A timing simulation system and method

Through the seasonal differential autoregressive sliding average model, the calculation resources and scheduling power generation resources are optimized, and the problem of insufficient energy conversion efficiency and stability of distributed photovoltaic smart grids is solved, and more efficient energy conversion and prediction accuracy is achieved.

CN118801348BActive Publication Date: 2024-12-06ELECTRIC POWER PLANNING & ENG INST CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410802777.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-20
Publication Date
2024-12-06
Estimated Expiration
2044-06-20

AI Technical Summary

Technical Problem

Existing timing simulation systems are difficult to effectively handle the high data volume and complexity of distributed photovoltaic smart grids, resulting in insufficient energy conversion efficiency and unable to meet the grid operation needs in the context of new energy.

Method used

Adopting seasonal differential autoregressive sliding average model, through data acquisition and preprocessing, model construction and prediction, allocation and optimization of computing resources, and intelligent scheduling modules, optimize computing resource management and allocation, dispatch power generation resources, and improve the energy of photovoltaic power stations. Conversion efficiency and stability.

Benefits of technology

It improves the energy conversion efficiency and stability of distributed photovoltaic smart grids, enhances the prediction accuracy of photovoltaic power generation, and solves the problem of insufficient energy conversion efficiency and stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118801348B_ABST
    Figure CN118801348B_ABST
Patent Text Reader

Abstract

The present invention discloses a time series simulation system and method, and relates to the technical field of energy data processing. The system includes: a data acquisition and preprocessing module, a model building and prediction module, a computing resource allocation and optimization module, and an intelligent scheduling module. The present invention optimizes the calculation method, computing task scheduling, and model parameters by constructing a seasonal difference autoregressive sliding average model and managing and allocating computing resources, and uses a Bayesian optimization algorithm to adjust the parameters of the seasonal difference autoregressive sliding average model to find the best parameter combination, thereby improving the prediction accuracy of the model, inputting the preprocessed real-time power grid characteristic data into the model, and obtaining the predicted correction value of the power generation efficiency in the energy storage system, thereby improving the energy conversion efficiency of the distributed photovoltaic power station, and effectively solving the problem of insufficient energy conversion efficiency of the distributed photovoltaic smart grid in the prior art.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of energy data processing, and in particular to a timing simulation system and method. Background Art

[0002] With the expansion of the scale of new power systems, the rapid growth of the proportion of new energy access such as wind and solar power, the impact of large-capacity DC expansion transmission and new loads, the traditional typical method has a large disconnect between single-period power balance analysis and extensive power balance analysis, and has gradually failed to meet the requirements of power grid operation under the influence of massive uncertain factors. The typical characteristics are no longer obvious, and the typical operation mode can no longer cover all possible operation boundaries and cannot accurately simulate the power grid operation boundaries.

[0003] The existing timing simulation systems and methods are unable to accurately simulate the operation of the power system under the background of a high proportion of renewable energy, and are therefore unable to meet production needs. We must make full use of existing technologies, continuously improve the precision of timing operation simulation, software efficiency and expansion capabilities, increase the coverage of software functions, and improve the flexibility of human-computer interaction in the software, so as to support the power grid in achieving its goals from the perspectives of power supply and grid planning, grid operation, and electricity market participation.

[0004] For example, the patent application with publication number: CN117574684A discloses a method and system for simulating the time-series production of an electric-hydrogen-carbon integrated energy system, including: constructing a comprehensive objective function and constraints of the electric-hydrogen-carbon integrated energy system based on the electric-hydrogen conversion model, the carbon-gas-electricity coupling model and the non-green hydrogen penalty cost; the comprehensive objective function of the electric-hydrogen-carbon integrated energy system includes a first objective function with the goal of minimizing the operating cost of the integrated energy system and a second objective function with the goal of minimizing the non-green hydrogen penalty cost; the wind power energy constraint is determined by the wind power output data; the photovoltaic power constraint is determined by the photovoltaic output data, and the comprehensive objective function of the electric-hydrogen-carbon integrated energy system is solved to obtain a time-series production simulation scheme. The present invention comprehensively considers the randomness and volatility of new energy output, the utilization rate of green hydrogen energy, etc., establishes an electric-hydrogen-carbon integrated energy system model, and improves the flexibility, efficiency and accuracy of the integrated energy system.

[0005] For example, the operation simulation and index evaluation method based on the energy cell group of urban morphological units disclosed in the patent application with publication number: CN115221685A includes: establishing the equipment model in the electricity-gas-heat-cold integrated energy system; establishing the objective function that considers the economic cost and reliability cost of the integrated energy system, and setting the coupling constraint of the electricity-gas-heat-cold system; based on the constructed electricity-gas-heat-cold integrated energy scheduling and operation model, formulating the system economic cost, system reliability cost, system environmental protection, new energy consumption, and the proportion of energy supply of multi-energy coupling equipment as the evaluation index of the integrated energy system; based on the time series Monte Carlo sampling method, the time series operation simulation of the multi-energy system is carried out to quantitatively obtain the output of the coupling equipment and various evaluation indicators. It realizes the time series operation scheduling of the integrated energy system, analyzes the output of each equipment in the system, and evaluates the reliable, economical, and environmentally friendly operation indicators of the system, which has a great market application prospect.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In the existing technology, due to the continuous expansion of distributed photovoltaic smart grids, the amount and complexity of data that the timing simulation system needs to process are also increasing. The energy conversion efficiency of distributed photovoltaic power stations is optimized based on timing data, and the coordinated operation of photovoltaic power stations with energy storage systems and other renewable energy sources is supported. However, there is a problem of insufficient energy conversion efficiency of distributed photovoltaic smart grids. Summary of the invention

