Power market deduction multi-scene generation method and related device

Abstract

The invention belongs to a scene generation method based on prediction data, and provides a power market deduction multi-scene generation method and a related device for solving the problems that a deduction scene generated during power market simulation deduction at present lacks a rigorous theoretical basis and has the capability of covering small-probability extreme events in actual production. The method comprises the steps that after historical operation data is obtained, error preprocessing is conducted on the historical operation data, preprocessing is conducted on the operation data through histogram visualization, then a Gaussian distribution fitting curve is overlaid on an error histogram, and a corresponding error probability density function is obtained according to a fitting result; and a new energy uncertainty problem in the electricity market is converted into a prediction error uncertainty problem. And sampling the cumulative distribution function by adopting a Latin hypercube probability sampling method to obtain a corresponding combined sample, and finally superposing the combined sample onto historical operation data to obtain prediction data for power market deduction, and generating power market deduction multiple scenes according to the prediction data.

Description

Technical Field

[0001] The present application relates to a scenario generation method based on forecast data, and specifically to a multi-scenario generation method for power market simulation and related devices. Background Art

[0002] In the new power system, the high proportion of new energy access has brought unprecedented challenges to the operation of the power market, especially the randomness and intermittency of new energy, which has significantly increased the uncertainty of the balance of power supply and demand. In this context, analyzing and verifying policies and rules through power market simulation has become a key means to explore the operation mechanism of the new power system, and an important tool to support policy formulation, market optimization and risk assessment.

[0003] The core of power market simulation is to accurately reproduce the real behavior of market operation and evaluate the adaptability and robustness of market mechanisms to different operating conditions in a variety of scenarios. However, this process places extremely high demands on the rationality, comprehensiveness and diversity of input data. Usually, the simulation system is based on historical operating data and generates simulation scenarios through fine-tuning of certain rules or experience. However, this method is limited to the simulation of known conditions and conventional operating scenarios, and lacks a rigorous theoretical basis and the ability to cover small-probability extreme events in actual production. Summary of the invention

[0004] This application aims to address the technical problems of the simulation scenarios generated during the current power market simulation, which lack a rigorous theoretical basis and the ability to cover small-probability extreme events in actual production, and provides a method and related devices for generating multiple scenarios for power market simulation.

[0005] In order to achieve the above objectives, this application adopts the following technical solutions: In the first aspect, the present application proposes a method for generating multiple scenarios for power market simulation, including: Get historical operation data; Performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data using a histogram to obtain an error histogram; Superimposing a Gaussian distribution fitting curve on the error histogram, and obtaining a corresponding error probability density function according to the fitting result; Calculate the cumulative distribution function corresponding to the Gaussian distribution according to the error probability density function; Using the Latin hypercube probability sampling method, sampling the cumulative distribution function to obtain a corresponding combined sample; Superimposing the combined sample on the historical operation data to obtain forecast data for power market deduction; Multiple scenarios for power market simulation are generated based on the forecast data.

[0006] Furthermore, the error preprocessing of the historical operation data includes abnormal data correction, missing value supplementation and data normalization.

[0007] Furthermore, the abnormal data correction includes: Raida criterion was used to screen outliers; Use at least one of the following methods to correct outliers: fill in the mean of outliers, correct missing values ​​by treating outliers as missing data, and correct outliers. Replace similar days and delete outliers.

[0008] Furthermore, the missing value supplementation includes: supplementing the missing values ​​using Lagrange interpolation method.

[0009] Furthermore, the electricity market includes wind power and photovoltaics; The formula for obtaining the corresponding error probability density function according to the fitting result is as follows:

[0010] in, is the error probability density function of wind power, for t The error value of photovoltaic prediction at the moment, for t The standard deviation of the photovoltaic prediction error at the moment, for t The expected value of the photovoltaic prediction error at time is the error probability density function of photovoltaic, for t The error value of wind power forecast at the moment, for t The standard deviation of wind power forecast error at each moment, for t The expected value of wind power forecast error at that moment.

