Method for generating optimal scheduling strategy of intelligent park and related device

Abstract

The present application discloses a method for generating an optimized scheduling strategy for an intelligent park and related devices. The method includes: obtaining historical data of wind turbines, photovoltaic units, and conventional units in the intelligent park, constructing a wind power output model and a photovoltaic power output model under different output scenarios based on the historical data, obtaining output data under different output scenarios through Monte Carlo simulation, determining the conditional risk value of the corresponding output scenario according to the output data, calculating the carbon emission cost in combination with the output data of the conventional unit, and finally comprehensively considering the conditional risk value and the carbon emission cost, screening out a target scheduling strategy under preset constraint conditions. The target scheduling strategy is the optimal scheduling strategy, realizing the optimized scheduling of the energy in the intelligent park, making full use of distributed resources, reducing the carbon emission of the power system and improving the economic benefit.

Description

Technical Field

[0001] This application relates to the technical field of optimal scheduling, and particularly to a method for generating an optimal scheduling strategy for an intelligent park and related devices. Background Art

[0002] An intelligent park refers to an area that integrates information technology, energy technology, and modern management concepts. By intelligently integrating and collaboratively managing various resources within the intelligent park, efficient energy utilization, convenient services, and sustainable development can be achieved. Distributed energy resources (DERs), such as wind power generation and photovoltaic power generation, have gradually become a key part of the energy supply system in intelligent parks due to their clean, environmentally friendly, and dispersible layout characteristics.

[0003] In the context of reducing carbon emissions, intelligent parks need to conduct low-carbon optimal scheduling. At the same time, with the increase in energy costs, when conducting low-carbon optimal scheduling, intelligent parks also need to make full use of distributed resources to reduce the operating costs of intelligent parks and improve economic benefits. Although distributed energy can reduce carbon emissions, it has problems such as unstable output, difficulty in prediction, and discontinuous production. These problems are likely to increase the risks in the electricity market transaction and bring difficulties to cost control. Existing optimal scheduling methods are not precise enough in quantifying the fluctuations of wind power and photovoltaic power outputs, resulting in the scheduling plan being unable to fully consider the characteristics of new energy, affecting the reliability and stability of energy supply. Moreover, in terms of carbon emissions and carbon trading management, it is difficult to effectively regulate the use of carbon quotas and unable to effectively improve carbon trading revenues, restricting the development of intelligent parks in the low-carbon economy.

[0004] Therefore, how to obtain an optimal scheduling strategy to make full use of distributed resources, reduce the carbon emissions of the power system, and improve economic benefits. Summary of the Invention

[0005] An embodiment of this application provides a method for generating an optimal scheduling strategy for an intelligent park and related devices. By constructing a wind power output model and a photovoltaic power output model under different output scenarios of the intelligent park, calculating the conditional risk value under different output scenarios, calculating the carbon emission cost based on the output data of conventional units, and comprehensively considering the conditional risk value and the carbon emission cost to obtain an optimal scheduling strategy, this optimal scheduling strategy effectively reduces the risks brought by the fluctuations of wind power and photovoltaic power outputs, reduces carbon emissions, improves energy utilization efficiency, and improves the economic benefits and overall scheduling efficiency of the intelligent park in a multi-market environment.

[0006] In a first aspect, an embodiment of this application provides a method for generating an optimal scheduling strategy for an intelligent park, which is applied to a control device of the intelligent park. The intelligent park includes: wind turbines, photovoltaic units, and conventional units. The method includes:

[0007] Obtain a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbine, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic unit;

[0008] Classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenarios; w is equal to m multiplied by n, and both n and m are positive integers;

[0009] Determine w wind power output models according to the first wind speed data set and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to a wind power output model and a photovoltaic output model;

[0010] Perform Monte Carlo simulation through the w wind power output models and the w photovoltaic output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic output data;

[0011] Determine w target conditional risk values according to the k second wind power output data and the k second photovoltaic output data;

[0012] Obtain the conventional output data of the conventional unit under the w output scenarios to obtain w conventional output data;

[0013] Determine the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data, and the w conventional output data to obtain w carbon emission costs;

[0014] Determine a target scheduling strategy according to the w target conditional risk values and the w carbon emission costs.

[0015] In a second aspect, an optimization scheduling strategy generation device for an intelligent park provided by an embodiment of the present application is applied to a control device of the intelligent park. The intelligent park includes: a wind turbine, a photovoltaic unit, and a conventional unit; the optimization scheduling strategy generation device of the intelligent park includes: a data acquisition module, a scenario classification module, a model training module, a data simulation module, a data processing module, and a strategy generation module, where,

[0016] The data acquisition module is used to obtain a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbine, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic unit;

[0017] The scenario classification module is used to classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenarios; w is equal to m multiplied by n, and both n and m are positive integers;

[0018] The model training module is used to determine w wind power output models according to the first wind speed data set and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to a wind power output model and a photovoltaic output model;

[0019] The data simulation module is used to perform Monte Carlo simulation through the w wind power output models and the w photovoltaic output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic output data;

[0020] The data processing module is used to determine w target conditional risk values according to the k second wind power output data and the k second photovoltaic output data;

[0021] The data acquisition module is further used to obtain the conventional output data of the conventional unit under the w output scenarios to obtain w conventional output data;

[0022] The data processing module is further used to determine the carbon emission cost under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data and the w conventional output data to obtain w carbon emission costs;

[0023] The policy generation module is used to determine a target scheduling policy according to the w target conditional risk values and the w carbon emission costs.

[0024] In a third aspect, an embodiment of the present application provides an electronic device, including: a processor, a memory, a communication interface, and one or more programs, wherein the above one or more programs are stored in the above memory and are configured to be executed by the above processor, and the above programs include instructions for executing the steps in the first aspect of the embodiment of the present application.

[0025] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the above computer-readable storage medium stores a computer program for electronic data exchange, and wherein the above computer program enables a computer to execute some or all of the steps described in the first aspect of the embodiment of the present application.

[0026] Fifth aspect, an embodiment of the present application provides a computer program product, where the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute some or all of the steps described in the first aspect of the embodiment of the present application. The computer program product may be a software installation package.

[0027] It can be seen that by adopting the embodiment of the present application, the following beneficial effects are achieved:

[0028] By implementing the embodiment of the present application, historical data of wind turbines, photovoltaic units, and conventional units in the smart park are obtained, and a wind power output model and a photovoltaic power output model under different output scenarios are constructed based on the historical data. Output data under different output scenarios are obtained through Monte Carlo simulation, the conditional risk value of the corresponding output scenario is determined according to the output data, the carbon emission cost is calculated in combination with the output data of the conventional unit, and finally, considering the conditional risk value and the carbon emission cost comprehensively, the target scheduling strategy is screened under the preset constraint conditions. It can be seen that the target scheduling strategy obtained by comprehensively considering the conditional risk value and the carbon emission cost under different wind power and photovoltaic output scenarios is the optimal scheduling strategy, which can effectively reduce the risk brought by the fluctuations of wind power and photovoltaic output, reduce carbon emissions, improve energy utilization efficiency, and improve the economic benefits and overall scheduling efficiency of the smart park in a multi-market environment. Description of the Drawings

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background art, the drawings required to be used in the embodiments of the present application or the background art will be described below.

[0030] Figure 1 It is a schematic flowchart of a method for generating an optimal scheduling strategy for a smart park provided by an embodiment of the present application;

[0031] Figure 2 It is a schematic diagram of the energy structure of a smart park provided by an embodiment of the present application;

[0032] Figure 3 It is a schematic flowchart of calculating the conditional risk value under an output scenario provided by an embodiment of the present application;

[0033] Figure 4 It is an operation structure diagram of a smart park operator participating in multi-time electricity markets and carbon markets provided by an embodiment of the present application;

[0034] Figure 5 It is an application scenario diagram of a method for generating an optimal scheduling strategy for a smart park provided by an embodiment of the present application;

[0035] Figure 6It is a schematic structural diagram of an optimization scheduling strategy generation system for an intelligent park provided by an embodiment of the present application;

[0036] Figure 7 It is a schematic structural diagram of an optimization scheduling strategy generation device for an intelligent park provided by an embodiment of the present application;

[0037] Figure 8 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0038] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0039] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0040] Referring to "embodiment" herein means that a specific feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0041] The relevant content, concepts, meanings, technical problems, technical solutions, beneficial effects, etc. involved in the embodiments of the present application will be described below.

[0042] Please refer to Figure 1 , Figure 1 It is a schematic flowchart of an optimization scheduling strategy generation method for an intelligent park provided by an embodiment of the present application. This method is applied to a control device of an intelligent park, and the intelligent park includes: wind turbines, photovoltaic units, and conventional units; this method includes but is not limited to the following steps:

[0043] S101. Obtain a first wind speed data set, a first wind power output data set corresponding to the wind turbine, a first solar irradiance data set, and a first photovoltaic power output data set corresponding to the photovoltaic unit within a historical time period.

[0044] Please refer to Figure 2 , Figure 2 which is a schematic diagram of the energy structure of an intelligent park provided by an embodiment of the present application. As Figure 2 shown, the intelligent park includes various energy devices to achieve comprehensive utilization and power supply of energy. The intelligent park includes x conventional units, where x is a positive integer. Conventional units generally refer to traditional thermal power generation units, such as coal-fired and gas-fired units, which can generate electricity stably but have environmental problems such as carbon emissions and can be used to adjust production energy in the energy supply of the intelligent park. The intelligent park includes y wind turbines, where y is a positive integer. Wind turbines are devices that convert wind energy into electrical energy, which is a clean energy source. The power generation is greatly affected by wind speed and direction and is intermittent. The intelligent park includes z photovoltaic units, where z is a positive integer. Photovoltaic units convert solar energy into electrical energy through the photovoltaic effect, which is also a clean energy source. The power generation is unstable due to the limitations of light intensity and duration. The intelligent park includes s energy storage devices, which are used to store electrical energy, charge when the power is excessive, and discharge when the power is insufficient, and can smooth the fluctuations of new energy power generation and adjust the power supply and demand.

[0045] In the embodiment of the present application, a wind turbine refers to a device that converts wind energy into electrical energy. As a clean and renewable energy source, wind energy enables the wind turbine not to generate greenhouse gas emissions during the power generation process, which helps to achieve the low-carbon goal. However, the power output of the wind turbine is greatly affected by wind speed. Unstable wind speed may cause significant fluctuations in the power output of the wind turbine. For example, when the wind speed is low, the wind turbine may not reach the rated power generation, and when the wind speed is too high, the wind turbine may stop operating for equipment safety considerations.

[0046] In the embodiment of the present application, a photovoltaic unit refers to a device that converts solar energy into electrical energy through the photovoltaic effect. The solar energy utilized by the photovoltaic unit is also a clean and renewable energy source. However, the power generation capacity of the photovoltaic unit depends on solar irradiance and light time. At night or on rainy and cloudy days, the solar irradiance is low, and the power output of the photovoltaic unit will decrease significantly or even stop generating electricity. At the same time, factors such as cloud changes will also cause unstable solar irradiance, resulting in fluctuations in the power output of the photovoltaic unit.

[0047] In the embodiments of the present application, a conventional unit refers to traditional thermal power generation equipment, such as gas turbines, coal-fired units, etc. The power generation of conventional units is relatively stable. In the energy supply system of an intelligent park, when wind turbines and photovoltaic units are restricted by natural conditions, such as too low wind speed or insufficient sunlight, conventional units can provide stable power to ensure the stability of power supply in the park and meet the needs of electrical equipment. However, the power generation of conventional units consumes energy such as fossil fuels and generates carbon emissions.