[0008] The embodiment of the present application solves the problem of insufficient energy conversion efficiency of distributed photovoltaic smart grids in the prior art by providing a timing simulation system and method, thereby achieving the effect of improving the energy conversion efficiency of distributed photovoltaic smart grids.

[0009] The embodiment of the present application provides a time series simulation system, including: a data acquisition and preprocessing module, a model building and prediction module, a computing resource allocation and optimization module and an intelligent scheduling module; the data acquisition and preprocessing module is used to collect real-time power grid data from sensors at various monitoring points of the power grid, clean and feature process the collected real-time power grid data, and obtain preprocessed real-time power grid feature data; the model building and prediction module is used to construct a seasonal difference autoregressive sliding average model, divide the preprocessed real-time power grid feature data into a training set and a test set, use the training set to train the seasonal difference autoregressive sliding average model to obtain the trained seasonal difference autoregressive sliding average model, use the trained seasonal difference autoregressive sliding average model to predict on the test set, and evaluate the prediction error of the seasonal difference autoregressive sliding average model; the computing resource allocation and optimization module is used to manage and allocate computing resources in constructing the seasonal difference autoregressive sliding average model, and optimize the computing method, computing task scheduling and model parameters; the intelligent scheduling module is used to optimize the allocation of power generation resources according to the prediction results of the seasonal difference autoregressive sliding average model in combination with the power grid operation status and power generation plan.

[0010] Furthermore, the specific processing process of cleaning and feature processing the collected real-time power grid data is as follows: the collected real-time power grid data is cleaned in combination with the historical data of the power grid to obtain pre-processed real-time power grid data, and the state and dynamic change characteristics of the power grid are extracted from the cleaned data to obtain pre-processed real-time power grid feature data.

[0011] Furthermore, the specific analysis process of the seasonal difference autoregressive moving average model is as follows: the seasonal pattern of the preprocessed real-time power grid characteristic data is analyzed, and seasonal differences are taken to remove the seasonal influence of the preprocessed real-time power grid characteristic data, so as to obtain the order, periodic time interval and residual of the seasonal difference autoregressive moving average model, which are used to test the fitting degree of the seasonal difference autoregressive moving average model.

[0012] Furthermore, the specific process of managing and allocating computing resources in constructing the seasonal difference autoregressive moving average model is as follows: evaluate the available existing computing resources, decompose the computing task into multiple subtasks according to the process of constructing the seasonal difference autoregressive moving average model, manage task execution according to the time step of the seasonal difference autoregressive moving average model, allocate computing resources to each subtask according to the computing requirements of the task, optimize resource utilization for parallel processing tasks using a parallel computing framework, and evenly distribute them through load balancing technology to obtain the computing load evaluation coefficient of the distributed power station.

[0013] Furthermore, the computing resource allocation and optimization module includes a computing task scheduling optimization unit; the computing task scheduling optimization unit: according to the computing resource allocation result, adopts the task scheduling strategy of shortest job first, and reduces the number of parameter combinations according to the system load and resource status.

[0014] Furthermore, the computing resource allocation and optimization module also includes a model parameter optimization unit; the model parameter optimization unit: finds the parameter combination of the seasonal difference autoregressive moving average model by adopting a Bayesian optimization algorithm, uses the mean square error to evaluate the model performance, obtains the energy storage device evaluation coefficient in the energy storage system, and adjusts the parameters of the seasonal difference autoregressive moving average model according to the energy storage device evaluation coefficient in the energy storage system.

[0015] Furthermore, the specific test process for testing the degree of fit of the seasonal difference autoregressive moving average model is as follows: white noise test is performed through residuals, the correlation of residuals is tested using the Durbin-Watson test method, the preprocessed real-time power grid characteristic data is divided into a training set and a test set, the seasonal difference autoregressive moving average model is trained using the training set to obtain the trained seasonal difference autoregressive moving average model, the trained seasonal difference autoregressive moving average model is used to perform predictions on the test set, the prediction error of the seasonal difference autoregressive moving average model is evaluated, the degree of fit of the seasonal difference autoregressive moving average model is obtained according to the prediction error of the seasonal difference autoregressive moving average model, and the predicted correction value of the power generation efficiency in the energy storage system is obtained according to the degree of fit of the seasonal difference autoregressive moving average model.

