike Information Criterion (AIC) and the Quasi-Newton Method (BFGS algorithm), which can effectively cope with the challenge of predicting energy fluctuations in the new energy industry. Through the combination of time series decomposition and the VARMA model, the present invention can capture the various components of energy fluctuations in the new energy industry, and consider the interaction of multiple factors in the optimization process, thereby providing accurate energy consumption forecasts and reasonable load scheduling recommendations for power grid planning. This method not only has high theoretical value, but also has strong practical application prospects, and can provide strong support for load management, resource allocation and stability of new energy power grids.

[0051] Specifically, for the collected energy consumption data, the Z-score standardization method is used to convert the data into standardized values ​​to eliminate the deviation between different dimensions and ensure the uniformity of the data. The Z-score standardization formula is as follows:

[0052] in, represents the standardized energy consumption data, Represents the raw energy consumption data at the time point The value at represents the sample mean and the center position of the data. Represents the sample standard deviation, which measures the degree of fluctuation of the data. Specifically, the maximum likelihood estimation (MLE) method is used to estimate the model parameters and Then, the quasi-Newton method (BFGS algorithm) is used to optimize and minimize the log-likelihood function. The specific steps are as follows: Initialize model parameters

[0053] Calculate the initial log-likelihood function

[0054] Compute the gradient and Hessian matrix:

[0055] in, represents the gradient of the log-likelihood function, represents the partial derivative of the log-likelihood function with respect to the model parameters, Represents the autoregressive coefficient, which represents the relationship with the historical value (autoregressive term). In the model, Used to describe the time series. The impact of energy consumption data at a certain moment on the current moment, Represents the moving average coefficient, which expresses the relationship with the error term (noise term). To capture the past The impact of the error of the moment on the current moment, represents the log-likelihood function with respect to the autoregressive coefficient The partial derivative of , which represents the autoregressive coefficient The degree of influence on the value of the log-likelihood function, Represents the log-likelihood function with respect to the moving average coefficient The partial derivative of , which represents the moving average coefficient The degree of influence on the value of the log-likelihood function, represents the Hessian matrix, which represents the second-order derivative matrix of the log-likelihood function and describes the relationship between each pair of parameters (e.g. and ) Denotes the log-likelihood function with respect to and The second-order partial derivative of , which represents the autoregressive coefficient and The second-order effect on the log-likelihood function is represents the autoregressive coefficients and represents the two parameters of the autoregressive part of the model.

[0056] gradient is the partial derivative of the log-likelihood function with respect to each parameter, and the Hessian matrix Represents the second derivative, which is used to capture the curvature of the curve.

[0057] In each iteration, the parameters are adjusted using the update rule of the BFGS algorithm:

[0058] in, Indicates The parameter vector in the iteration represents all the parameters of the model in the current iteration step, including And parameters such as the variance of the noise term, Represents the step size, which indicates the size of the update in each iteration, usually calculated by a line search algorithm so that the update of each step minimizes the log-likelihood function. Represents the Hessian matrix, which represents the second-order derivative matrix of the current iteration step, and is used to adjust the direction and pace of the gradient. Represents the gradient of the current iteration, which represents the partial derivative of the current parameter with respect to the log-likelihood function. Indicated in The updated parameters in the iteration.

[0059] Repeat the above steps until the log-likelihood function converges to the maximum value, that is, the optimal parameter is found and In maximum likelihood estimation, the autoregressive coefficients of the optimization model are and the moving average coefficient The model parameters are updated by the quasi-Newton method, and finally the AIC is used to select the optimal order.

[0060] Initialization parameters: In the initial iteration, set the autoregressive order and the moving average order , and initialize all coefficients and

[0061] Calculate the gradient and Hessian matrix: Based on the log-likelihood function and the current parameter value, calculate the gradient and Hessian matrix to indicate the impact on the parameters.

[0062] Parameter update: Use the BFGS algorithm to update the parameters and continue iterating until the log-likelihood function converges.

[0063] AIC selects the optimal model: By minimizing the AIC value, the optimal order is selected and Finally determine the model parameters.

[0064] The division of power grid areas includes calculating the energy demand fluctuation of different power grid areas based on energy consumption data generated by an energy consumption fluctuation prediction model; according to the differences in the energy demand fluctuations, combined with the geographical information and load distribution of the power grid, a K-means clustering algorithm is used to determine the division results of the power grid areas by minimizing the variance of the load fluctuations in each area.