[0048] In an intelligent park, wind turbines and photovoltaic units provide clean energy to help achieve the low-carbon goal, but there is a problem of unstable output. Conventional units can ensure stable power supply, but they will generate carbon emissions. Therefore, by generating an optimal scheduling strategy, the respective advantages can be fully utilized to achieve efficient energy utilization and low-carbon development.

[0049] In a specific embodiment, a first wind speed data set, a first wind power output data set corresponding to a wind turbine, a first solar irradiance data set, and a first photovoltaic output data set corresponding to a photovoltaic unit within a historical time period can be obtained. Since wind speed and solar irradiance directly affect the output of wind turbines and photovoltaic units, collecting data that can reflect the historical operating conditions of wind turbines and photovoltaic units in an intelligent park can be used for model construction, scenario analysis, and risk assessment.

[0050] In a possible embodiment, various monitoring devices will be installed in the intelligent park to measure data within a historical time period. For example, a wind speed sensor is used to measure wind speed to obtain a first wind speed data set, and a solar irradiance sensor collects solar irradiance information to obtain a first solar irradiance data set. At the same time, the monitoring system of the wind turbine itself can record the output at different times to obtain a first wind power output data set, and the monitoring system of the photovoltaic unit itself can record the photovoltaic output data at different times to obtain a first photovoltaic output data set. These data can be stored in the local database of the intelligent park or transmitted to the data center of the intelligent park through a sensor network for storage and management.

[0051] S102. Classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenarios.

[0052] Among them, w is equal to m multiplied by n, where both n and m are positive integers. For example, classification can be performed according to characteristics such as the numerical range and variation trend of wind speed and solar irradiance. The wind speed can be divided into three wind power categories: strong wind, normal wind, and weak wind, and the solar irradiance can be divided into three photovoltaic categories: strong light, normal light, and weak light. Then, the wind power categories and photovoltaic categories are combined. Different combinations of wind speed and solar irradiance will result in different energy output situations, so nine different output scenarios can be formed. For example, the output scenario of strong wind and normal light.

[0053] In a specific embodiment, the first wind speed dataset and the first solar irradiance dataset can be classified according to historical experience and data distribution to obtain m wind power categories and n photovoltaic categories. Combining the m wind power categories and the n photovoltaic categories can obtain w output scenarios, that is, each wind power category can be combined with each photovoltaic category, thus forming m multiplied by n output scenarios.

[0054] By classifying and combining to obtain multiple output scenarios, various possible energy production situations can be considered more comprehensively, so that in the subsequent generation process of the optimization scheduling strategy, targeted analysis can be carried out for different output scenarios, and how to reasonably arrange energy production and distribution in various situations can be planned in advance. For example, in the output scenario of strong wind and strong light, how to make full use of the high output of wind turbines and photovoltaic units, and in the output scenario of weak wind and weak light, how to coordinate conventional units to supplement electricity to ensure the stable power supply and economic benefits of the entire intelligent park.

[0055] S103. Determine w wind power output models according to the first wind speed dataset and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance dataset and the w output scenarios.

[0056] Among them, each output scenario corresponds to a wind power output model and a photovoltaic output model. The output of wind turbines and photovoltaic units in an intelligent park is affected by various factors such as wind speed and solar irradiance. These factors change at any time, resulting in the diversity and uncertainty of energy production. Therefore, different combinations of wind speed and solar irradiance form different output scenarios, and the characteristics of energy production in each scenario are different. By constructing output models for different output scenarios, the energy production situation under different output scenarios can be simulated more accurately.

[0057] In specific embodiments, w wind power output models under w output scenarios can be constructed based on the Weibull distribution according to the first wind speed dataset, and w photovoltaic output models under w output scenarios can be constructed based on the beta distribution according to the first solar irradiance dataset. The wind power output models and photovoltaic output models constructed based on the Weibull distribution and the beta distribution can more accurately describe the uncertainty of wind speed and solar irradiance, and thus more accurately predict the output of wind turbines and photovoltaic units under different output scenarios.

[0058] Optionally, the above step of determining w wind power output models according to the first wind speed dataset and the w output scenarios may specifically include the following steps:

[0059] A301. Select the wind speed data corresponding to the first output scenario from the first wind speed dataset to obtain a training wind speed dataset; the first output scenario is any one of the w output scenarios;

[0060] A302. Obtain the target likelihood function;

[0061] A303. Solve the target likelihood function through the preset maximum likelihood estimation method and the training wind speed dataset to obtain the first shape parameter and the first scale parameter;

[0062] A304. Determine the first probability density function corresponding to the first output scenario according to the first shape parameter and the first scale parameter; the first probability density function is the Weibull probability density function;

[0063] A305. Determine the wind power output model corresponding to the first output scenario according to the preset mapping relationship between wind speed and wind power output and the first probability density function.

[0064] Among them, the preset maximum likelihood estimation method (Maximum Likelihood Estimation, MLE) refers to a preset parameter estimation method, which is used to estimate model parameters through observed data when the probability distribution model form of the known data is known but the parameters are unknown.

[0065] Among them, the preset mapping relationship between wind speed and wind power output refers to the corresponding relationship established in advance between the wind speed and the output power of the wind turbine according to the physical characteristics and operating principles of the wind turbine.

[0066] In the case of weak wind, when the wind speed is lower than the cut-in wind speed, the wind turbine may not be able to start, and the output power is 0 at this time. When the wind speed is lower than the rated wind speed of the wind turbine and not lower than the cut-in wind speed, the wind turbine can operate and generate electricity, but due to the small wind speed, the power increases slowly and the output is low. For example, when the cut-in wind speed is 3 m / s and the wind speed is 3 - 5 m / s, for every 1 m / s increase in the wind speed, the wind power output may only increase by 5% - 10% of the rated power.

[0067] In the case of normal wind, when the wind speed is not lower than the rated wind speed, the relationship between the wind power output and the wind speed is relatively stable, and as the wind speed increases, the wind power output increases significantly. For example, when the wind speed is 8 - 12 m / s, for every 1 m / s increase in the wind speed, the wind power output may increase by 15% - 20% of the rated power. However, in some cases, such as when the wind speed is close to the cut-out wind speed, to protect the safety of the wind turbine equipment, the wind turbine will limit the output power by adjusting the blade angle and other means to keep it near the rated power.

[0068] In the case of strong wind, when the wind speed is not lower than the cut-out wind speed, the excessive wind speed may cause excessive pressure and wear on components such as the blades and gearbox of the wind turbine. At this time, it is necessary to limit the power output to protect the equipment, and the wind turbine will stop running and the output power will drop to 0.

[0069] In a specific embodiment, to construct a wind power output model under a specific output scenario, it is necessary to select the wind speed data under this scenario as training samples in a targeted manner to ensure that the model can accurately reflect the relationship between the wind speed and the wind power output in this scenario. Therefore, the wind speed data corresponding to the first output scenario can be selected from the first wind speed data set to obtain a training wind speed data set, where the first output scenario is any one of the w output scenarios.

[0070] Obtain the target likelihood function. For the Weibull distribution, its probability density function has a specific form. According to this form and the training wind speed data set, the target likelihood function can be constructed. The target likelihood function is as follows:

[0071]

[0072] In the above formula, represents the likelihood function with respect to the first shape parameter and the first scale parameter ; represents the number of training wind speed data in the training wind speed data set; represents the th training wind speed data; represents the quadrature formula.

[0073] Next, the target likelihood function is solved by using the preset maximum likelihood estimation method and the training wind speed data set to obtain the first shape parameter and the first scale parameter. Specifically, since directly maximizing the likelihood function is usually complex, the log-likelihood function can be maximized. Maximizing the log-likelihood function is as follows:

[0074]

[0075] In the above formula, represents the log-likelihood function. By solving the maximum value of the log-likelihood function, the first shape parameter and the first scale parameter can be obtained.

[0076] The first probability density function corresponding to the first output scenario is determined according to the first shape parameter and the first scale parameter, where the first probability density function is the Weibull probability density function. The Weibull probability density function is as follows:

[0077]

[0078] In the above formula, represents the real-time wind speed at time ; represents the probability of wind speed ; represents the first shape parameter; represents the first scale parameter.

[0079] The wind power output model corresponding to the first output scenario is determined according to the preset mapping relationship between wind speed and wind power output and the first probability density function, that is, the probability distribution of wind speed is transformed into the probability distribution of wind power output to obtain the wind power output model for predicting the wind power output in the first output scenario. The wind power output model can accurately predict the wind power output in the first output scenario. For example, when the wind power category in the first output scenario is weak wind, the wind power output probability score in the case of weak wind can be determined through the preset mapping relationship between wind speed and wind power output.

[0080] Optionally, the above step of determining the w photovoltaic output models according to the first solar irradiance data set and the w output scenarios may specifically include the following steps:

[0081] B301. Select the first solar irradiance data corresponding to the second output scenario from the first solar irradiance data set to obtain the training solar irradiance data set; the second output scenario is any one of the w output scenarios;

[0082] B302. Determine the average value and standard deviation of the solar irradiance corresponding to the training solar irradiance data set;

[0083] B303. Determine a second shape parameter and a third shape parameter according to the average solar irradiance and the standard deviation of the solar irradiance;

[0084] B304. Determine a second probability density function corresponding to the second output scenario according to the second shape parameter and the third shape parameter; the second probability density function is a beta distribution probability density function;

[0085] B305. Determine a photovoltaic output model corresponding to the second output scenario according to a preset normalization coefficient and the second probability density function.

[0086] In a specific embodiment, under different output scenarios, the distribution of solar irradiance is different. To accurately reflect the relationship between solar irradiance and photovoltaic output in this scenario, solar irradiance data under a specific scenario can be selected specifically. Therefore, select the first solar irradiance data corresponding to the second output scenario from the first solar irradiance dataset to obtain a training solar irradiance dataset, where the second output scenario is any one of the w output scenarios.

[0087] Next, determine the average solar irradiance and the standard deviation of the solar irradiance corresponding to the training solar irradiance dataset. Among them, the average solar irradiance reflects the overall level of solar irradiance in the second output scenario, while the standard deviation reflects the degree of dispersion of the solar irradiance data relative to the average value.

[0088] Determine a second shape parameter and a third shape parameter according to the average solar irradiance and the standard deviation of the solar irradiance. The calculation formula for the second shape parameter is as follows:

[0089]

[0090] In the above formula, represents the second shape parameter; represents the average solar irradiance; represents the standard deviation of the solar irradiance.

[0091] The calculation formula for the third shape parameter is as follows;

[0092]

[0093] In the above formula, represents the third shape parameter.

[0094] Determine a second probability density function corresponding to the second output scenario according to the second shape parameter and the third shape parameter. Among them, the second probability density function is a beta distribution probability density function, and the beta distribution probability density function is as follows:

[0095]

[0096] In the above formula, represents the solar irradiance at the current moment and the preset maximum irradiance ratio; represents probability; represents the gamma function; represents the second shape parameter; represents the third shape parameter.