[0016] Furthermore, the specific optimization process for optimizing the allocation of power generation resources is as follows: evaluating the pre-processed real-time grid characteristic data, including the load of photovoltaic power generation and the capacity of energy storage equipment, maximizing the stability of the load of photovoltaic power generation and the capacity of energy storage equipment, minimizing environmental impact, and adjusting and optimizing the scheduling strategy based on the load evaluation coefficient calculated by the distributed power station, the energy storage equipment evaluation coefficient in the energy storage system and the predicted correction value of the power generation efficiency in the energy storage system. The adjustment and optimization includes adjusting the angle of the photovoltaic panels, replacing the photovoltaic panels and maintaining the photovoltaic panels.

[0017] Furthermore, the specific constraint formula of the power system evaluation coefficient is:

[0018]

[0019] In the formula, Represents the power system assessment coefficient, which is used to evaluate the comprehensive impact of distributed power station load and power generation efficiency in energy storage system on the power system. represents the load assessment factor for distributed power stations, It represents the evaluation coefficient of the energy storage equipment in the energy storage system, α represents the predicted correction value of the power generation efficiency in the energy storage system, and e represents a natural constant.

[0020] Furthermore, a time series simulation method is: collecting real-time power grid data from sensors at various monitoring points of the power grid, cleaning and feature processing the collected real-time power grid data to obtain pre-processed real-time power grid feature data, constructing a seasonal difference autoregressive moving average model based on the pre-processed real-time power grid feature data, dividing the pre-processed real-time power grid feature data into a training set and a test set, using the training set to train the seasonal difference autoregressive moving average model to obtain the trained seasonal difference autoregressive moving average model, using the trained seasonal difference autoregressive moving average model to perform predictions on the test set, evaluating the prediction error of the seasonal difference autoregressive moving average model, managing and allocating computing resources in constructing the seasonal difference autoregressive moving average model, optimizing the computing method, computing task scheduling and model parameters, and optimizing the allocation of power generation resources based on the prediction results of the seasonal difference autoregressive moving average model in combination with the power grid operation status and power generation plan.

[0021] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0022] 1. By constructing the management and allocation of computing resources in the seasonal difference autoregressive moving average model, the calculation method, computing task scheduling and model parameters are optimized, thereby improving the energy conversion efficiency of distributed photovoltaic power stations and effectively solving the problem of insufficient energy conversion efficiency of distributed photovoltaic smart grids in existing technologies.

[0023] 2. By predicting the results based on the seasonal difference autoregressive sliding average model, the allocation of power generation resources is optimized in combination with the grid operation status and power generation plan, thereby improving the stability of the distributed photovoltaic smart grid and effectively solving the problem of insufficient stability of the distributed photovoltaic smart grid in the existing technology.

[0024] 3. The Durbin-Watson test method is used to test the correlation of residuals, thereby evaluating the prediction error of the seasonal difference autoregressive moving average model. The degree of fit of the seasonal difference autoregressive moving average model is obtained according to the prediction error of the seasonal difference autoregressive moving average model, thereby improving the prediction accuracy of photovoltaic power generation, and effectively solving the problem of insufficient prediction accuracy of photovoltaic power generation in distributed photovoltaic smart grids in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of the structure of a timing simulation system provided in an embodiment of the present application;

[0026] Figure 2 A schematic diagram of the structure of a computing resource allocation and optimization module in a timing simulation system provided in an embodiment of the present application;

[0027] Figure 3 An image of a distributed power station calculating a load assessment coefficient in a timing simulation system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The embodiment of the present application solves the problem of insufficient energy conversion efficiency of distributed photovoltaic smart grids in the prior art by providing a time series simulation system. By constructing a seasonal difference autoregressive sliding average model to manage and allocate computing resources, the calculation method, computing task scheduling and model parameters are optimized, thereby achieving improved energy conversion efficiency of distributed photovoltaic power stations.

[0029] The technical solution in the embodiment of the present application is to solve the problem of insufficient energy conversion efficiency of the above-mentioned distributed photovoltaic smart grid. The overall idea is as follows:

[0030] By constructing a seasonal difference autoregressive moving average model based on preprocessed real-time power grid characteristic data, the computing resources in the seasonal difference autoregressive moving average model are managed and allocated, and the calculation method, computing task scheduling and model parameters are optimized. According to the prediction results of the seasonal difference autoregressive moving average model, the allocation of power generation resources is optimized in combination with the power grid operation status and power generation plan, thereby improving the energy conversion efficiency of distributed photovoltaic power stations.

[0031] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0032] like Figure 1As shown, it is a structural schematic diagram of a time series simulation system provided in an embodiment of the present application. A time series simulation system provided in an embodiment of the present application includes: a data acquisition and preprocessing module, a model building and prediction module, a computing resource allocation and optimization module and an intelligent scheduling module; the data acquisition and preprocessing module is used to collect real-time power grid data from sensors at various monitoring points of the power grid, clean and feature process the collected real-time power grid data, and obtain preprocessed real-time power grid feature data; the model building and prediction module is used to construct a seasonal difference autoregressive sliding average model, divide the preprocessed real-time power grid feature data into a training set and a test set, and use the training set The seasonal difference autoregressive moving average model is trained to obtain the trained seasonal difference autoregressive moving average model, and the trained seasonal difference autoregressive moving average model is used to make predictions on the test set to evaluate the prediction error of the seasonal difference autoregressive moving average model; the computing resource allocation and optimization module is used to manage and allocate computing resources in the construction of the seasonal difference autoregressive moving average model, and optimize the calculation method, computing task scheduling and model parameters; the intelligent scheduling module is used to optimize the allocation of power generation resources according to the prediction results of the seasonal difference autoregressive moving average model in combination with the power grid operation status and power generation plan.