[0011] Furthermore, the cumulative distribution function corresponding to the Gaussian distribution is calculated using the following formula:

[0012] in, is the cumulative distribution function corresponding to the Gaussian distribution of wind power prediction error, is the cumulative distribution function corresponding to the Gaussian distribution obeyed by the photovoltaic prediction error.

[0013] Furthermore, the Latin hypercube probability sampling method is used to sample the cumulative distribution function to obtain the corresponding combined sample, including: Divide the cumulative distribution function into n intervals, and obtain a set T of all intervals; where n is the number of preset generated scenes; Randomly select a uniformly distributed value in each of the n intervals; All the uniformly distributed values ​​selected are mapped to the Gaussian distribution through the inverse transform sampling method to obtain the sample points of the Gaussian distribution; According to the sample points of Gaussian distribution, the prediction errors of wind power and photovoltaic power are arranged according to the time period to form a combined sample.

[0014] Furthermore, the generating of multiple scenarios of power market simulation according to the forecast data includes: A distributed computing module is used to perform parallel calculations on each set of forecast data to generate corresponding power market simulation scenarios.

[0015] Furthermore, before using the distributed computing module to perform calculations on each set of prediction data in parallel, the method further includes: packaging the prediction data and the tasks for performing the calculations into a mirror image.

[0016] In the second aspect, the present application proposes a multi-scenario generation system for power market simulation, including: Data module, used to obtain historical operation data; An error preprocessing module, used for performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data with a histogram to obtain an error histogram; A Gaussian fitting module, used to superimpose a Gaussian distribution fitting curve on the error histogram, and obtain a corresponding error probability density function according to the fitting result; A statistical module, used for calculating a cumulative distribution function corresponding to a Gaussian distribution according to the error probability density function; A sampling module, used for sampling the cumulative distribution function by using a Latin hypercube probability sampling method to obtain a corresponding combined sample; A superposition module, used for superimposing the combined sample onto the historical operation data to obtain forecast data for power market deduction; A deduction module is used to generate multiple scenarios of power market deduction based on the forecast data.

[0017] In the third aspect, the present application proposes an electronic device, comprising: a memory, and one or more processors; the memory is coupled to the processor; wherein computer program code is stored in the memory, and the computer program code includes computer instructions, and when the computer instructions are executed by the processor, the electronic device performs the steps of the above-mentioned method for generating multiple scenarios for electricity market simulation.

[0018] In a fourth aspect, the present application proposes a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for generating multiple scenarios for power market simulation are implemented.

[0019] Compared with the prior art, this application has the following beneficial effects: This application proposes a method for generating multiple scenarios for power market simulation. After obtaining historical operating data, in order to ensure the accuracy in the subsequent data processing process, the historical operating data is firstly preprocessed for error, and then the operating data is preprocessed through histogram visualization to understand the distribution trend of the error, and then the Gaussian distribution fitting curve is superimposed on the error histogram. According to the fitting result, the corresponding error probability density function is obtained, and the uncertainty problem of new energy in the power market is transformed into the problem of prediction error uncertainty. The Latin hypercube probability sampling method is then used to sample the cumulative distribution function to obtain the corresponding combined samples, ensuring that the sample space is evenly covered within the uncertainty range, which is suitable for multi-scenario analysis. Finally, the combined samples are superimposed on the historical operating data to obtain the forecast data for power market simulation, and multiple scenarios for power market simulation are generated according to the forecast data. This application improves the adaptability and comprehensiveness of power market simulation to complex operating conditions.