[0097] Determine the photovoltaic output power model corresponding to the second output scenario according to the preset normalization coefficient and the second probability density function. The photovoltaic output power model indicates that the photovoltaic output power also follows a beta distribution. Among them, the preset normalization coefficient refers to the coefficient preset to ensure the correct normalization of the probability density function. Among them, the photovoltaic output power model is as follows:

[0098]

[0099] In the above formula, represents the output power of the photovoltaic, that is, the photovoltaic output; represents the photovoltaic output probability; is the preset normalization coefficient, usually representing the coefficient determined according to the number of photovoltaic panels in the photovoltaic unit. The more the number of its panels, the stronger the power generation ability of the entire photovoltaic system under the same lighting conditions, and the larger this coefficient; is the preset normalization coefficient, usually representing the coefficient determined according to the area of the photovoltaic panels in the photovoltaic unit. The larger the area of the panels, the more sunlight radiation can be received, and thus more electric energy can be generated; is the preset normalization coefficient, usually representing the coefficient determined according to the photoelectric conversion efficiency of the photovoltaic panels in the photovoltaic unit. It reflects the ability of the photovoltaic panels to convert solar energy into electric energy. The higher the conversion efficiency, the more electric energy is generated under the same lighting; represents the maximum output power of the photovoltaic.

[0100] The photovoltaic output power model constructed by combining the hardware parameters of the photovoltaic unit, such as the number of panels, area, conversion efficiency, and the statistical characteristics of the beta distribution, can accurately describe the probability distribution of the photovoltaic output, provide accurate information for the energy scheduling of the smart park, and predict the output of the photovoltaic unit according to this model under different output scenarios, reasonably arrange energy production, storage and distribution, improve energy utilization efficiency, reduce energy costs, and at the same time help to ensure the stability and reliability of the power supply in the park.

[0101] S104. Perform Monte Carlo simulation using the w wind power output models and the w photovoltaic power output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic power output data.

[0102] In the embodiments of the present application, Monte Carlo simulation is a numerical method for solving mathematical, physical, engineering, etc. problems through random sampling. By using the selected wind power output model and photovoltaic power output model, multiple random simulation calculations can be performed to obtain multiple second wind power output data and multiple second photovoltaic power output data.

[0103] In a specific embodiment, for each output scenario, select the wind power output model corresponding to this scenario from the w wind power output models, and at the same time select the corresponding photovoltaic power output model from the w photovoltaic power output models, and perform Monte Carlo simulation using the selected wind power output model and photovoltaic power output model to obtain k second wind power output data and k second photovoltaic power output data. Among them, the larger k is, the closer the simulation result is to the real situation, but the calculation amount will also increase accordingly.

[0104] Wind power output and photovoltaic power output are greatly affected by natural factors and have high uncertainty. Through Monte Carlo simulation, multiple random simulations are performed based on probability distributions. By simulating the output situations under different output scenarios, the change ranges and possibilities of wind power output and photovoltaic power output under different output scenarios can be more comprehensively shown, so as to accurately simulate the energy production situation corresponding to the output scenario.

[0105] Among them, k is equal to w multiplied by e, where e is an integer greater than or equal to 1, and each output scenario corresponds to e second wind power output data and e second photovoltaic power output data.

[0106] Optionally, the above step of performing Monte Carlo simulation using the w wind power output models and the w photovoltaic power output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic power output data may specifically include the following steps:

[0107] A401. Select the wind power output model corresponding to the target output scenario from the w wind power output models to obtain the target wind power output model, and select the photovoltaic power output model corresponding to the target output scenario from the w photovoltaic power output models to obtain the target photovoltaic power output model; the target output scenario is any one of the w output scenarios;

[0108] A402. Perform Monte Carlo simulation using the target wind power output model and the target photovoltaic power output model to obtain x second wind power output data and x second photovoltaic power output data; x is an integer greater than or equal to e;

[0109] A403. Remove the outliers from the x second wind power output data to obtain e second wind power output data;

[0110] A404. Remove the outliers from the x second photovoltaic output data to obtain e second photovoltaic output data.

[0111] In a specific embodiment, a wind power output model corresponding to the target output scenario is selected from the w wind power output models to obtain a target wind power output model, and a photovoltaic output model corresponding to the target output scenario is selected from the w photovoltaic output models to obtain a target photovoltaic output model, where the target output scenario is any one of the w output scenarios. For example, if the target output scenario is a strong wind and strong light scenario, a wind power output model and a photovoltaic output model constructed based on strong wind speed data and strong light solar irradiance data are selected. By selecting an appropriate target output model, the Monte Carlo simulation can be made more in line with the actual situation and the accuracy of the simulation results can be improved.

[0112] According to the process of the Monte Carlo simulation, in the target wind power output model, according to the Weibull probability density function on which the target wind power output model is based, a wind speed value is randomly generated, and then the wind power output is calculated according to the preset mapping relationship between the wind speed and the wind power output. In the target photovoltaic output model, a solar irradiance value is randomly generated according to the beta distribution probability density function, and the photovoltaic output is calculated by combining the preset normalization coefficient. By performing the simulation, x second wind power output data and x second photovoltaic output data can be obtained, where x is an integer greater than or equal to e.

[0113] During the simulation process, due to the randomness of random sampling, some outliers that deviate from the normal range may be generated. These outliers may affect the accuracy of subsequent data analysis. Therefore, it is necessary to remove them to obtain more reliable wind power output data. By removing the outliers from the x second wind power output data, e second wind power output data can be obtained. At the same time, by removing the outliers from the x second photovoltaic output data, e second photovoltaic output data can be obtained. The e second wind power output data and the e second photovoltaic output data after removing the outliers can better represent the true distribution of wind power output and photovoltaic output under the target output scenario, and can improve the accuracy of subsequent value-at-risk assessment and scheduling strategy formulation.

[0114] Optionally, the above step of removing the outliers from the x second wind power output data to obtain e second wind power output data may specifically include the following steps:

[0115] B401. Select x target wind power output data corresponding to the target output scenario from the first wind power output data set;

[0116] B402. Determine a target Pearson correlation coefficient and a target mean square error based on the x second wind power output data and the x target wind power output data;

[0117] B403. Determine a first weight coefficient corresponding to the target Pearson correlation coefficient and a second weight coefficient corresponding to the target mean square error; the sum of the first weight coefficient and the second weight coefficient is 1;

[0118] B404. Determine a target comprehensive score based on the target Pearson correlation coefficient, the first weight coefficient, the target mean square error, and the second weight coefficient;

[0119] B405. If the target comprehensive score is less than or equal to a preset comprehensive score threshold, determine the wind power output difference between the x second wind power output data and the x target wind power output data to obtain x wind power output differences;

[0120] B406. Select the wind power output differences less than a preset difference threshold from the x wind power output differences to obtain e wind power output differences;

[0121] B407. Select the second wind power output data corresponding to the e wind power output differences from the x second wind power output data to obtain the e second wind power output data;

[0122] B408. If the target comprehensive score is greater than the preset comprehensive score threshold, obtain the average value of the x second wind power output data, and select e second wind power output data with the smallest absolute value of the difference from the x second wind power output data.

[0123] Among them, the preset comprehensive score threshold refers to a boundary value set when evaluating the matching degree between the second wind power output data and the target wind power output data, and is used to judge whether the similarity between the simulated data and the actual data reaches an acceptable standard, and can be specifically set according to historical data experience and actual application requirements.

[0124] In a specific embodiment, the first wind power output data set is the actually recorded wind power output data. x data matching the target output scenario are screened out from the first wind power output data set to obtain x target wind power output data, which are used to evaluate and compare with the second wind power output data obtained by Monte Carlo simulation to judge whether the simulation result is reasonable.

[0125] Determine the target Pearson correlation coefficient and the target mean squared error based on x second wind power output data and x target wind power output data. The target Pearson correlation coefficient is used to measure the linear correlation degree between the simulated data and the actual data, and the mean squared error is an index to measure the difference size between the simulated data and the actual data. To comprehensively consider the influence of the target Pearson correlation coefficient and the target mean squared error on the evaluation of the simulated data, the first weight coefficient corresponding to the target Pearson correlation coefficient and the second weight coefficient corresponding to the target mean squared error can be determined, where the sum of the first weight coefficient and the second weight coefficient is 1.

[0126] Perform weighted summation on the target Pearson correlation coefficient and the target mean squared error according to their respective weights to obtain the target comprehensive score. The target comprehensive score is used to evaluate the matching degree between the second wind power output data and the target wind power output data, and its calculation formula is as follows:

[0127]

[0128] In the above formula, represents the target comprehensive score; represents the first weight coefficient, which represents the weight of the Pearson correlation coefficient in the comprehensive score; represents the target Pearson correlation coefficient; represents the second weight coefficient, which represents the weight of the mean squared error in the comprehensive score; represents the target mean squared error.

[0129] If the target comprehensive score is less than or equal to the preset comprehensive score threshold, it indicates that the matching degree between the second wind power output data and the target wind power output data is poor, and outliers can be further screened out. Determine the wind power output difference between the x second wind power output data and the x target wind power output data to obtain x wind power output differences. Select the wind power output differences less than the preset difference threshold from the x wind power output differences to obtain e wind power output differences, and select the second wind power output data corresponding to the e wind power output differences from the x second wind power output data to obtain e second wind power output data. Among them, the larger the wind power output difference, the greater the deviation between the second wind power output data and the target wind power output data. The preset difference threshold is a value preset to determine whether the wind power output difference is within an acceptable range. By setting the difference threshold for screening, data with too large a difference from the actual data, that is, outliers, can be removed.

[0130] When the target comprehensive score is greater than the preset comprehensive score threshold, it indicates that the matching degree between the second wind power output data and the target wind power output data is good. At this time, the data with the smallest difference from the average value can be directly selected as the second wind power output data with the average value of the second wind power output data as the center. Specifically, the average value of x second wind power output data can be obtained, and e second wind power output data with the smallest absolute value of the difference from the average value can be selected from the x second wind power output data.

[0131] Please refer to Figure 3 , Figure 3 FIG. is a schematic diagram of the calculation process of conditional value at risk in an output scenario provided by an embodiment of the present application. As shown in the figure, by collecting historical data of an intelligent park, including wind speed and solar irradiance data, and classifying them according to intensity, multiple output scenarios are obtained. Then, using the Monte Carlo simulation method, for each output scenario, considering the uncertainty of new energy (wind power, photovoltaic) output, random simulation is carried out to obtain new energy output data under multiple output scenarios. For each simulated output scenario, the comprehensive score of wind power and photovoltaic output is calculated according to specific rules to comprehensively reflect the new energy power generation performance in this scenario. According to the comprehensive score, abnormal simulation values are identified and removed to make the simulation data more representative, and the final simulation data is obtained.

[0132] Check whether the number of simulations is less than the set number. The set number represents the predetermined number of simulations. The more the number of times, the closer the simulation result can approach the true new energy output distribution, and the more accurately it can capture the uncertainty characteristics of wind power and photovoltaic output, improving the simulation accuracy. If the number of simulations is less than the set number, return to continue the simulation. Otherwise, based on the final simulation data, calculate the conditional value at risk corresponding to different output scenarios to measure the potential risk of the intelligent park under the uncertainty of new energy output.

[0133] S105. Determine w target conditional value at risks according to the k second wind power output data and the k second photovoltaic output data.