[0033] Furthermore, the specific processing process of cleaning and feature processing the collected real-time power grid data is as follows: the collected real-time power grid data is cleaned in combination with the historical data of the power grid to obtain pre-processed real-time power grid data, and the state and dynamic change characteristics of the power grid are extracted from the cleaned data to obtain pre-processed real-time power grid feature data.

[0034] In this embodiment, the historical data of the power grid include historical data of power generation of the power grid, historical load rate data, historical operating time data, power quality data and environmental parameter data; the real-time power grid data include environmental temperature, humidity, wind speed, light intensity, power injected into the power grid, power received from the power grid, voltage, current, harmonic content and power factor; the principal component analysis algorithm is used to reduce the dimension of the data, extract data features, establish a hierarchical tree of clusters, and merge or split clusters according to distance and cohesion to obtain preprocessed real-time power grid feature data.

[0035] Furthermore, the specific analysis process of the seasonal difference autoregressive moving average model is as follows: the seasonal pattern of the preprocessed real-time power grid characteristic data is analyzed, seasonal differences are taken to remove the seasonal influence of the preprocessed real-time power grid characteristic data, and the order, periodic time interval and residual of the seasonal difference autoregressive moving average model are determined to test the fitting degree of the seasonal difference autoregressive moving average model.

[0036] In this embodiment, the preprocessed real-time power grid characteristic data is analyzed to obtain a seasonal pattern of the preprocessed real-time power grid characteristic data, the seasonal pattern includes determined seasonality, random stationary seasonality and random non-stationary seasonality, and the preprocessed real-time power grid characteristic data is seasonally differentiated, the periodic difference order is D, and the seasonal difference is usually between 0-3. Determine the seasonal difference number D from 0 to 3 cycles, stop after stability, when testing model parameters D = 1, the data is stable, autoregressive term order p, moving average term order q, seasonal autoregressive term order P, seasonal moving average term order Q, use Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to determine the order, p is usually between 1-3, through the cycle, it can be obtained that when p = 2, the AIC value and BIC value are the smallest, q is usually between 1-3, through the cycle, it can be obtained that when q = 1, the AIC value and BIC value are the smallest, P is usually between 1-2, through the cycle, it can be obtained that when P = 2, the AIC value and BIC value are the smallest, Q is usually between 1-2, through the cycle, it can be obtained that when Q = 1, the AIC value and BIC value are the smallest, according to the order, period time interval and residual of the seasonal difference autoregressive moving average model, the fitting degree of the seasonal difference autoregressive moving average model is tested, and according to the fitting degree of the seasonal difference autoregressive moving average model, accurate prediction can be made and used to allocate resources.

[0037] Furthermore, the specific process of managing and allocating computing resources in constructing the seasonal difference autoregressive moving average model is as follows: evaluate the available existing computing resources, decompose the computing task into multiple subtasks according to the process of constructing the seasonal difference autoregressive moving average model, manage task execution according to the time step of the seasonal difference autoregressive moving average model, allocate computing resources to each subtask according to the computing requirements of the task, optimize resource utilization for parallel processing tasks using a parallel computing framework, and evenly distribute them through load balancing technology to obtain the computing load evaluation coefficient of the distributed power station.

[0038] In this embodiment, the existing computing resources are evaluated, including the number of CPU cores, memory size and storage speed of the server. Subsequently, the seasonal difference autoregressive moving average model construction process is decomposed into subtasks such as data preprocessing, model selection, parameter tuning and model evaluation. In the data preprocessing stage, due to the large amount of data, we decompose it into multiple sub-datasets. The task scheduling tool used here is the MapReduce parallel computing framework. The specific processing steps of using the MapReduce parallel computing framework to process these sub-datasets simultaneously are: Integrate the selected distributed task queue system in the application, distribute subtasks to the distributed task queue, and the scheduling system reasonably allocates tasks to worker nodes according to the principles of task priority, execution dependency, etc. The worker node obtains tasks from the task queue, executes tasks, and handles possible exceptions to speed up the preprocessing speed;

[0039] In the model training phase, a distributed task queue is used to manage the specific management process of the execution of different subtasks: design the logic of the Map phase, implement a Map function, which reads the input sub-dataset, sorts and groups the key-value pairs output by the Map phase according to the key, design the logic of the Reduce phase, implement a Reduce function, which receives the same key and the corresponding value list and processes it. The MapReduce framework usually has a fault-tolerant mechanism that can re-execute tasks when tasks fail, ensuring that they are queued and executed on appropriate computing resources;

[0040] In order to achieve load balancing, use Nginx to allocate tasks on different servers. The specific allocation process is as follows: set the server list in the Nginx configuration file, set Nginx's load balancing strategy, configure Nginx's health check function, the client sends a request to the Nginx server, and after Nginx receives the request, it selects a server according to the configured load balancing strategy. The target server receives the request and processes the task to ensure that each server can run efficiently.