[0065] Specifically, the goal of grid regional division is to identify the load fluctuation patterns of different regions based on the energy consumption data generated by the energy consumption fluctuation prediction model, and then reasonably divide the grid area to ensure that the subsequent energy storage system configuration and renewable energy access can be more effectively optimized. In this process, the load fluctuation degree of each region is first calculated through the energy consumption fluctuation prediction model. This indicator reflects the magnitude of grid load changes over time. In order to describe the load fluctuation degree, this indicator is defined as the fluctuation amplitude of the load in each region within a certain time window. The calculation formula is as follows:

[0066] in, It is a region The load fluctuation, It's time point Lower area The actual load value, is the load forecast value, is the time window size, which represents the calculation period of load fluctuation.

[0067] Using these fluctuation data, the K-means clustering algorithm is used to divide the regions by combining the geographic information and load distribution of the power grid. The K-means algorithm is a distance-based clustering algorithm that aims to divide the power grid into several regions with similar load fluctuation characteristics by minimizing the variance of load fluctuation within each cluster. The steps are as follows: Initialize cluster centers: randomly selected There are initial cluster centers, each cluster corresponds to a preliminary division of a power grid area.

[0068] Calculate load fluctuation: Based on the historical load data of each area, calculate the load fluctuation ( ).

[0069] Assign regions to clusters: Assign each grid region to the cluster center closest to it and calculate the variance of load fluctuation within the cluster.

[0070] Update cluster center: Calculate the mean of load fluctuations of all regions in each cluster and update the mean as the new cluster center.

[0071] Iterative update: Repeat steps 3 and 4 until the load fluctuation variance converges.

[0072] The configuration of the energy storage system includes optimizing the capacity and distribution of the energy storage system using a particle swarm optimization algorithm based on the energy demand fluctuation of each power grid area and the results of the energy fluctuation prediction model; optimizing the energy storage system configuration of each power grid area by minimizing the difference between the energy storage demand and the power grid load fluctuation, thereby improving the load regulation capability and reducing the power grid load fluctuation.

[0073] Specifically, after completing the grid area division, the next step is to optimize the configuration of the energy storage system based on the energy demand fluctuation of each grid area and the results of the energy fluctuation prediction model. This step is critical because reasonable energy storage configuration can effectively improve the load regulation capability of the grid and reduce the impact of load fluctuations on grid stability.

[0074] Particle swarm optimization is the core algorithm for optimizing the capacity and distribution of energy storage systems. PSO simulates the foraging behavior of bird flocks and relies on the position and speed of particles to search the solution space. The PSO algorithm is used to minimize the difference between energy storage demand and grid load fluctuations. Its optimization goal is to reduce the fluctuation amplitude of the grid by configuring appropriate energy storage capacity. The optimization process includes the following steps: Initialize particle swarm: Initialize multiple particles for each grid area. The position of the particles represents the capacity of the energy storage system, while the speed controls the adjustment direction and amplitude of the energy storage capacity.

[0075] Fitness evaluation: For each particle, calculate its fitness value, which is determined by the change in grid load fluctuation. The fitness function is expressed as:

[0076] in, It is a region The load fluctuation, is the load fluctuation after optimization of the energy storage system, and the goal is to minimize this difference. is the number of grid areas.

[0077] Particle update: Update the speed and position of the particle according to the fitness value. The position of the particle represents the optimized configuration of the energy storage system.

[0078] Termination condition: When the fitness of the particle reaches the preset optimal level or after multiple iterations, the algorithm ends and returns the final optimization result.

[0079] The access to renewable energy includes optimizing the timing and proportion of renewable energy access based on the configuration parameters of the energy storage system and the results of the energy consumption fluctuation prediction model by using a genetic algorithm; optimizing the utilization efficiency of renewable energy by minimizing grid load fluctuations and maximizing the access proportion of renewable energy, thereby further improving the stability and sustainability of the grid.

[0080] Specifically, after the energy storage system is configured, the next task is to optimize the timing and proportion of renewable energy access. The access of renewable energy can not only improve the sustainability of the power grid, but also reduce dependence on traditional energy. However, too much renewable energy access may aggravate load fluctuations, so scientific optimization is needed.