[0134] In the embodiment of the present application, the value at risk (VaR) refers to At the confidence level, the maximum possible loss caused by the uncertainty of wind and light output in a specific future period. For example, at a 95% confidence level, if VaR is 100, there is a 95% probability that the loss will not exceed 100, that is, there is a 5% probability that the loss will exceed 100. However, VaR cannot well reflect the impact of extreme loss events because it ignores tail losses. In contrast, the conditional value at risk (CVaR) represents when the confidence level is greater than The average loss at that time. Therefore, CVaR can better measure the tail loss. For example, at a 95% confidence level, in addition to knowing that there is a 95% probability that the loss does not exceed 100, it can also determine the average loss when the loss exceeds 100. It pays more attention to the tail of the loss distribution and reflects the risk under extreme events.

[0135] In a specific embodiment, the uncertainty of wind-solar power output can be quantified based on k second wind power output data and k second photovoltaic power output data to obtain w target conditional risk values corresponding to w power output scenarios. By quantifying the energy supply risk under different power output scenarios, the potential loss caused by insufficient power supply under various possible power output situations can be understood through the target conditional risk values, so as to better evaluate the risk status of the energy system, and then a more reasonable optimal dispatching strategy can be formulated according to the conditional risk values, improving the stability and reliability of the energy system and reducing the operation cost and risk.

[0136] Optionally, the above step of determining w target conditional risk values according to the k second wind power output data and the k second photovoltaic power output data may specifically include the following steps:

[0137] A501. Obtain e second wind power output data and e second photovoltaic power output data corresponding to the target power output scenario;

[0138] A502. Select the actual wind power output data corresponding to the target power output scenario from the first wind power output data set, and select the actual photovoltaic power output data corresponding to the target power output scenario from the first photovoltaic power output data set;

[0139] A503. Determine e penalty costs according to the e second wind power output data, the e second photovoltaic power output data, the actual wind power output data, and the actual photovoltaic power output data;

[0140] A504. Determine the first risk value according to the e penalty costs and the preset confidence level;

[0141] A505. Determine y first penalty costs according to the e penalty costs and the first risk value; y is a positive integer less than e;

[0142] A506. Determine the operation of the y first penalty costs and the first risk value to obtain y risk differences;

[0143] A507. Determine the target conditional risk value according to the first risk value and the y risk differences.

[0144] In specific embodiments, e second wind power output data and e second photovoltaic output data corresponding to the target output scenario can be obtained from the k second wind power output data and the k second photovoltaic output data, and the actual wind power output data corresponding to the target output scenario can be selected from the first wind power output data set, and the actual photovoltaic output data corresponding to the target output scenario can be selected from the first photovoltaic output data set.

[0145] Next, e penalty costs are determined according to the e second wind power output data, the e second photovoltaic output data, the actual wind power output data, and the actual photovoltaic output data. The penalty costs are used to measure the deviation degree between the simulated wind power and photovoltaic output data and the actual data. The greater the deviation, the higher the penalty cost, so as to reflect the potential risk brought by the inconsistency between the simulation result and the actual situation.

[0146] The first risk value is determined according to the e penalty costs and the preset confidence level, and its calculation formula is as follows:

[0147]

[0148] In the above formula, represents the first risk value corresponding to the target output scenario; represents the th penalty cost value after sorting all the penalty costs from small to large, is rounded up, so represents the total number of samples of the e penalty costs multiplied by the preset confidence level , and the rounded-up index value is obtained, and the penalty cost selected from the sorted e penalty costs according to the index value.

[0149] Among the e penalty costs, the penalty costs exceeding the first risk value are screened out to obtain y first penalty costs, where y is a positive integer less than e. The first penalty costs represent extreme cases outside the preset confidence level. The difference between each first penalty cost and the first risk value among the y first penalty costs is calculated to obtain y risk differences. The risk differences reflect the additional losses exceeding the first risk value. Then, the average value of the y risk differences is calculated first, and then the average value is added to the first risk value to obtain the target conditional risk value. The calculation formula of the target conditional risk value is as follows:

[0150]

[0151] In the above formula, represents the target conditional risk value corresponding to the target output scenario; represents the first risk value; represents the total number of samples of the e penalty costs; represents the confidence level; represents the th first penalty cost.

[0152] Optionally, the above steps of determining e penalty costs according to the e second wind power output data, the e second photovoltaic power output data, the actual wind power output data, and the actual photovoltaic power output data may specifically include the following steps:

[0153] B501. Determine the total wind and photovoltaic power output according to the e second wind power output data and the e second photovoltaic power output data, and obtain e total wind and photovoltaic power outputs;

[0154] B502. Determine the target total wind and photovoltaic power output according to the actual wind power output data and the actual photovoltaic power output data;

[0155] B503. Determine the difference between each total wind and photovoltaic power output in the e total wind and photovoltaic power outputs and the target total wind and photovoltaic power output, and obtain e output total differences;

[0156] B504. Determine the e penalty costs according to the e output total differences, a preset first penalty coefficient, and a preset second penalty coefficient; wherein, if the output total difference i is greater than or equal to 0, then determine the penalty cost i according to the output total difference i and the preset first penalty coefficient; if the output total difference i is less than 0, then determine the penalty cost i according to the output total difference i and the preset second penalty coefficient; the output total difference i is any one of the e output total differences.

[0157] Among them, the preset first penalty coefficient is used to calculate the penalty cost when the simulated total wind and photovoltaic power output is greater than or equal to the actual value. In the context of energy management in an intelligent park, although the excess new energy power generation can meet the power demand, there will also be some problems. For example, if the excess electric energy cannot be stored or utilized effectively in time, it may need to be sold at a low price, resulting in economic losses, and storing the excess electric energy will generate storage costs, including the purchase, maintenance of energy storage equipment, and losses during the electric energy storage process, etc. The preset first penalty coefficient represents a comprehensive quantification of potential economic losses, and its specific value setting can be based on the actual situation of the intelligent park, such as factors like the cost of energy storage equipment, power market price fluctuations, and electric energy storage efficiency. For example, if the cost of energy storage equipment in the intelligent park is high and the electricity price is low when the local power market supply exceeds demand, then the preset first penalty coefficient should be adjusted upward accordingly to more accurately reflect the economic impact of power generation surplus.

[0158] The preset second penalty coefficient is used to calculate the penalty cost when the total simulated wind and photovoltaic power output is less than the actual value. When the new energy generation is insufficient and there is a power supply gap in the smart park, in order to ensure stable power supply, it is necessary to purchase electricity at a high price from the external power grid, thereby increasing the electricity cost. At the same time, the shortage of power supply may affect the normal production and operation of enterprises in the park, causing indirect economic losses such as production interruption and equipment damage. The preset second penalty coefficient represents a comprehensive quantification of the direct and indirect economic losses caused by insufficient power generation. The specific value setting can consider factors such as the external power purchase price and the average economic loss of park enterprises due to power outages. For example, if the power purchase price of the power grid around the smart park is high, and the smart park is sensitive to power interruption and has large power outage losses, the preset second penalty coefficient can be set higher to accurately reflect the losses caused by insufficient power generation.

[0159] In a specific embodiment, the total wind and photovoltaic power output is determined according to e second wind power output data and e second photovoltaic power output data, and e total wind and photovoltaic power output values are obtained. In the energy supply scenario of the smart park, wind power and photovoltaic power jointly provide power for the smart park. Therefore, by combining the wind power output and the photovoltaic power output for calculation, the total new energy power generation output under each simulation sample can be obtained. The target total wind and photovoltaic power output is determined according to the actual wind power output data and the actual photovoltaic power output data.

[0160] The difference between each total wind and photovoltaic power output value in the e total wind and photovoltaic power output values and the target total wind and photovoltaic power output value is determined, and e output value differences are obtained. The output value difference reflects the difference degree between the simulated total new energy power generation and the actual total power generation. The larger the difference, the greater the deviation between the simulated value and the actual value.

[0161] According to the e output value differences, the preset first penalty coefficient, and the preset second penalty coefficient, e penalty costs are determined. Among them, if the output value difference i is greater than or equal to 0, the penalty cost i is determined according to the output value difference i and the preset first penalty coefficient; if the output value difference i is less than 0, the penalty cost i is determined according to the output value difference i and the preset second penalty coefficient. The output value difference i is any one of the e output value differences. The calculation formula of the penalty cost is as follows:

[0162]

[0163] In the above formula, represents the wind and photovoltaic simulated output in the output scenario ; represents the wind and photovoltaic actual output in the output scenario ; represents the th penalty cost, that is, the penalty cost corresponding to the output value difference ; represents a preset first penalty coefficient, which specifically represents the penalty coefficient for over-high output. represents a preset second penalty coefficient, which specifically represents the penalty coefficient for under-low output.

[0164] In actual energy management, the impacts and costs brought about by over-generation and under-generation of new energy power are different. For example, over-generation may involve the cost of power storage or the loss of selling excess power at a low price, while under-generation may lead to a shortage of power supply in the park, and it is necessary to purchase power at a high price from the outside. By determining the penalty cost and converting the deviation between the simulated value and the actual value into a quantifiable cost index, the potential economic impact caused by the inconsistency between the simulated result and the actual situation can be intuitively reflected.

[0165] S106. Obtain the conventional output data of the conventional unit under the w output scenarios to obtain w pieces of conventional output data.

[0166] The energy supply of an intelligent park is usually jointly affected by wind turbines, photovoltaic units and conventional units. The output of wind turbines and photovoltaic units is greatly affected by natural conditions, with uncertainty and volatility. The power generation of conventional units is relatively stable and can play a supplementary role when the new energy power generation is insufficient. By obtaining the conventional output data of conventional units under different output scenarios, the energy supply capacity of the intelligent park under different output scenarios can be comprehensively understood, so as to formulate a more reasonable optimal dispatching strategy to ensure the stability and reliability of power supply.

[0167] In a specific embodiment, the data of the unit can be recorded in real time by collecting the monitoring equipment equipped with the conventional unit itself. The monitoring equipment can record the operation data such as the power generation power of the unit in real time. By obtaining the conventional output data of the conventional unit under the w output scenarios, w pieces of conventional output data can be obtained. When calculating the carbon emission cost, evaluating the overall performance of the energy system, and formulating an optimal dispatching strategy, the power generation situation of the conventional unit needs to be considered. For example, when evaluating the carbon emission cost, the carbon emissions generated by different conventional units are different. Through the conventional output data, the carbon emissions of the conventional unit in each scenario can be accurately calculated, and then the carbon emission cost of the entire park can be obtained.

[0168] S107. Determine the carbon emission cost under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data and the w conventional output data to obtain w carbon emission costs.

[0169] In the energy supply system of an intelligent park, carbon emissions mainly come from the power generation process of conventional units. Since wind turbines and photovoltaic units hardly produce carbon emissions during power generation, the changes in wind power and photovoltaic power output will affect the power generation power of conventional units, and thus affect the total carbon emissions. For example, when the wind power and photovoltaic power output are high, the power generation power of conventional units may decrease, thereby reducing carbon emissions. Otherwise, when the wind power and photovoltaic power output are insufficient, conventional units need to increase the power generation power to meet the power demand of the park, and the carbon emissions will also increase accordingly.

[0170] In a specific embodiment, the total carbon emissions in w output scenarios can be determined according to k second wind power output data, k second photovoltaic power output data, and w conventional power output data, and the carbon emission cost can be calculated based on the total carbon emissions to obtain w carbon emission costs. When calculating the total carbon emissions, since the increase in wind power and photovoltaic power output will reduce the power generation demand of conventional units, thereby reducing carbon emissions, therefore, by analyzing k second wind power output data and k second photovoltaic power output data, the total carbon emissions can be adjusted.