[0041] Furthermore, the computing resource allocation and optimization module includes a computing task scheduling optimization unit; the computing task scheduling optimization unit: according to the computing resource allocation result, adopts the task scheduling strategy of shortest job first, and reduces the number of parameter combinations according to the system load and resource status.

[0042] In this embodiment, the computing resource allocation and optimization module first allocates computing resources according to the characteristics of the task and the resource status of the system. For example, the demand forecasting task with high real-time requirements is allocated to the server with stronger performance. The task scheduling strategy of shortest job first is adopted, which will give priority to those tasks with short expected running time, such as fast equipment status check, to ensure that these tasks can be processed quickly. Through optimization, the module can reduce unnecessary parameter combinations, such as reducing excessive historical data features in the prediction model to reduce the computing burden. By optimizing the scheduling of computing tasks, computing resources can be used more efficiently and waiting time can be reduced.

[0043] Furthermore, the computing resource allocation and optimization module also includes a model parameter optimization unit; the model parameter optimization unit: by adopting the Bayesian optimization algorithm to find the parameter combination of the seasonal difference autoregressive moving average model, using the mean square error to evaluate the model performance, obtaining the energy storage device evaluation coefficient in the energy storage system, and adjusting the parameters of the seasonal difference autoregressive moving average model according to the energy storage device evaluation coefficient in the energy storage system.

[0044] In this embodiment, the model parameter optimization unit uses a Bayesian optimization algorithm to find the best parameter combination of the seasonal difference autoregressive moving average model. This process involves searching for the values ​​of multiple parameters to find the parameter configuration that minimizes the model prediction error, using the mean square error to evaluate the performance of the seasonal difference autoregressive moving average model, and obtaining the energy storage device evaluation coefficient. According to the optimized energy storage device evaluation coefficient, the parameters of the seasonal difference autoregressive moving average model are adjusted to more accurately simulate the performance of the energy storage device. By optimizing the model parameters, the model can be better adapted to the data characteristics, thereby improving the prediction accuracy.

[0045] Furthermore, the specific test process for testing the degree of fit of the seasonal difference autoregressive moving average model is as follows: white noise test is performed through residuals, the correlation of residuals is tested using the Durbin-Watson test method, the preprocessed real-time power grid characteristic data is divided into a training set and a test set, the seasonal difference autoregressive moving average model is trained using the training set to obtain the trained seasonal difference autoregressive moving average model, the trained seasonal difference autoregressive moving average model is used to perform predictions on the test set, the prediction error of the seasonal difference autoregressive moving average model is evaluated, the degree of fit of the seasonal difference autoregressive moving average model is obtained according to the prediction error of the seasonal difference autoregressive moving average model, and the predicted correction value of the power generation efficiency in the energy storage system is obtained according to the degree of fit of the seasonal difference autoregressive moving average model.

[0046] In this embodiment, the preprocessed data is divided into a training set and a test set. The seasonal difference autoregressive moving average model is trained using the training set to obtain the trained seasonal difference autoregressive moving average model. The trained seasonal difference autoregressive moving average model is used to make predictions on the test set to evaluate the prediction error of the seasonal difference autoregressive moving average model. The parameters of the seasonal difference autoregressive moving average model are determined based on the historical data of electricity demand. The power load data of the past year is used to train the seasonal difference autoregressive moving average model. In order to ensure the accuracy of the model, the power load data of the past week is used as the validation set to test the performance of the model. The prediction accuracy of the seasonal difference autoregressive moving average model on the validation set is evaluated. The correlation of the prediction residuals is tested by the Durbin-Watson test method to determine whether the model is overfitting or underfitting. The fitting degree of the seasonal difference autoregressive moving average model is obtained according to the prediction error of the seasonal difference autoregressive moving average model, thereby improving the prediction accuracy of the model. The preprocessed real-time power grid characteristic data is input into the seasonal difference autoregressive moving average model to obtain the predicted correction value of the power generation efficiency in the energy storage system. By evaluating the fitting degree of the model, it can be ensured that the prediction model can accurately reflect the characteristics of the actual data, thereby improving the accuracy of the prediction.

[0047] Furthermore, the specific optimization process for optimizing the allocation of power generation resources is as follows: evaluating the pre-processed real-time grid characteristic data, including the load of photovoltaic power generation and the capacity of energy storage equipment, maximizing the stability of the load of photovoltaic power generation and the capacity of energy storage equipment, minimizing environmental impact, and adjusting and optimizing the scheduling strategy based on the load evaluation coefficient calculated by the distributed power station, the energy storage equipment evaluation coefficient in the energy storage system and the predicted correction value of the power generation efficiency in the energy storage system. The adjustment and optimization includes adjusting the angle of the photovoltaic panels, replacing the photovoltaic panels and maintaining the photovoltaic panels.