[0020] The present application also proposes a power market simulation multi-scenario generation system, electronic equipment and computer storage medium, which have all the advantages of the above-mentioned power market simulation multi-scenario generation method. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 This is a first schematic diagram of a multi-scenario generation method for power market simulation in this application; Figure 2 A second flow chart of the multi-scenario generation method for power market simulation in this application; Figure 3 This is a schematic diagram of the principle of the DistRun distributed computing module; Figure 4 A schematic diagram of the process principle of the multi-scenario generation method for power market simulation in this application; Figure 5 A schematic diagram of a multi-scenario generation system for power market simulation in this application. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0024] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0025] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0026] In the description of the embodiments of the present application, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the drawings, or the orientation or position relationship in which the invented product is usually placed when used. It is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0027] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0028] In the description of the embodiments of the present application, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0029] In the new power system, with the substantial increase in the proportion of new energy (such as wind power, solar energy, etc.), the operation of the power market faces unprecedented challenges. These new energy sources have significant randomness and intermittent characteristics. The power generation output of new energy sources is affected by many uncontrollable factors such as weather conditions and seasonal changes, resulting in a significant increase in the uncertainty of the balance of power supply and demand. This uncertainty also brings tremendous pressure to the stable operation of the power system and the effective dispatch of the power market. In order to meet these challenges, power market simulation has become an important analytical tool. By constructing a simulation model of the power market and conducting simulations based on actual or assumed market conditions, the actual effects of different policies and rules in the power market can be deeply analyzed and verified. This simulation not only helps to explore the operating mechanism of the new power system, but also provides a scientific basis for policymakers to help formulate more reasonable and effective power market policies. At the same time, power market simulation is also an important tool for market optimization and risk assessment. By simulating the power supply and demand under different market conditions, the adaptability and robustness of the market mechanism can be evaluated, and potential market problems and risk points can be discovered. It helps market operators to adjust market rules in a timely manner, optimize market structure, and improve the efficiency and stability of the power market.

[0030] The core of power market simulation is to accurately reproduce the real behavior of market operation, which requires that the simulation system must be able to comprehensively and accurately reflect the various complex relationships and dynamic changes in the power market. In order to achieve this goal, the simulation system needs to be built and operated based on a large amount of input data. These input data include historical operation data, market rules, cost structure of power generation enterprises, user demand behavior, etc. Their rationality and comprehensiveness are crucial to the accuracy of the simulation results. However, it is far from enough to rely solely on historical operation data for simulation. Because historical data can only reflect past market conditions, but cannot cover all possible future scenarios. Especially for those extreme events with low probability but great impact, there is often a lack of sufficient samples in historical data for simulation. Therefore, the simulation system also needs to be fine-tuned through certain rules or experience to generate more diverse and realistic simulation scenarios. However, this rule- and experience-based fine-tuning method also has limitations. It mainly relies on manually set rules and parameters and lacks a rigorous theoretical basis. At the same time, due to the limited cognition and experience of people, this method may not be able to fully cover all possible operating conditions and extreme events. Therefore, in order to improve the accuracy and reliability of electricity market simulation, it is necessary to continuously explore and introduce new theories and methods to better simulate and predict the dynamic changes of the electricity market.

[0031] In actual operation, low-probability scenarios such as extreme weather events and large-scale new energy grid connection may cause market operations to deviate from the normal trajectory, triggering sharp price fluctuations, market failures, and even power supply crises. Although these scenarios are rare, their potential impacts may be catastrophic. Therefore, how to construct an input data set that can reflect low-probability extreme events in power market simulation has become the focus and difficulty of current research. Therefore, for new power systems with a high proportion of new energy access, it is an important part of improving market resilience and efficiency to reasonably construct input data and conduct power market simulation. By constructing comprehensive and scientific simulation scenarios, the power market will be able to better adapt to the challenges of complex operating environments in the future, while providing key decision support for policymakers to ensure the realization of system stability and economic goals.

[0032] Based on the above situation, the present application proposes a method and related devices for generating multiple scenarios for power market simulation. The present application is described in detail below in conjunction with embodiments and drawings.

[0033] like Figure 1 As shown, it is a first schematic diagram of the method for generating multiple scenarios for power market simulation in this application, which may include: S101, obtaining historical operation data.