[0171] By calculating the carbon emission costs in different output scenarios, the environmental costs generated by the energy supply in the intelligent park in different output scenarios can be intuitively reflected. By comparing the carbon emission costs in different output scenarios, the impact of different energy scheduling strategies on the environment can be evaluated, so as to select a strategy with a lower carbon emission cost to achieve the goal of energy conservation and emission reduction. At the same time, the calculation of carbon emission costs can also help the park managers understand the operating costs of the energy system, reasonably arrange energy production and consumption, improve energy utilization efficiency, and reduce operating costs.

[0172] Optionally, the above step of determining the carbon emission costs in the w output scenarios according to the k second wind power output data, the k second photovoltaic power output data, and the w conventional power output data to obtain w carbon emission costs may specifically include the following steps:

[0173] A701. Obtain w total load values corresponding to the intelligent park;

[0174] A702. Determine w first carbon emissions according to a preset power supply carbon emission quota benchmark value and the w total load values;

[0175] A703. Determine w second carbon emissions according to the w conventional power output data;

[0176] A704. Determine w third carbon emissions according to the k second wind power output data and the k second photovoltaic power output data;

[0177] A705. Determine w target carbon emissions according to the w first carbon emissions, the w second carbon emissions, and the w third carbon emissions;

[0178] A706. Determine the w carbon emission costs according to the mapping relationship between the w target carbon emissions and the preset carbon emissions and costs.

[0179] In a specific embodiment, in different output scenarios of the intelligent park, its total power load will be different. Specifically, the power consumption data in different output scenarios can be collected through the power monitoring equipment and relevant data recording systems in the intelligent park, and the total load value corresponding to each output scenario can be obtained through sorting and statistics. Determine w first carbon emissions according to the preset power supply carbon emission quota benchmark value and the w total load values. Among them, the preset power supply carbon emission quota benchmark value is a preset standard used to measure the allowable carbon emissions per unit power supply. Specifically, the total load value of each output scenario can be multiplied by the preset power supply carbon emission quota benchmark value to obtain the corresponding first carbon emission. The calculation formula of the first carbon emission is as follows:

[0180]

[0181] In the above formula, represents the first carbon emission; represents the preset power supply carbon emission quota benchmark value, and the unit is usually tons per megawatt-hour; represents the output scenario the total load value under.

[0182] Calculating the first carbon emission can be used to judge whether the requirements of the carbon emission quota are met in the current output scenario. If the actual carbon emission exceeds the first carbon emission, it may be necessary to purchase additional carbon emission rights or take measures to reduce carbon emissions, which will increase costs.

[0183] Determine w second carbon emissions according to the w normal output data. The calculation formula of the second carbon emission is as follows:

[0184]

[0185] In the above formula, represents the second carbon emission; represents the number of conventional units in the intelligent park; , , are the power supply carbon emission coefficients of the conventional units respectively. Among them, represents the fixed carbon emission coefficient of the conventional unit , which is related to factors such as the type, manufacturing process, and initial installation of the conventional unit, and does not change with the change of the power generation power of the conventional unit, indicating a part of the carbon emissions generated when the conventional unit does not generate electricity, represents the conventional unit The carbon emission coefficient related to the first-order term of power generation, which represents the part where the carbon emission is linearly related to the power generation. Indicates a conventional unit The carbon emission coefficient related to the second-order term of power generation, which represents the part where the carbon emission is related to the square of the power generation, meaning that as the power generation increases, the growth rate of this part of the carbon emission will be faster than the linear part. Indicates a conventional unit Under the output scenario The output.

[0186] When the wind turbines and photovoltaic units in the smart park transmit power to the superior power grid through the tie line, it actually reduces the operation of thermal power in the region and the superior power grid area, thereby reducing carbon emissions. Therefore, according to k second wind power output data and k second photovoltaic power output data, w third carbon emissions are determined, and the calculation formula for the third carbon emission is as follows:

[0187]

[0188] In the above formula, Represents the third carbon emission; , Represent the numbers of wind turbines and photovoltaic units in the smart park respectively; , Represent the wind turbines And the photovoltaic units Under the output scenario The output; , Represent the wind turbines And the photovoltaic units The corresponding carbon emission coefficients, which represent the carbon emissions that can be reduced per unit of electricity transmitted by the wind turbines and photovoltaic units.

[0189] Based on w first carbon emissions, w second carbon emissions and w third carbon emissions, w target carbon emissions are determined, and the calculation formula for the target carbon emission is as follows:

[0190]

[0191] In the above formula, Represents the target carbon emission; Represents the first carbon emission; Represents the second carbon emission; Represents the third carbon emission.

[0192] When the target carbon emissions are greater than 0, w carbon emission costs can be determined according to the w target carbon emissions and the preset mapping relationship between carbon emissions and costs. The preset mapping relationship between carbon emissions and costs is a pre-set rule for converting the total carbon emissions into corresponding costs, and the preset mapping relationship between carbon emissions and costs is as follows:

[0193]

[0194] In the above formula, represents the carbon emission cost; represents the base price of carbon emission rights trading; represents the target carbon emissions; represents the unit quota step interval of carbon emissions, which is used to divide different intervals of carbon emissions, and its value is pre-set according to the rules of the carbon emission rights trading market and actual needs; represents the price growth rate.

[0195] When the target carbon emissions are not greater than 0, it means that the smart park has surplus carbon emission rights. At this time, the smart park can sell these redundant carbon emission rights in the carbon emission rights trading market to obtain economic benefits and further reduce carbon emissions.

[0196] Determining the carbon emission costs corresponding to different output scenarios can provide an economic reference for the energy management of the smart park. At the same time, according to the carbon emission costs under different output scenarios, the energy dispatch strategy can be adjusted to preferentially select scenarios and energy combinations with lower carbon emission costs.

[0197] S108. Determine the target dispatch strategy according to the w target conditional risk values and the w carbon emission costs.

[0198] In the embodiments of the present application, by comprehensively considering the target conditional risk values and carbon emission costs under different output scenarios, the optimal optimization dispatch strategy is determined to balance risks and costs and achieve efficient, stable and sustainable supply of energy in the smart park.

[0199] In specific embodiments, multiple scheduling strategies can be pre-generated. Different scheduling strategies include the operating parameters of various energy devices, the charge and discharge plans of energy storage devices, and the interaction strategies with the power grid under different output scenarios. For example, in the output scenario of strong wind and strong light, the outputs of wind turbines and photovoltaic units are relatively high, and the excess electricity can be charged into the energy storage device. In the scenario of weak wind and weak light, the conventional unit generates electricity according to the load demand, and at the same time, the energy storage device discharges to supplement the power gap. The risks and carbon emissions corresponding to different output scenarios are different. Therefore, among the multiple scheduling strategies, the target scheduling strategy is determined according to the target conditional risk value and carbon emission cost corresponding to different scheduling strategies. The target scheduling strategy is the optimal scheduling strategy among the multiple scheduling strategies, which can control the energy cost while reducing risks and improve the economic benefits of the smart park.

[0200] Optionally, the above step of determining the target scheduling strategy according to the w target conditional risk values and the w carbon emission costs may specifically include the following steps:

[0201] A801. Generate s scheduling strategies according to preset constraint conditions;

[0202] A802. Determine the output scenario corresponding to each scheduling strategy among the s scheduling strategies to obtain s output scenarios;

[0203] A803. Select the target conditional risk values and carbon emission costs corresponding to the s output scenarios from the w target conditional risk values and the w carbon emission costs to obtain s target conditional risk values and s carbon emission costs;

[0204] A804. Determine s strategy scores according to the s target conditional risk values, the s carbon emission costs, and a preset optimal scheduling model for the smart park;

[0205] A805. Select the highest strategy score from the s strategy scores to obtain the highest strategy score;

[0206] A806. Take the scheduling strategy corresponding to the highest strategy score as the target scheduling strategy.

[0207] Please refer to Figure 4 , Figure 4It is an operation structure diagram of an intelligent park operator participating in multi-time electricity markets and carbon markets provided by an embodiment of the present application. As shown in the figure, the operation structure includes main equipment such as conventional units, wind turbine units, photovoltaic units, and energy storage devices. Among them, the conventional unit is a device that can use natural gas, etc. as fuel and generate electricity through combustion work. It can be flexibly started and stopped and adjust the power generation power according to the electricity demand of the park, and serve as a supplementary power source when new energy power generation is insufficient to ensure stable power supply. However, carbon emissions will be generated during the power generation process. The wind turbine unit is a device that uses wind energy to generate electricity and converts wind energy into electrical energy to provide a clean energy source for the park's power supply. The photovoltaic unit is a device that converts solar energy into electrical energy through the photovoltaic effect of solar energy and depends on the light intensity and sunshine duration. The wind turbine unit and the photovoltaic unit are greatly affected by the environment and have intermittency and volatility. The energy storage device is a device used to store excess electrical energy. It can be charged during periods of excess power supply (such as when wind power and photovoltaic power are booming) and discharged during peak power demand or when new energy power generation is insufficient, playing a role in smoothing power fluctuations, regulating power supply and demand, and improving the stability and reliability of power supply.

[0208] In terms of participating in the electricity market, the intelligent park operator can sign electricity trading contracts in advance in the futures market to lock in the electricity trading price and quantity for a certain future period, plan the electricity sales or procurement strategy in advance, and avoid the risk of market price fluctuations. At the same time, based on information such as the predicted electricity demand and new energy power generation for the next day, the operator can bid for the purchase and sale of electricity in the day-ahead market to determine the electricity trading plan for each period of the next day and optimize the allocation of electricity resources. During the real-time operation of electricity, the intelligent park operator can adjust the real-time purchase and sale of electricity in the real-time market according to the actual electricity supply and demand deviation, such as the actual new energy power generation not matching the prediction and sudden changes in the park load, etc., to balance the electricity supply and demand and ensure real-time balance and stable supply of electricity.

[0209] During the operation process, the intelligent park operator generates carbon emissions due to the power generation of conventional units, etc., and needs to participate in the carbon market. It can meet the emission requirements by purchasing carbon emission rights, or reduce carbon emissions by optimizing the energy structure, such as increasing the proportion of wind power and photovoltaic power and improving the energy utilization efficiency, etc. It can also sell the excess carbon emission rights to make a profit and achieve a balance between economic and environmental benefits.

[0210] The operating states of the conventional unit, wind turbine unit, photovoltaic unit, and energy storage device, such as power generation power and energy storage power, etc., can be fed back to the park operator. Based on this information and the market price signals from the futures, day-ahead, and real-time markets, the operator can formulate the dispatching strategies for each device, participate in electricity market transactions, and at the same time, consider the carbon emission situation and participate in carbon market transactions to achieve coordinated optimization operation in multiple markets, ensuring stable power supply and cost optimization in the park.

[0211] Among them, the preset constraints include energy storage device charge and discharge constraints, futures market output constraints, day-ahead market bidding constraints, time-of-use power balance constraints in intelligent parks, conditional value-at-risk constraints, etc.

[0212] Among them, the energy storage device charge and discharge constraints of the energy storage device include the following constraint conditions:

[0213]

[0214] The above formula represents the device capacity constraint in the energy storage device charge and discharge constraints, which means that before the moment, the energy state of the energy storage device needs to meet a certain range, that is, the energy state needs to be greater than or equal to a lower limit value calculated according to the discharge power and discharge efficiency constant of the energy storage device, and the energy state is less than or equal to the maximum capacity of the energy storage device .