[0048] In this embodiment, the pre-processed real-time grid characteristic data is evaluated, including the load characteristics of the power station, the capacity and power generation efficiency of the energy storage equipment, and the overall power generation efficiency of the photovoltaic power station, to maximize the power generation efficiency and the stability of the power supply, while minimizing the environmental impact. According to the load evaluation coefficient of the distributed power station, the evaluation coefficient of the energy storage equipment in the energy storage system, and the predicted correction value of the power generation efficiency in the energy storage system, the scheduling strategy is adjusted and optimized. The adjustment and optimization includes adjusting the angle of the photovoltaic panels, replacing the photovoltaic panels, and maintaining the photovoltaic panels. As the photovoltaic panels are used for a longer time, dust, dirt, and even plant leaves and other debris will accumulate on the surface, affecting the absorption of light. Regular cleaning and maintenance can remove these impurities, restore the optical efficiency of the photovoltaic panels, and thus increase the power generation. Replacing old or damaged photovoltaic panels can also restore the power generation capacity in time to avoid affecting the overall power generation efficiency due to problems with a single component.

[0049] Furthermore, the specific constraint formula of the power system evaluation coefficient is:

[0050]

[0051] In the formula, Represents the power system assessment coefficient, which is used to evaluate the comprehensive impact of distributed power station load and power generation efficiency in energy storage system on the power system. represents the load assessment factor for distributed power stations, It represents the evaluation coefficient of the energy storage equipment in the energy storage system, α represents the predicted correction value of the power generation efficiency in the energy storage system, and e represents a natural constant.

[0052] In this embodiment, the specific constraint formula for calculating the load assessment coefficient of the distributed power station is:

[0053]

[0054] In the formula, It represents the calculation load assessment coefficient of distributed power station, which is used to assess the impact of distributed power station load on the power system. A represents the intensity of solar radiation, B represents the power generation efficiency of photovoltaic power station, C represents the standard temperature of semiconductor in photovoltaic power station, D represents the actual temperature of semiconductor in photovoltaic power station, and e represents a natural constant.

[0055] In a specific embodiment, an example of data for calculating load assessment coefficients of a distributed power station is shown in the following table.

[0056] Table 1 Example of data for calculating load assessment coefficients for distributed power stations

[0057]

[0058] From the data in Table 1, it can be seen that when the solar radiation intensity, the power generation efficiency of the photovoltaic power station, and the standard temperature of the semiconductor in the photovoltaic power station remain unchanged, by observing groups 1-5 of data, the greater the actual temperature of the semiconductor in the photovoltaic power station, the higher the calculated load assessment coefficient of the distributed power station, and the actual temperature of the semiconductor in the photovoltaic power station is positively correlated with the calculated load assessment coefficient of the distributed power station.

[0059] Figure 3An image of a distributed power station calculating a load evaluation coefficient in a timing simulation system provided by an embodiment of the present application, wherein a function curve a represents a fixed value 1 obtained by formula calculation after the solar radiation intensity, the power generation efficiency of the photovoltaic power station, the standard temperature of the semiconductor in the power station and the actual temperature of the semiconductor in the photovoltaic power station, and the distributed power station calculates the load evaluation coefficient increases with the increase of the actual temperature of the semiconductor in the photovoltaic power station, and the function curve b represents a fixed value 3 obtained by formula calculation after the solar radiation intensity, the power generation efficiency of the photovoltaic power station, the standard temperature of the semiconductor in the power station and the actual temperature of the semiconductor in the photovoltaic power station, and the distributed power station calculates the load evaluation coefficient increases with the increase of the actual temperature of the semiconductor in the photovoltaic power station, and the function curve c represents a fixed value 5 obtained by formula calculation after the solar radiation intensity, the power generation efficiency of the photovoltaic power station, the standard temperature of the semiconductor in the power station and the actual temperature of the semiconductor in the photovoltaic power station, and the distributed power station calculates the load evaluation coefficient increases with the increase of the actual temperature of the semiconductor in the photovoltaic power station, and the curve a, curve b and curve c represent function images under different independent variables, and when the actual temperature of the semiconductor in the photovoltaic power station takes the same value, the distributed power station calculates the load evaluation coefficient of the curve c is the largest, the curve b is the second, and the curve a is the smallest.

[0060] Solar radiation intensity refers to the solar radiation energy received per unit area per unit time. It can be measured in real time by solar radiation sensors installed at the photovoltaic power station site, or estimated by meteorological data and historical statistical data. The range of solar radiation intensity depends on the geographical location and weather conditions, and is approximately 100-500kWh / m 2 .

[0061] The power generation efficiency of a photovoltaic power station refers to the efficiency of the photovoltaic power station in converting the received solar radiation energy into electrical energy. The power generation efficiency of a photovoltaic power station can be tested when manufacturing photovoltaic panels and maintained and evaluated through performance monitoring during actual operation. The power generation efficiency of a photovoltaic power station usually ranges from 15% to 20%.