[0081] Genetic algorithm is an optimization algorithm that imitates the biological evolution process and is used to optimize the access timing and proportion of renewable energy. Genetic algorithm gradually approaches the optimal solution based on operations such as selection, crossover and mutation. The specific steps are as follows: Initialize the population: randomly generate a set of combinations of renewable energy access timing and proportions, and each individual in the population represents an access plan.

[0082] Fitness evaluation: Calculate the fitness of each solution. The goal of the fitness function is to minimize the load fluctuation of the power grid and maximize the access ratio of renewable energy. The fitness function is expressed as:

[0083] in, It is a region The proportion of renewable energy access is a weight coefficient used to balance the relationship between minimizing load fluctuations and maximizing renewable energy.

[0084] Selection, crossover and mutation: The selection operation selects individuals with better fitness from the population, the crossover operation generates new individuals, and the mutation operation increases the diversity of the population.

[0085] Termination condition: When the fitness reaches the predetermined optimal value or after a sufficient number of generations, the algorithm ends and returns the optimal renewable energy access plan.

[0086] In terms of optimization of grid load regulation and renewable energy access, the method of this embodiment provides a more accurate and efficient grid planning method by innovatively combining energy fluctuation prediction models and multiple intelligent optimization algorithms. Compared with traditional methods, the present invention can fully consider the load fluctuation characteristics of the grid area, reasonably divide the grid area, optimize the configuration of the energy storage system, and improve the access efficiency of renewable energy, thereby improving the stability of the grid while reducing the operating cost and environmental impact of the grid.

[0087] Traditional methods of dividing power grid regions are often based only on the geographical information of the power grid or static analysis of the load, ignoring the dynamic characteristics of energy fluctuations. This division method cannot reflect the actual situation of power grid load fluctuations, resulting in insufficient power grid load regulation capabilities and an inability to maintain system stability when load fluctuations are large. In contrast, the present invention calculates the volatility of power grid regions through load fluctuation data generated by an energy fluctuation prediction model, and uses a K-means clustering algorithm to divide the power grid into regions. The advantage of this method is that it can dynamically divide power grid regions through accurate load fluctuation data, minimize load fluctuations within the region, and thus achieve more accurate load regulation and power grid optimization. This method not only ensures the stability of loads in each region of the power grid, but also provides a more reasonable basis for the subsequent configuration of the energy storage system.

[0088] In terms of energy storage system configuration, traditional methods usually determine the energy storage capacity and layout based on load peak and valley values ​​or simple rules of thumb, which often leads to problems of over-configuration or under-configuration. This method lacks a detailed analysis of the load fluctuation characteristics of the power grid, resulting in the inability to maximize the efficiency of the energy storage system and may even cause waste of resources. The present invention uses a particle swarm optimization algorithm, combined with the results of the load fluctuation and energy consumption fluctuation prediction model of each power grid area, to accurately optimize the capacity and layout of the energy storage system. The introduction of the particle swarm optimization algorithm enables the configuration of the energy storage system to be optimized on a global scale, thereby effectively improving the load regulation capability of the energy storage system, reducing power grid fluctuations, and reducing investment and operating costs.

[0089] For the access of renewable energy, traditional methods usually only consider the impact of load fluctuations to a certain extent, and often rely on static load balancing models. This method cannot fully utilize the power generation capacity of renewable energy when the grid load fluctuates greatly, and may even lead to excessive waste of renewable energy. The present invention uses a genetic algorithm to dynamically optimize the access timing and proportion of renewable energy based on the results of the energy storage system configuration and the energy consumption fluctuation prediction model. The genetic algorithm can comprehensively consider the grid load fluctuations and the access benefits of renewable energy, ensuring that the access proportion of renewable energy is increased without increasing the grid load fluctuations. This optimization process effectively improves the efficiency of renewable energy use, while reducing the potential risks brought by grid fluctuations and ensuring the stable operation of the grid.

[0090] Example 2 An embodiment of the present invention is different from the previous embodiment in that: If the functions are implemented in the form of 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 the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0091] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0092] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0093] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0094] Example 3 As an embodiment of the present invention, a grid planning method considering the energy consumption characteristics of emerging industries is provided. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0095] This experiment was conducted in a simulated medium-sized urban power grid environment. The power grid was divided into 10 regions, each of which represents a typical emerging industry energy consumption characteristic, such as electric vehicle charging, smart manufacturing, photovoltaic power generation, etc. The energy consumption data of each region comes from the actual collected load fluctuation data, covering multiple time periods. The specific data source includes the power load data of the past year. In order to ensure the authenticity and operability of the experiment, the experiment uses MATLAB and Python platforms for simulation. The experimental environment simulates the operation of the power grid in high-load seasons, low-load seasons, and when the load fluctuations are large.