[0215]

[0216] The above formula represents the SOC constraint in the energy storage device charge and discharge constraints, which means that the state of charge of the energy storage device under the output scenario at the moment must be between the minimum value and the maximum value to ensure that the state of charge of the energy storage device is within a reasonable range and avoid damage to the device caused by overcharging or over-discharging.

[0217] Among them, the formula for calculating the state of charge is as follows:

[0218]

[0219] In the above formula, the at the moment of the energy storage device can be calculated through its corresponding energy state and the maximum capacity of the energy storage device .

[0220]

[0221] The above formula represents the charge and discharge power constraint in the energy storage device charge and discharge constraints, which limits the at ​Charging power at a moment and discharging power The value ranges of charging and discharging power should both be greater than or equal to 0 and should not exceed the maximum charging power allowed by the device and the maximum discharging power .

[0222]

[0223] The above formula represents the energy state update formula in the charge-discharge constraints of the energy storage device, which represents the energy storage device at the energy state at a moment is updated according to the energy state at the previous moment , and the charging electricity at a moment and the discharging electricity .

[0224]

[0225] The above formula represents the charging electricity calculation formula in the charge-discharge constraints of the energy storage device, which represents the energy storage device at the charging electricity at a moment is obtained by multiplying the charging power by the efficiency constant of energy loss during the charging process , which takes into account the energy loss during the charging process

[0226]

[0227] The above formula represents the discharging electricity calculation in the charge-discharge constraints of the energy storage device, which represents the energy storage device at the discharging electricity at a moment is obtained by dividing the discharging power by the efficiency constant of energy loss during the discharging process , which takes into account the energy loss during the discharging process

[0228] Among them, the output constraints of the futures market include the following constraint conditions:

[0229]

[0230] The above formula represents the output range constraint in the output constraints of the futures market, which limits the output scenario under the actual output of the smart park in the futures market at a moment the value range of is the power of the difference contract, is the actual contract time-sharing electricity volume, is the deviation tolerance of the actual demand in the futures market and the time-sharing electricity of the contract for difference, that is, the actual output should be within the range of the sum of the contract for difference electricity adjusted according to the deviation tolerance and the actual contract time-sharing electricity volume.

[0231]

[0232] The above formula represents the total electricity balance constraint in the output constraint of the futures market, which means that in the output scenario during the entire time period the actual total output of the smart park in the futures market should be equal to the sum of the contract for difference electricity and the actual contract time-sharing electricity volume during the entire period to ensure the electricity balance in the futures market.

[0233] Among them, the day-ahead market bidding constraints include the following constraint conditions:

[0234]

[0235] The above formula represents the at the time, the power purchase and sale electric power of the smart park in the day-ahead market , which is obtained by subtracting the sum of the power of each wind turbine in the wind turbines and the sum of the power of each photovoltaic unit in the photovoltaic units from the sum of the power of each load in the loads in this scenario, which represents the power supply and demand relationship of the park in the day-ahead market.

[0236] Among them, the time-sharing electric energy balance constraint of the smart park includes the following constraint conditions:

[0237]

[0238] The above formula represents the electric energy balance relationship of the smart park at the at the time in the output scenario, represents the power purchase and sale electric power to the main power grid, represents the th conventional unit among the conventional units, represents the actual output of the th wind turbine among the wind turbines, represents the th photovoltaic unit among the photovoltaic units, Represents the discharge power of the energy storage device, Represents the electricity purchase and sale power of the day-ahead market, Represents the actual output of the futures market, Represents the actual load of the th non-flexible load among Represents the charging power of the energy storage device. The time-of-use power balance constraint of this smart park means that it is necessary to ensure the balance between power supply and demand in the smart park at each moment.

[0239]

[0240] The above formula limits the range of the electricity purchase and sale power of the smart park from and to the main power grid at time . If is greater than 0, it means selling electricity to the main power grid; otherwise, it means purchasing electricity from the main power grid, and the absolute value of the electricity purchase and sale power cannot exceed the maximum exchange power between the smart park and the main power grid to prevent the power exchange between the park and the main power grid from exceeding the allowable range.

[0241] Among them, constraints such as conditional value-at-risk constraints include the following constraint conditions:

[0242]

[0243] The above formula is a constraint based on conditional value-at-risk, represents the penalty cost of the th case under the output scenario , represents the risk value, represents the conditional risk value, represents the confidence level, represents the number of cases with penalty costs. This conditional value-at-risk constraint means that the sum of the risk values of the tail losses should be greater than or equal to the sum of the conditional risk values of the tail losses to ensure risk control at a certain confidence level. The sum of the risk values of the tail losses needs to be greater than or equal to 0.

[0244] Based on the above constraint conditions, and constructing an optimal scheduling model for the smart park, the objective function of the optimal scheduling model for the smart park can achieve the overall revenue maximization of the smart park in a multi-market environment by comprehensively considering various factors. Its objective function is as follows:

[0245]

[0246] In the above formula, represents the maximum overall revenue of the smart park under the output scenario , , Indicates the output scenario The calculation time period under Indicates the output scenario Under The actual output of the smart park in the futures market at the moment Indicates the contract electricity price in the futures market Indicates the output scenario Under The power purchase and sale electricity of the smart park in the day-ahead market at the moment. If Is positive, it means selling electricity and bringing benefits. If Is negative, it means purchasing electricity, which is a cost Indicates the output scenario Under The electricity price in the day-ahead market at the moment 、 Is the output scenario Under The power purchase and sale electricity in the real-time market at the moment 、 Is the output scenario Under The power purchase and sale electricity price in the real-time market at the moment Indicates the output scenario The total cost generated by the operation of conventional units and energy storage devices under Indicates the output scenario Under The carbon emission cost at the moment Indicates the output scenario The conditional value at risk under

[0247] Among them, the total cost generated by the operation of conventional units and energy storage devices is calculated as follows:

[0248]

[0249] In the above formula, Indicates the total cost generated by the operation of conventional units and energy storage devices under the output scenario Indicates the operating cost of the th conventional unit among conventional units Indicates th energy storage device among energy storage devices

[0250] Among them, the operating cost of the energy storage device is calculated as follows:

[0251]

[0252] In the above formula, Output scenario Lower The operating cost of the th energy storage device among the energy storage devices, And represents the charging and discharging power of the energy storage device.

[0253] The objective function provides a clear economic goal for the energy optimization scheduling of the smart park by comprehensively considering the revenues and various costs in different markets. When making energy scheduling decisions, such as determining the electricity purchase and sale strategies in different markets, the power generation power of distributed gas turbines, the charge and discharge plans of energy storage devices, etc., all are oriented towards maximizing this objective function, so as to achieve the optimal economic benefits of the smart park in a multi-market environment. At the same time, it also prompts the park to consider cost factors during the energy usage process, reasonably arrange energy production and consumption, and improve energy utilization efficiency.

[0254] In specific implementation, the preset constraint conditions include, for example, equipment capacity limitations, power balance requirements, energy market rules, etc. These constitute the boundaries for generating scheduling strategies. Different energy scheduling strategies can be generated according to the preset constraint conditions through algorithms or artificial experience, and finally s strategies are obtained.

[0255] Each scheduling strategy corresponds to a specific output scenario. By analyzing each scheduling strategy, the output scenario corresponding to each scheduling strategy among the s scheduling strategies can be determined, obtaining s output scenarios. Then, from the w objective condition risk values and w carbon emission costs, the objective condition risk values and carbon emission costs corresponding to the s output scenarios are selected, obtaining s objective condition risk values and s carbon emission costs.

[0256] The preset smart park optimization scheduling model includes an objective function and constraint conditions. Its objective function is a function that comprehensively considers risks, costs, and revenues. Using this model, based on the s objective condition risk values and s carbon emission costs, each scheduling strategy is evaluated, thereby obtaining s strategy scores. The strategy scores reflect the comprehensive performance of each scheduling strategy in aspects such as risk control, cost control, and meeting the energy demands of the park.

[0257] Select the highest strategy score from the s strategy scores to obtain the highest strategy score. The scheduling strategy corresponding to the highest strategy score is used as the target scheduling strategy. This target scheduling strategy performs optimally in aspects such as risk control, cost control, and meeting the energy demands of the park, so it can be used to guide the energy scheduling work of the smart park.

[0258] Please refer to Figure 5 , Figure 5It is an application scenario diagram of an optimization scheduling strategy generation method for an intelligent park provided by an embodiment of the present application. As shown in the figure, the current time is 12:00, the current wind speed is in a strong wind state, and the current light intensity is normal light. At the same time, a prediction diagram of power generation output is given. The solid line represents the actual output, and the dashed line represents the predicted output. The horizontal axis is time. This prediction diagram of power generation output can provide a reference for scheduling decisions.

[0259] Real-time dynamic information of the electricity market and the carbon market can include real-time data such as electricity prices and trading conditions in each electricity market, such as the futures market, the day-ahead market, and the real-time market, and carbon prices and quota trading in the carbon market.

[0260] According to meteorological conditions, power generation output conditions, and market information, multiple scheduling strategies are calculated through a preset intelligent park optimization scheduling model, namely Scheduling Strategy 1, Scheduling Strategy 2, and Scheduling Strategy 3. Among them, the score of Scheduling Strategy 1 is 85, the score of Scheduling Strategy 2 is 86, and the score of Scheduling Strategy 3 is 90. Based on the scores, the optimal scheduling strategy can be determined, and detailed information on the optimal scheduling strategy is given to guide the current energy scheduling of the intelligent park to achieve the optimal balance of economy and risk.

[0261] It can be seen that by comprehensively considering the target conditional risk value and carbon emission cost, and according to the preset constraint conditions, an optimal scheduling strategy that conforms to the actual situation can be generated, providing an accurate decision-making basis for park energy management.

[0262] In summary, by implementing the embodiment of the present application, historical data of wind turbines, photovoltaic units, and conventional units in the intelligent park are obtained, and wind power output models and photovoltaic output models under different output scenarios are constructed based on the historical data. Output data under different output scenarios are obtained through Monte Carlo simulation, the conditional risk value of the corresponding output scenario is determined according to the output data, the carbon emission cost is calculated in combination with the output data of the conventional unit, and finally, considering the conditional risk value and carbon emission cost comprehensively, the target scheduling strategy is screened under the preset constraint conditions. It can be seen that the target scheduling strategy obtained by comprehensively considering the conditional risk value and carbon emission cost under different wind power and photovoltaic output scenarios is the optimal scheduling strategy, which can effectively reduce the risks brought by the fluctuations of wind power and photovoltaic output, reduce carbon emissions, improve energy utilization efficiency, and improve the economic benefits and overall scheduling efficiency of the intelligent park in a multi-market environment.

[0263] Please refer to Figure 6 , Figure 6 It is a structural schematic diagram of an optimization scheduling strategy generation system for an intelligent park provided by an embodiment of the present application. The optimization scheduling strategy generation system for the intelligent park includes a data acquisition unit, a data simulation unit, and a strategy generation unit.

[0264] Among them, the data acquisition unit is responsible for collecting various energy-related data in the smart park, such as the real-time power generation of wind turbines and photovoltaic units, the operating parameters of conventional units, the power and charge-discharge status of energy storage devices, the load demand of the park at different times, as well as electricity market prices, carbon market trading information, etc.

[0265] The data simulation unit uses the data obtained by the data acquisition unit and adopts methods such as Monte Carlo simulation to simulate the uncertainty of new energy output, generate multiple output scenarios and corresponding data, and can also simulate power supply and demand, carbon emissions, etc. under different scenarios according to historical data and preset models, providing diverse scenario analyses for strategy generation.