[0062] The standard temperature and actual temperature of semiconductors in photovoltaic power stations. The standard temperature of semiconductors in photovoltaic power stations usually refers to the temperature under standard test conditions, that is, the ambient temperature is 25°C, the humidity is 50%, and the photovoltaic panels are working in maximum power point tracking. The actual temperature of semiconductors in photovoltaic power stations usually refers to the temperature of semiconductor materials in photovoltaic panels under actual working conditions. The temperature of semiconductors can be obtained by real-time monitoring through the temperature sensor integrated in the photovoltaic panel. The standard temperature of semiconductors is 25°C, and the actual temperature of semiconductors ranges from 40°C to 60°C.

[0063] In this embodiment, the specific constraint formula of the energy storage device evaluation coefficient in the energy storage system is:

[0064]

[0065] In the formula, It represents the evaluation coefficient of the energy storage equipment in the energy storage system, which is used to evaluate the impact of the energy storage equipment in the distributed power station on the power system. H represents the rated capacity of the energy storage system, G represents the regulation frequency of the energy storage system, F represents the intervention frequency of the energy storage system, J represents the refractive index of the air, and e represents a natural constant.

[0066] The rated capacity of an energy storage system refers to the maximum amount of energy that the energy storage device can store and release under specific conditions, which is determined by actually measuring the charge and discharge curve of the energy storage system.

[0067] The regulation frequency of the energy storage system refers to the rate at which the energy storage system can regulate the grid frequency. The energy storage system can regulate the frequency by providing or absorbing active power between power generation and load. This can be obtained through the design parameters of the control system and the dynamic response characteristics of the energy storage system. The regulation frequency range is 50-60Hz.

[0068] The intervention frequency of the energy storage system refers to the frequency threshold at which the energy storage system begins to intervene in grid regulation. When the grid frequency deviates from this threshold, the energy storage system will start working to return to the set frequency range. The intervention frequency of the energy storage system can be obtained based on the safety and stability requirements of the system operators. The intervention frequency of the energy storage system is 60.2-60.5Hz.

[0069] The refractive index of air usually refers to the refraction of visible light by air. The refractive index will affect the propagation path of sunlight when it enters photovoltaic modules, which may affect the power generation efficiency of photovoltaic modules. The refractive index of air can be measured experimentally. The normal range of the refractive index of air is 1.0003-1.003.

[0070] Furthermore, a time series simulation method is: collecting real-time power grid data from sensors at various monitoring points of the power grid, cleaning and feature processing the collected real-time power grid data to obtain pre-processed real-time power grid feature data, constructing a seasonal difference autoregressive moving average model based on the pre-processed real-time power grid feature data, dividing the pre-processed real-time power grid feature data into a training set and a test set, using the training set to train the seasonal difference autoregressive moving average model to obtain the trained seasonal difference autoregressive moving average model, using the trained seasonal difference autoregressive moving average model to perform predictions on the test set, evaluating the prediction error of the seasonal difference autoregressive moving average model, managing and allocating computing resources in constructing the seasonal difference autoregressive moving average model, optimizing the computing method, computing task scheduling and model parameters, and optimizing the allocation of power generation resources based on the prediction results of the seasonal difference autoregressive moving average model in combination with the power grid operation status and power generation plan.

[0071] The technical solutions in the above-mentioned embodiments of the present application have at least the following technical effects or advantages: relative to a method and system for simulating the time-series production of an electric-hydrogen-carbon integrated energy system disclosed in the patent application with publication number: CN117574684A, the embodiment of the present application optimizes the calculation method, calculation task scheduling and model parameters by constructing a seasonal difference autoregressive sliding average model to manage and allocate computing resources, thereby improving the energy conversion efficiency of distributed photovoltaic power stations; relative to the operation simulation and index evaluation method based on urban morphological unit energy cell groups disclosed in publication number: CN115221685A, the embodiment of the present application optimizes the allocation of power generation resources in combination with the grid operation status and power generation plan according to the prediction results of the seasonal difference autoregressive sliding average model, thereby improving the stability of the distributed photovoltaic smart grid.