[0034] Historical operation data is the basis for power market deduction and helps to understand the market's operating rules and potential risks. The collected historical operation data may include power generation, power consumption, electricity prices, market participants' behavior records, weather conditions, equipment failure records, etc. The source of data may include power company operation reports, market regulators' public data, meteorological departments' weather records, etc. The time span of the data can be set according to the forecast requirements. For example, it can cover different market cycles and extreme events to ensure the comprehensiveness and representativeness of the data.

[0035] S102, performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data with a histogram to obtain an error histogram.

[0036] Error preprocessing helps eliminate noise and bias in data and improves data reliability and availability. Histogram visualization provides an intuitive view of data distribution and helps identify data features and potential problems.

[0037] When formulating the day-ahead power generation plan, the grid dispatching department needs to grasp the forecast value of the next day's active power of the entire network's renewable energy and the system load demand. Due to the large scale of the power system and the complex and changeable operating conditions, there are orderly power consumption, channel obstruction, extreme weather and equipment failure, and some data is missing and abnormal in the historical operation database. In order to prevent bad data from interfering with the accuracy of renewable energy forecasts and load forecasts, historical data should be preprocessed first. After the preprocessing is completed, the preprocessed operating data is visualized using a histogram to show the distribution characteristics and error conditions of the data and form an error histogram.

[0038] S103, superimposing a Gaussian distribution fitting curve on the error histogram, and obtaining a corresponding error probability density function according to the fitting result.

[0039] Gaussian distribution fitting helps to understand the distribution law of data errors and provides a theoretical basis for subsequent sampling and deduction. The probability density function is an important tool for describing the probability of random variable values ​​and can provide detailed statistical information about data errors. On the error histogram, Gaussian distribution is used for fitting to describe the probability distribution characteristics of the data. By adjusting the parameters of the Gaussian distribution, the fitting curve is made as close to the contour of the histogram as possible. After the fitting is completed, the probability density function describing the data error can be obtained based on the fitting results.

[0040] S104: Calculate a cumulative distribution function corresponding to a Gaussian distribution according to the error probability density function.

[0041] The cumulative distribution function is the basis for probability sampling and provides a method for extracting samples from the probability density function. In power market simulation, the cumulative distribution function helps to generate forecast data that conforms to the actual error distribution. Using the error probability density function, the corresponding Gaussian distribution cumulative distribution function can be calculated. The cumulative distribution function describes the probability that a random variable takes a value less than or equal to a certain value.

[0042] S105, using a Latin hypercube probability sampling method to sample the cumulative distribution function to obtain a corresponding combined sample.

[0043] Latin hypercube sampling is an efficient probability sampling method that can reduce the number of samples while ensuring the representativeness of the samples. By performing Latin hypercube sampling on the cumulative distribution function, a set of samples with a specific probability distribution can be obtained, which constitute the combined samples required for the deduction. Latin hypercube sampling can improve the efficiency and accuracy of sampling and ensure the diversity and representativeness of the deduction scenarios. Combined samples are the basis for generating deduction scenarios, and their quality and quantity directly affect the reliability and comprehensiveness of the deduction results.

[0044] S106, superimposing the combined sample onto the historical operation data to obtain forecast data for power market deduction.

[0045] The combined samples obtained by Latin hypercube sampling are superimposed on the historical operating data to form forecast data containing error information. These forecast data not only retain the overall characteristics of historical data, but also incorporate the error distribution information obtained through sampling. By superimposing combined samples, forecast data can more truly reflect the potential risks and uncertainties of the market.

[0046] S107, generating multiple scenarios for power market simulation based on the forecast data.

[0047] Based on the forecast data, multiple power market simulation scenarios can be generated. Each scenario represents a possible operating state of the market in the future, and can include forecast values ​​of key variables such as power generation, power consumption, electricity prices, and market participant behavior. These scenarios can cover different market conditions and extreme events to ensure the comprehensiveness and robustness of the simulation.