[0266] The strategy generation unit, based on the results output by the data simulation unit, combines preset constraint conditions and optimization goals, such as cost reduction, risk control, carbon emission reduction, etc., generates multiple scheduling strategies through a preset optimization scheduling model for the smart park, evaluates and scores these strategies, and finally determines the optimal scheduling strategy to guide the energy scheduling operation of the smart park.

[0267] Please refer to Figure 7 , Figure 7 FIG. is a schematic structural diagram of an optimization scheduling strategy generation device for a smart park provided by an embodiment of the present application. The optimization scheduling strategy generation device 700 for the smart park is applied to a control device of the smart park. The smart park includes: wind turbines, photovoltaic units, and conventional units. The optimization scheduling strategy generation device 700 for the smart park includes: a data acquisition module 701, a scenario classification module 702, a model training module 703, a data simulation module 704, a data processing module 705, and a strategy generation module 706. Among them,

[0268] The data acquisition module 701 is used to obtain a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbines, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic units.

[0269] The scenario classification module 702 is used to classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenarios; w is equal to m multiplied by n, and both n and m are positive integers.

[0270] The model training module 703 is used to determine w wind power output models according to the first wind speed data set and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to a wind power output model and a photovoltaic output model.

[0271] The data simulation module 704 is configured to perform Monte Carlo simulation under the w output scenarios through the w wind power output models and the w photovoltaic power output models, so as to obtain k second wind power output data and k second photovoltaic power output data;

[0272] The data processing module 705 is configured to determine w target conditional risk values according to the k second wind power output data and the k second photovoltaic power output data;

[0273] The data acquisition module 701 is further configured to obtain the conventional output data of the conventional unit under the w output scenarios, so as to obtain w conventional output data;

[0274] The data processing module 705 is further configured to determine the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic power output data and the w conventional output data, so as to obtain w carbon emission costs;

[0275] The policy generation module 706 is configured to determine a target scheduling policy according to the w target conditional risk values and the w carbon emission costs.

[0276] Optionally, in terms of determining the w wind power output models according to the first wind speed data set and the w output scenarios, the model training module 703 is further specifically configured to:

[0277] Select the wind speed data corresponding to the first output scenario from the first wind speed data set to obtain a training wind speed data set; the first output scenario is any one of the w output scenarios;

[0278] Obtain a target likelihood function;

[0279] Solve the target likelihood function through a preset maximum likelihood estimation method and the training wind speed data set to obtain a first shape parameter and a first scale parameter;

[0280] Determine a first probability density function corresponding to the first output scenario according to the first shape parameter and the first scale parameter; the first probability density function is a Weibull probability density function;

[0281] Determine a wind power output model corresponding to the first output scenario according to a preset mapping relationship between wind speed and wind power output and the first probability density function.

[0282] Optionally, in terms of determining the w photovoltaic power output models according to the first solar irradiance data set and the w output scenarios, the model training module 703 is further specifically configured to:

[0283] Select first solar irradiance data corresponding to a second output scenario from the first solar irradiance dataset to obtain a training solar irradiance dataset; the second output scenario is any one of the w output scenarios;

[0284] Determine the average solar irradiance and the standard deviation of the solar irradiance corresponding to the training solar irradiance dataset;

[0285] Determine a second shape parameter and a third shape parameter according to the average solar irradiance and the standard deviation of the solar irradiance;

[0286] Determine a second probability density function corresponding to the second output scenario according to the second shape parameter and the third shape parameter; the second probability density function is a beta distribution probability density function;

[0287] Determine a photovoltaic output model corresponding to the second output scenario according to a preset normalization coefficient and the second probability density function.

[0288] Optionally, k is equal to w multiplied by e, where e is an integer greater than or equal to 1; each output scenario corresponds to e second wind power output data and e second photovoltaic output data; in terms of performing Monte Carlo simulation using the w wind power output models and the w photovoltaic output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic output data, the data simulation module 704 is further specifically configured to:

[0289] Select a wind power output model corresponding to a target output scenario from the w wind power output models to obtain a target wind power output model, and select a photovoltaic output model corresponding to the target output scenario from the w photovoltaic output models to obtain a target photovoltaic output model; the target output scenario is any one of the w output scenarios;

[0290] Perform Monte Carlo simulation using the target wind power output model and the target photovoltaic output model to obtain x second wind power output data and x second photovoltaic output data; x is an integer greater than or equal to e;

[0291] Remove outliers from the x second wind power output data to obtain e second wind power output data;

[0292] Remove outliers from the x second photovoltaic output data to obtain e second photovoltaic output data.

[0293] Optionally, in terms of removing outliers from the x second wind power output data to obtain e second wind power output data, the data simulation module 704 is further specifically configured to:

[0294] Select x target wind power output data corresponding to the target output scenario from the first wind power output dataset;

[0295] Determine the target Pearson correlation coefficient and the target mean square error according to the x second wind power output data and the x target wind power output data;

[0296] Determine the first weight coefficient corresponding to the target Pearson correlation coefficient and the second weight coefficient corresponding to the target mean square error; the sum of the first weight coefficient and the second weight coefficient is 1;

[0297] Determine the target comprehensive score according to the target Pearson correlation coefficient, the first weight coefficient, the target mean square error and the second weight coefficient;

[0298] If the target comprehensive score is less than or equal to the preset comprehensive score threshold, determine the wind power output difference between the x second wind power output data and the x target wind power output data to obtain x wind power output differences;

[0299] Select the wind power output differences less than the preset difference threshold from the x wind power output differences to obtain e wind power output differences;

[0300] Select the second wind power output data corresponding to the e wind power output differences from the x second wind power output data to obtain the e second wind power output data;

[0301] If the target comprehensive score is greater than the preset comprehensive score threshold, obtain the average value of the x second wind power output data, and select the e second wind power output data with the smallest absolute value of the difference from the x second wind power output data.

[0302] Optionally, in terms of determining w target conditional risk values according to the k second wind power output data and the k second photovoltaic output data, the data processing module 705 is further specifically configured to:

[0303] Obtain e second wind power output data and e second photovoltaic output data corresponding to the target output scenario;

[0304] Select the actual wind power output data corresponding to the target output scenario from the first wind power output dataset, and select the actual photovoltaic output data corresponding to the target output scenario from the first photovoltaic output dataset;

[0305] Determine e penalty costs according to the e second wind power output data, the e second photovoltaic output data, the actual wind power output data and the actual photovoltaic output data;

[0306] Determine a first risk value according to the e penalty costs and a preset confidence level;

[0307] Determine y first penalty costs according to the e penalty costs and the first risk value; y is a positive integer less than e;

[0308] Determine the operation of the y first penalty costs and the first risk value to obtain y risk differences;

[0309] Determine a target conditional risk value according to the first risk value and the y risk differences.

[0310] Optionally, in terms of determining e penalty costs according to the e second wind power output data, the e second photovoltaic output data, the actual wind power output data, and the actual photovoltaic output data, the data processing module 705 is further specifically configured to:

[0311] Determine the total wind power and photovoltaic output according to the e second wind power output data and the e second photovoltaic output data to obtain e total wind power and photovoltaic output values;

[0312] Determine the target total wind power and photovoltaic output according to the actual wind power output data and the actual photovoltaic output data;

[0313] Determine the difference between each total wind power and photovoltaic output value in the e total wind power and photovoltaic output values and the target total wind power and photovoltaic output value to obtain e output value differences;

[0314] Determine the e penalty costs according to the e output value differences, a preset first penalty coefficient, and a preset second penalty coefficient; wherein, if the output value difference i is greater than or equal to 0, determine the penalty cost i according to the output value difference i and the preset first penalty coefficient; if the output value difference i is less than 0, determine the penalty cost i according to the output value difference i and the preset second penalty coefficient; the output value difference i is any one of the e output value differences.

[0315] Optionally, in terms of determining carbon emission costs in the w output scenarios according to the k second wind power output data, the k second photovoltaic output data, and the w conventional output data to obtain w carbon emission costs, the data processing module 705 is further specifically configured to:

[0316] Obtain the w total load values corresponding to the smart park;

[0317] Determine w first carbon emissions according to a preset power supply carbon emission right quota benchmark value and the w total load values;

[0318] Determine w second carbon emissions according to the w conventional output data;

[0319] Determine w third carbon emissions based on the k second wind power output data and the k second photovoltaic power output data;

[0320] Determine w target carbon emissions based on the w first carbon emissions, the w second carbon emissions, and the w third carbon emissions;

[0321] Determine the w carbon emission costs based on the w target carbon emissions and the preset mapping relationship between carbon emissions and costs.

[0322] Optionally, in terms of determining the target scheduling strategy according to the w target conditional risk values and the w carbon emission costs, the policy generation module 706 is further specifically configured to:

[0323] Generate s scheduling strategies according to the preset constraint conditions;

[0324] Determine the output scenarios corresponding to each scheduling strategy among the s scheduling strategies, and obtain s output scenarios;

[0325] Select the target conditional risk values and carbon emission costs corresponding to the s output scenarios from the w target conditional risk values and the w carbon emission costs, and obtain s target conditional risk values and s carbon emission costs;

[0326] Determine s policy scores according to the s target conditional risk values, the s carbon emission costs, and the preset intelligent park optimal scheduling model;

[0327] Select the highest policy score from the s policy scores to obtain the highest policy score;

[0328] Take the scheduling strategy corresponding to the highest policy score as the target scheduling strategy.

[0329] The optimal scheduling strategy generation device 700 of the intelligent park described in this application can obtain the historical data of wind turbines, photovoltaic units, and conventional units in the intelligent park, construct a wind power output model and a photovoltaic power output model under different output scenarios based on the historical data, obtain the output data under different output scenarios through Monte Carlo simulation, determine the conditional risk value of the corresponding output scenario according to the output data, calculate the carbon emission cost in combination with the output data of the conventional unit, and finally comprehensively consider the conditional risk value and the carbon emission cost, and screen out the target scheduling strategy under the preset constraint conditions. It can be seen that the target scheduling strategy obtained by comprehensively considering the conditional risk value and the carbon emission cost under different wind power and photovoltaic output scenarios is the optimal scheduling strategy, which can effectively reduce the risk brought by the fluctuations of wind power and photovoltaic output, reduce carbon emissions, improve energy utilization efficiency, and improve the economic benefits and overall scheduling efficiency of the intelligent park in a multi-market environment.

[0330] Please refer toFigure 8 , Figure 8 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device may include a processor, a memory, a communication interface, and one or more programs. The processor, the memory, and the communication interface may be interconnected through a bus; the above one or more programs are stored in the above memory and are configured to be executed by the above processor; in the embodiment of the present application, the above program includes instructions for performing the following steps:

[0331] Obtain a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbine generator set, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic generator set;

[0332] Classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenarios; w is equal to m multiplied by n, and both n and m are positive integers;

[0333] Determine w wind power output models according to the first wind speed data set and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to a wind power output model and a photovoltaic output model;

[0334] Perform Monte Carlo simulation through the w wind power output models and the w photovoltaic output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic output data;

[0335] Determine w target conditional risk values according to the k second wind power output data and the k second photovoltaic output data;

[0336] Obtain the conventional output data of the conventional unit under the w output scenarios to obtain w conventional output data;

[0337] Determine the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data, and the w conventional output data to obtain w carbon emission costs;

[0338] Determine a target scheduling strategy according to the w target conditional risk values and the w carbon emission costs.