[0096] Traditional methods mainly rely on grid area division, energy storage system configuration and renewable energy access based on grid geographic information, load density and historical load data. The experiment first divides the grid area based on the geographic information of the grid and the historical load data of each grid area. The traditional method uses a fixed load density threshold to divide the grid according to the load level. Each divided area does not fully consider the difference in energy fluctuations, resulting in areas with large fluctuations being divided into the same area, resulting in low scheduling efficiency. Next, the traditional method configures an energy storage system for each grid area. The energy storage capacity configuration is usually allocated according to a certain proportion (10%) of the total grid capacity, ignoring the differences in energy fluctuation characteristics between regions. Finally, in the renewable energy access stage, the traditional method usually sets a fixed renewable energy access ratio (30% of renewable energy is fixedly connected to each area), and does not optimize according to grid load fluctuations and energy storage system configuration, which may lead to low renewable energy utilization in some areas and increased grid fluctuations.

[0097] The method of the present invention first divides the power grid area, and analyzes and predicts the energy consumption fluctuation characteristics of each area by constructing an energy consumption fluctuation prediction model. Specifically, the VARMA multivariate autoregressive moving average model is used to analyze the energy consumption data of the power grid area, decompose the trend, seasonality and random fluctuation components, and optimize the model parameters through maximum likelihood estimation and quasi-Newton method (BFGS) to ensure the accuracy of the model. On this basis, the K-means clustering algorithm is used to combine load fluctuation, geographic information and load distribution to accurately divide the power grid area. The load fluctuation of each area is obtained through the output result of the energy consumption fluctuation prediction model. After the division, the load fluctuation characteristics of each area are ensured to be consistent, minimizing the risks brought by power grid load fluctuations.

[0098] In the energy storage system configuration stage, based on the load fluctuation and energy consumption fluctuation prediction results of each region, the particle swarm optimization algorithm (PSO) is used to dynamically adjust the capacity and distribution of the energy storage system, and optimize the configuration of the energy storage system in each region by minimizing the difference between the energy storage demand and the grid load fluctuation. Finally, in the renewable energy access stage, based on the configuration of the energy storage system and the output of the energy consumption fluctuation prediction model, the genetic algorithm (GA) is used to optimize the timing and proportion of renewable energy access, and to ensure the stability and sustainable operation of the grid by maximizing the utilization efficiency of renewable energy and minimizing grid load fluctuation. The experimental results are shown in Table 1.

[0099] Table 1 Comparison of experimental results index Method of the present invention Traditional methods Standard deviation of load fluctuation by region 0.12 0.23 Energy storage system configuration efficiency 85% 65% Renewable energy access ratio 40% 30% Grid load balancing error 3% 8% Renewable energy efficiency 90% 70% Energy storage system capacity configuration accuracy high Low By comparing the data in the table, first, when dividing the power grid area, the method of the present invention uses the energy fluctuation prediction model to accurately calculate the load fluctuation, thereby obtaining a more accurate regional division result. Compared with the traditional method, the method of the present invention can significantly reduce the load fluctuation difference between regions and improve the efficiency of power grid dispatching.

[0100] Secondly, in terms of energy storage system configuration, the traditional method configures energy storage capacity through a fixed ratio, fails to fully consider the load fluctuation characteristics of each area, and leads to low energy storage efficiency. In contrast, the method of the present invention uses a particle swarm optimization algorithm to dynamically adjust according to the specific needs of each area, so that the capacity and distribution of the energy storage system are more in line with the actual load fluctuation needs of the power grid, and the energy storage efficiency is improved by 20%.

[0101] In terms of renewable energy access, the traditional method uses a fixed ratio of access, which fails to optimize according to grid load fluctuations, resulting in low renewable energy utilization in some areas. After optimizing the access timing and ratio through genetic algorithms, the method of the present invention successfully increases the renewable energy access ratio of the grid, effectively reduces load fluctuations, and improves the utilization efficiency of renewable energy.