[0048] In view of the limitations of the existing technology in the simulation and deduction of multi-scenario electricity markets, this application proposes a method for generating multi-scenario deduction data based on probability distribution. This method can generate output curves of new energy units according to the requirements of different electricity market deduction scenarios, and combine these data with known quantities to form boundary conditions for electricity market simulation and deduction. This method provides data support for the continuous simulation and deduction domain analysis and verification of multi-scenario electricity markets. In terms of the prediction and uncertainty modeling of new energy output, this application uses Gaussian distribution to perform statistical modeling on new energy prediction errors. In order to generate diversified scenario data, this application combines Latin hypercube sampling technology to sample from the fitted Gaussian distribution to generate new energy output data and system load forecast data covering multiple scenarios. Among them, Latin hypercube sampling technology ensures uniform coverage of the sample space, so that the deduction data includes both conventional scenarios and low-probability extreme scenarios, thereby improving the adaptability and comprehensiveness of electricity market simulation and deduction to complex operating conditions.

[0049] like Figure 2 As shown, it is a second flow chart of the method for generating multiple scenarios for power market simulation in this application, which may include: S201, historical data preprocessing.

[0050] In order to prevent bad data from interfering with the accuracy of new energy forecast and load forecast, the historical data is preprocessed as follows: (1) Identification and correction of abnormal data.

[0051] Values ​​that seriously deviate from the sample distribution interval are considered outliers. In the power system operation reserve capacity demand assessment model, the sample data are the new energy day-ahead forecast deviation and the load day-ahead forecast deviation. When the sample size is large enough, the Raida criterion can usually be used to screen outliers. Generally, values ​​with a deviation of more than 3 standard deviations from the total sample mean value are considered outliers, that is:

[0052] in, is the new energy forecast error or load forecast error, is the sample mean, is the sample standard deviation.

[0053] When a value meets the Raida criterion, it means that the value is an outlier and the following outlier processing can be performed: 1) Mean filling: Take the average of the same number of samples before and after the outlier as the filling value.

[0054] 2) Missing value correction: outliers are treated as missing data in the sample.

[0055] 3) Similar day replacement: For a small number of data anomalies, similar days with higher prediction power correlation are used to replace the abnormal data.

[0056] 4) Direct deletion: For data anomalies for consecutive days in the historical database, the outliers can be directly deleted when the sample size is large enough.

[0057] (2) Missing value supplementation.

[0058] The missing values ​​of the prediction error are filled using the Lagrange interpolation method:

[0059] Where k is the total number of sampling periods, is the interpolation basis function, for The forecast error corresponding to the time period, is the sampling period number, is the sampling period number, is the sampling period, For sampling Time period, For sampling The missing items can be calculated by substituting the sample missing value period into the Lagrange interpolation formula.

[0060] (3) Normalization of sample data.

[0061] After completing the outlier replacement and missing value supplementation, in order to facilitate data processing, the sample population needs to be normalized:

[0062] in, is the actual value of the new energy prediction error, is the per unit value of the new energy prediction error, It is the installed capacity of new energy in the corresponding period.

[0063] S202, draw an error distribution histogram.

[0064] Use a histogram to visualize the distribution of error data. The histogram shows the frequency of samples within each error range, helping us understand the distribution trend of the error.

[0065] S203, new energy prediction error fitting based on Gaussian distribution.

[0066] The scenario generation method based on the probability model fits the probability density distribution of historical data and samples the probability distribution to obtain the output scenario.

[0067] This application regards the output of renewable energy as the sum of predicted power and uncertain predicted power error. To reduce the complexity of the model and make it easier to quantitatively analyze the uncertainty of wind power and photovoltaic power generation, the problem of wind and solar power output uncertainty is transformed into the problem of prediction error uncertainty. Therefore, the actual power generation of wind power and photovoltaic output can be expressed as:

[0068] in, is the photovoltaic power at time t, is the actual wind power output value at time t, is the predicted photovoltaic output value at time t, is the predicted output value of wind power, is the photovoltaic prediction error value at time t, is the wind power prediction error value at time t.