[0339] The electronic device described in this application can obtain the historical data of wind turbines, photovoltaic units, and conventional units in an intelligent park, construct a wind power output model and a photovoltaic power output model under different output scenarios based on the historical data, obtain the output data under different output scenarios through Monte Carlo simulation, determine the conditional risk value of the corresponding output scenario according to the output data, calculate the carbon emission cost in combination with the output data of the conventional unit, and finally, comprehensively consider the conditional risk value and the carbon emission cost, and screen out the target scheduling strategy under the preset constraints. It can be seen that the target scheduling strategy obtained by comprehensively considering the conditional risk value and the carbon emission cost under different wind power and photovoltaic output scenarios is the optimal scheduling strategy, which can effectively reduce the risks brought by the fluctuations in wind power and photovoltaic output, reduce carbon emissions, improve energy utilization efficiency, and improve the economic benefits and overall scheduling efficiency of the intelligent park in a multi-market environment.

[0340] The embodiment of this application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, and the computer program enables a computer to execute part or all of the steps of any method recorded in the above method embodiment. The above computer includes an electronic device.

[0341] The embodiment of this application also provides a computer program product. The above computer program product includes a non-transitory computer-readable storage medium storing a computer program. The above computer program is operable to enable a computer to execute part or all of the steps of any method recorded in the above method embodiment. The computer program product can be a software installation package. The above computer includes an electronic device.

[0342] Those of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by a computer program instructing relevant hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. The foregoing storage medium includes: various media such as ROM or random access memory RAM, magnetic disk, or optical disc that can store program code.

[0343] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the embodiments of this application. It should be understood that the above are only specific embodiments of the embodiments of this application and are not used to limit the protection scope of the embodiments of this application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the embodiments of this application should be included in the protection scope of the embodiments of this application.

Claims

1. A method for generating an optimization scheduling strategy for a smart park, characterized in that: A control device applied to an intelligent park, wherein the intelligent park includes: a wind turbine unit, a photovoltaic unit, and a conventional unit; the method includes: Acquire a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbine generator set, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic generator set; The first wind speed data set and the first solar irradiance data set are classified to obtain m wind power categories and n photovoltaic categories, and the m wind power categories and the n photovoltaic categories are combined to obtain w output scenarios; w is equal to m multiplied by n, and n and m are both positive integers; Determine w wind power output models according to the first wind speed data set and the w output scenarios, and determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to one wind power output model and one photovoltaic output model; Performing Monte Carlo simulation on the w wind power output models and the w photovoltaic output models under the w output scenarios to obtain k second wind power output data and k second photovoltaic output data; Determining w target condition risk values ​​according to the k second wind power output data and the k second photovoltaic output data; Obtaining conventional output data of the conventional unit in the w output scenarios to obtain w conventional output data; Determine the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data and the w conventional output data, to obtain w carbon emission costs; A target scheduling strategy is determined according to the w target condition risk values ​​and the w carbon emission costs.

2. The method according to claim 1, characterized in that The determining w wind power output models according to the first wind speed data set and the w output scenarios includes: Select wind speed data corresponding to a first output scenario from the first wind speed data set to obtain a training wind speed data set; the first output scenario is any one of the w output scenarios; Get the target likelihood function; Solving the target likelihood function by using a preset maximum likelihood estimation method and the training wind speed data set to obtain a first shape parameter and a first scale parameter; Determine a first probability density function corresponding to the first output scenario according to the first shape parameter and the first scale parameter; the first probability density function is a Weibull probability density function; A wind power output model corresponding to the first output scenario is determined according to a preset mapping relationship between wind speed and wind power output and the first probability density function.

3. The method according to claim 1, characterized in that The determining w photovoltaic output models according to the first solar irradiance data set and the w output scenarios includes: Selecting first solar irradiance data corresponding to a second output scene from the first solar irradiance data set to obtain a training solar irradiance data set; the second output scene is any one of the w output scenes; Determine the solar irradiance mean value and the solar irradiance standard deviation corresponding to the training solar irradiance data set; Determine a second shape parameter and a third shape parameter according to the solar irradiance average value and the solar irradiance standard deviation; Determine a second probability density function corresponding to the second output scenario according to the second shape parameter and the third shape parameter; the second probability density function is a Beta distribution probability density function; A photovoltaic output model corresponding to the second output scenario is determined according to a preset normalization coefficient and the second probability density function.

4. The method according to any one of claims 1 to 3, characterized in that: k is equal to w multiplied by e, where e is an integer greater than or equal to 1; each output scenario corresponds to e second wind power output data and e second photovoltaic output data; the Monte Carlo simulation is performed under the w output scenarios using the w wind power output models and the w photovoltaic output models to obtain k second wind power output data and k second photovoltaic output data, including: Selecting a wind power output model corresponding to a target output scenario from the w wind power output models to obtain a target wind power output model, and selecting a photovoltaic output model corresponding to the target output scenario from the w photovoltaic output models to obtain a target photovoltaic output model; the target output scenario is any one of the w output scenarios; Performing Monte Carlo simulation through the target wind power output model and the target photovoltaic output model to obtain x second wind power output data and x second photovoltaic output data; x is an integer greater than or equal to e; removing abnormal values ​​from the x pieces of second wind power output data to obtain e pieces of second wind power output data; Abnormal values ​​in the x second photovoltaic output data are removed to obtain e second photovoltaic output data.

5. The method according to claim 4, characterized in that The removing of abnormal values ​​from the x second wind power output data to obtain e second wind power output data includes: Selecting x target wind power output data corresponding to the target output scenario from the first wind power output data set; Determine a target Pearson correlation coefficient and a target mean square error according to the x second wind power output data and the x target wind power output data; Determine a first weight coefficient corresponding to the target Pearson correlation coefficient and a second weight coefficient corresponding to the target mean square error; the sum of the first weight coefficient and the second weight coefficient is 1; Determine a target comprehensive score according to the target Pearson correlation coefficient, the first weight coefficient, the target mean square error and the second weight coefficient; If the target comprehensive score is less than or equal to the preset comprehensive score threshold, determining the wind power output difference between the x second wind power output data and the x target wind power output data to obtain x wind power output difference values; Selecting a wind power output difference value that is less than a preset difference threshold value from the x wind power output difference values ​​to obtain e wind power output difference values; Selecting the second wind power output data corresponding to the e wind power output differences from the x second wind power output data to obtain the e second wind power output data; If the target comprehensive score is greater than the preset comprehensive score threshold, an average value of the x second wind power output data is obtained, and e second wind power output data with the smallest absolute value of the difference with the average value are selected from the x second wind power output data.

6. The method according to claim 4, characterized in that The determining w target condition risk values ​​according to the k second wind power output data and the k second photovoltaic output data comprises: Obtaining e second wind power output data and e second photovoltaic output data corresponding to the target output scenario; Selecting actual wind power output data corresponding to the target output scenario from the first wind power output data set, and selecting actual photovoltaic output data corresponding to the target output scenario from the first photovoltaic output data set; determining e penalty costs according to the e second wind power output data, the e second photovoltaic output data, the actual wind power output data and the actual photovoltaic output data; Determine a first risk value according to the e penalty costs and a preset confidence level; Determine y first penalty costs according to the e penalty costs and the first risk value; y is a positive integer less than e; Determine the y first penalty costs and the first risk value to perform operations to obtain y risk difference values; A target conditional risk value is determined according to the first risk value and the y risk difference values.

7. The method according to claim 6, characterized in that The determining e penalty costs according to the e second wind power output data, the e second photovoltaic output data, the actual wind power output data and the actual photovoltaic output data comprises: Determine the total wind power and photovoltaic output value according to the e second wind power output data and the e second photovoltaic output data, and obtain e total wind power and photovoltaic output values; Determine a target wind power and photovoltaic power total output value according to the actual wind power output data and the actual photovoltaic power output data; Determine the difference between each of the e wind power photovoltaic total output values ​​and the target wind power photovoltaic total output value to obtain e total output value differences; The e penalty costs are determined according to the e total output value differences, a preset first penalty coefficient, and a preset second penalty coefficient; wherein, if the total output value difference i is greater than or equal to 0, the penalty cost i is determined according to the total output value difference i and the preset first penalty coefficient; if the total output value difference i is less than 0, the penalty cost i is determined according to the total output value difference i and the preset second penalty coefficient; the total output value difference i is any total output value difference among the e total output value differences.

8. The method according to claim 1, characterized in that The step of determining the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data and the w conventional output data to obtain the w carbon emission costs includes: Obtain w total load values ​​corresponding to the smart park; Determine w first carbon emissions according to a preset power supply carbon emission quota reference value and the w total load values; Determining w second carbon emissions according to the w conventional output data; Determining w third carbon emissions according to the k second wind power output data and the k second photovoltaic output data; Determining w target carbon emissions according to the w first carbon emissions, the w second carbon emissions, and the w third carbon emissions; The w carbon emission costs are determined according to the w target carbon emissions and a mapping relationship between preset carbon emissions and costs.

9. The method according to claim 1, characterized in that The determining of the target scheduling strategy according to the w target condition risk values ​​and the w carbon emission costs includes: Generate s scheduling strategies according to the preset constraints; Determine the output scenario corresponding to each of the s scheduling strategies to obtain s output scenarios; Selecting the target condition risk values ​​and carbon emission costs corresponding to the s output scenarios from the w target condition risk values ​​and the w carbon emission costs to obtain s target condition risk values ​​and s carbon emission costs; Determine s strategy scores according to the s target condition risk values, the s carbon emission costs and a preset smart park optimization scheduling model; Select the highest strategy score from the s strategy scores to obtain the highest strategy score; The scheduling strategy corresponding to the highest strategy score is used as the target scheduling strategy.

10. A device for generating an optimization scheduling strategy for a smart park, characterized in that: A control device applied to a smart park, wherein the smart park includes: a wind turbine, a photovoltaic unit, and a conventional unit; the optimization scheduling strategy generation device of the smart park includes: a data acquisition module, a scene classification module, a model training module, a data simulation module, a data processing module, and a strategy generation module, wherein: The data acquisition module is used to obtain a first wind speed data set within a historical time period, a first wind power output data set corresponding to the wind turbine generator set, a first solar irradiance data set, and a first photovoltaic output data set corresponding to the photovoltaic generator set; The scene classification module is used to classify the first wind speed data set and the first solar irradiance data set to obtain m wind power categories and n photovoltaic categories, and combine the m wind power categories and the n photovoltaic categories to obtain w output scenes; w is equal to m multiplied by n, and n and m are both positive integers; The model training module is used to determine w wind power output models according to the first wind speed data set and the w output scenarios, and to determine w photovoltaic output models according to the first solar irradiance data set and the w output scenarios; each output scenario corresponds to one wind power output model and one photovoltaic output model; The data simulation module is used to perform Monte Carlo simulation under the w output scenarios through the w wind power output models and the w photovoltaic output models to obtain k second wind power output data and k second photovoltaic output data; The data processing module is used to determine w target condition risk values ​​according to the k second wind power output data and the k second photovoltaic output data; The data acquisition module is also used to obtain the conventional output data of the conventional unit in the w output scenarios to obtain w conventional output data; The data processing module is further used to determine the carbon emission costs under the w output scenarios according to the k second wind power output data, the k second photovoltaic output data and the w conventional output data, to obtain w carbon emission costs; The strategy generation module is used to determine the target scheduling strategy according to the w target condition risk values ​​and the w carbon emission costs.

Images