[0102] Finally, through the comparison of grid load balancing errors in the experiment, the load regulation error of the method of the present invention is only 3%, while that of the traditional method is 8%. The method of the present invention can more accurately regulate the grid load, reduce grid fluctuations, and thus improve the stability and reliability of the grid.

[0103] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0104] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0105] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A grid planning method considering the energy consumption characteristics of emerging industries, characterized in that: include: Identify the energy consumption characteristics of emerging industries and build energy consumption fluctuation prediction models; Solve the energy fluctuation prediction model through load prediction algorithm and divide the power grid area; Based on the grid area division result, configure an energy storage system; Based on the configuration result of the energy storage system, renewable energy is connected.

2. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 1 is characterized by: The construction of the energy consumption fluctuation prediction model includes determining the new energy industry category; Define energy consumption characteristic parameters for each industry category at different life cycle stages; Collecting energy consumption data of the new energy industry based on the energy consumption characteristic parameters; The Z-score standardization method is used to preprocess the collected energy consumption data; The energy consumption fluctuation prediction model is constructed using the preprocessed energy consumption data.

3. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 2 is characterized by: The energy consumption fluctuation prediction model is constructed by using a time series decomposition method to decompose the energy consumption data into three parts: trend, seasonality and random fluctuation, and a VARMA multivariate autoregressive moving average model is applied to establish a dynamic energy consumption fluctuation prediction model.

4. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 3 is characterized by: The time series decomposition method is expressed as, in, Represents the original energy consumption data, at time The actual value of the moment, Represents the trend component, which is the medium- and long-term trend of data. represents the seasonal component, which is the periodic fluctuation in the data that repeats over time. Represents the random fluctuation component, i.e., the residual part, which is the random fluctuation in the data that cannot be explained by the trend and seasonal patterns; The VARMA multivariate autoregressive moving average model is expressed as, in, Indicates the current time The random fluctuation component of represents the autoregressive coefficient, represents the moving average coefficient, represents the error term, represents the autoregressive order, Represents the moving average order.

5. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 4 is characterized by: The energy consumption fluctuation prediction model is constructed by using the maximum likelihood estimation method to optimize parameters, and the optimal autoregressive coefficient and moving average coefficient are determined by calculating the log-likelihood function and maximizing the function, which is expressed as: in, represents the log-likelihood function, represents the variance of the noise term of the model, Represents the random fluctuation component obtained from the time series decomposition, at the time point The value at Indicates the total length of the time series, indicating the total number of samples of the data; Use the quasi-Newton method to optimize the parameters in the maximum likelihood estimation to minimize the log-likelihood function, initialize the parameters, calculate the gradient and Hessian matrix, and use the update rule of the BFGS algorithm to adjust the parameters; The AIC information criterion is used to select the optimal autoregressive order and moving average order, expressed as, in, represents the maximum value of the log-likelihood function, represents the number of parameters in the model, , including parameters of the autoregressive coefficient, moving average coefficient, and noise term variance.

6. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 5 is characterized by: The dividing of the power grid regions includes calculating the energy demand fluctuation of different power grid regions based on the energy consumption data generated by the energy consumption fluctuation prediction model; According to the difference in energy demand fluctuation, combined with the geographical information and load distribution of the power grid, the K-means clustering algorithm is used to determine the division result of the power grid area by minimizing the variance of the load fluctuation in each area.

7. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 6 is characterized by: The configuring of the energy storage system includes optimizing the capacity and distribution of the energy storage system using a particle swarm optimization algorithm according to the energy demand fluctuation of each power grid area and the result of the energy fluctuation prediction model; The energy storage system configuration for each grid area is optimized by minimizing the difference between energy storage demand and grid load fluctuations.

8. The grid planning method considering the energy consumption characteristics of emerging industries as claimed in claim 7 is characterized by: The access to renewable energy includes optimizing the timing and proportion of renewable energy access using a genetic algorithm based on the configuration parameters of the energy storage system and the results of the energy fluctuation prediction model; Optimize the efficiency of renewable energy use by minimizing grid load fluctuations and maximizing the proportion of renewable energy access.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the grid planning method considering the energy consumption characteristics of emerging industries described in any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the grid planning method considering the energy consumption characteristics of emerging industries described in any one of claims 1 to 8 are implemented.

Images