[0072] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0073] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0074] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0075] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0076] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0077] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A timing simulation system, characterized in that: It includes data collection and preprocessing module, model building and prediction module, computing resource allocation and optimization module and intelligent scheduling module; The data acquisition and preprocessing module is used to collect real-time power grid data from sensors at various monitoring points of the power grid, clean and feature process the collected real-time power grid data, and obtain preprocessed real-time power grid feature data; The model building and prediction module is used to build a seasonal difference autoregressive moving average model, divide the preprocessed real-time power grid characteristic data into a training set and a test set, use the training set to train the seasonal difference autoregressive moving average model to obtain the trained seasonal difference autoregressive moving average model, use the trained seasonal difference autoregressive moving average model to predict on the test set, evaluate the prediction error of the seasonal difference autoregressive moving average model, and analyze the power system evaluation coefficient according to the preprocessed real-time power grid characteristic data; The computing resource allocation and optimization module is used to manage and allocate computing resources in constructing a seasonal difference autoregressive sliding average model, and optimize the computing method, computing task scheduling and model parameters; The intelligent dispatching module is used to optimize the allocation of power generation resources based on the prediction results of the seasonal difference autoregressive moving average model in combination with the power system evaluation coefficient; The specific analysis process of the seasonal difference autoregressive moving average model is as follows: Analyze the seasonal pattern of preprocessed real-time grid characteristic data, take seasonal difference to remove the seasonal influence of preprocessed real-time grid characteristic data, and obtain the order, period time interval and residual of seasonal difference autoregressive moving average model, which are used to test the fitting degree of seasonal difference autoregressive moving average model; The specific process of managing and allocating computing resources in constructing the seasonal difference autoregressive moving average model is as follows: Evaluate the available existing computing resources, decompose the computing tasks into multiple subtasks according to the process of building the seasonal difference autoregressive moving average model, manage the task execution according to the time step of the seasonal difference autoregressive moving average model, and use the task scheduling tool to manage the task execution. According to the computing requirements of the task, allocate computing resources to each subtask. For parallel processing tasks, use the parallel computing framework to optimize resource utilization, and evenly distribute them through load balancing technology to obtain the distributed power station computing load evaluation coefficient; The computing resource allocation and optimization module includes a computing task scheduling optimization unit; The computing task scheduling optimization unit: adopts a task scheduling strategy of shortest job first according to the allocation result of computing resources, and reduces the number of parameter combinations according to the system load and resource status; The computing resource allocation and optimization module also includes a model parameter optimization unit; The model parameter optimization unit: finds the parameter combination of the seasonal difference autoregressive moving average model by adopting the Bayesian optimization algorithm, uses the mean square error to evaluate the model performance, obtains the evaluation coefficient of the energy storage device in the energy storage system, and adjusts the parameters of the seasonal difference autoregressive moving average model according to the evaluation coefficient of the energy storage device in the energy storage system; The specific test process for testing the fitting degree of the seasonal difference autoregressive moving average model is as follows: Perform a white noise test on the residuals, use the Durbin-Watson test method to test the correlation of the residuals, divide the preprocessed real-time power grid characteristic data into a training set and a test set, use the training set to train the seasonal difference autoregressive moving average model to obtain the trained seasonal difference autoregressive moving average model, use the trained seasonal difference autoregressive moving average model to perform predictions on the test set, evaluate the prediction error of the seasonal difference autoregressive moving average model, obtain the degree of fit of the seasonal difference autoregressive moving average model based on the prediction error of the seasonal difference autoregressive moving average model, and obtain the predicted correction value of the power generation efficiency in the energy storage system based on the degree of fit of the seasonal difference autoregressive moving average model; The specific optimization process of optimizing the allocation of power generation resources is as follows: Evaluate the pre-processed real-time grid characteristic data, including the load of photovoltaic power generation and the capacity of energy storage equipment, maximize the stability of the load of photovoltaic power generation and the capacity of energy storage equipment, minimize environmental impact, and adjust and optimize the dispatch strategy based on the load evaluation coefficient calculated by the distributed power station, the energy storage equipment evaluation coefficient in the energy storage system, and the predicted correction value of the power generation efficiency in the energy storage system. The adjustment and optimization includes adjusting the angle of photovoltaic panels, replacing photovoltaic panels, and maintaining photovoltaic panels; The specific constraint formula of the power system evaluation coefficient is: ; In the formula, Represents the power system assessment coefficient, which is used to evaluate the comprehensive impact of distributed power station load and power generation efficiency in energy storage system on the power system. represents the load assessment factor for distributed power stations, Represents the evaluation coefficient of energy storage equipment in the energy storage system, It represents the predicted correction value of power generation efficiency in the energy storage system, and e represents a natural constant.

2. A timing simulation system as claimed in claim 1, characterized in that: The specific process of cleaning and feature processing the collected real-time power grid data is as follows: The collected real-time power grid data is cleaned in combination with the historical data of the power grid to obtain preprocessed real-time power grid data, and the state and dynamic change characteristics of the power grid are extracted from the cleaned data to obtain preprocessed real-time power grid characteristic data.

3. A method applied to the timing simulation system as claimed in any one of claims 1 to 2, characterized in that: The following steps are involved: Collect real-time grid data from sensors at various monitoring points of the grid, clean and feature process the collected real-time grid data, and obtain pre-processed real-time grid feature data; A seasonal difference autoregressive moving average model is constructed based on preprocessed real-time power grid characteristic data, the preprocessed real-time power grid characteristic data is divided into a training set and a test set, the training set is used to train the seasonal difference autoregressive moving average model to obtain the trained seasonal difference autoregressive moving average model, the trained seasonal difference autoregressive moving average model is used to make predictions on the test set, and the prediction error of the seasonal difference autoregressive moving average model is evaluated; Manage and allocate computing resources in constructing seasonal difference autoregressive moving average models, and optimize computing methods, computing task scheduling, and model parameters; According to the prediction results of seasonal difference autoregressive moving average model, the allocation of power generation resources is optimized in combination with the grid operation status and power generation plan.