[0069] The wind power and photovoltaic prediction errors obey Gaussian distribution, and the corresponding probability density function can be expressed as:

[0070] in, is the error probability density function of wind power, for t The error value of photovoltaic prediction at the moment, for t The standard deviation of the photovoltaic prediction error at the moment, for t The expected value of the photovoltaic prediction error at time is the error probability density function of photovoltaic, for t The error value of wind power forecast at the moment, for t The standard deviation of wind power forecast error at each moment, for t The expected value of wind power forecast error at that moment.

[0071] S204, Latin hypercube probability sampling.

[0072] After fitting the Gaussian distribution of renewable energy forecast errors, the Latin Hypercube Sampling (LHS) method can be used to generate forecast data for power market simulation. This method ensures uniform coverage of the sample space within the uncertainty range and is very suitable for multi-scenario analysis.

[0073] Latin hypercube sampling is a stratified sampling method that ensures that a sample point is selected from each distribution interval and is evenly distributed throughout the space. For the application of Gaussian distribution, the key steps are as follows: 1) Set the number of samples.

[0074] Define the number of samples n that need to be generated, that is, the number of scenes you want to generate is n.

[0075] 2) Division of stratified intervals.

[0076] The Cumulative Distribution Function (CDF) describes the probability that a random variable takes a value less than or equal to a certain value. When sampling, the inverse transformation method of the CDF is usually relied on to generate samples that conform to a certain distribution. The CDF corresponding to the Gaussian distribution obeyed by the above-mentioned wind power and photovoltaic prediction errors is as follows:

[0077] The cumulative distribution function (CDF) of the Gaussian distribution is divided into n intervals. The range of the cumulative distribution function F(x) is [0, 1]. It is divided into n segments, and the width of each interval is 1 / n. At this time, the set of all intervals T can be expressed as: .

[0078] 3) Draw samples within the interval.

[0079] A sample point is randomly selected in each interval to ensure coverage of the entire distribution.

[0080] For i intervals, randomly select a uniformly distributed value : .

[0081] 4) Map the sample points to Gaussian distribution.

[0082] These uniformly distributed points are mapped to a Gaussian distribution through inverse transform sampling.

[0083] Inverse transformation of the cumulative distribution function using Gaussian distribution ,Will Sample points converted to Gaussian distribution :

[0084] in, It is the inverse function of the Gaussian distribution and can be calculated numerically.

[0085] above That is the i The prediction error values ​​corresponding to the stratified intervals can be used to obtain samples of wind power and photovoltaic output prediction errors. These error samples are independent and can be arranged according to the time period to form a combined sample:

[0086] S205, generating multi-scenario data based on Latin hypercube.

[0087] The generated combined samples are superimposed on the original new energy forecast data to obtain new energy forecast data that can be used for deduction and analysis. Specifically, the forecast value at each time point is added to its corresponding error sample value, so that the forecast value at each time point becomes closer to the actual operation due to the addition of the error sample.

[0088] S206, parallel calculation of multiple deduction scenarios.

[0089] Assume that there are n sets of new energy forecast data in the simulation scenario. These data are extracted and generated from the specified sample set according to the simulation requirements. Each set of data reflects different new energy output trends, covering various types of energy changes such as wind energy and solar energy.

[0090] As an example, in order to quickly process these simulation scenarios and obtain calculation results, the DistRun distributed computing module of MindOptStudio can be used, such as Figure 3The figure shows the principle diagram of the DistRun distributed computing module. It should be noted that MindOpt Studio is a cloud platform that provides an easy-to-use interface and powerful background computing capabilities. The DistRun distributed computing module is a key component in MindOpt Studio. It can use distributed computing technology to improve the efficiency of solving optimization problems. By breaking down large computing tasks into multiple small tasks and processing them in parallel on multiple computers, the DistRun distributed computing module can significantly shorten the solution time and improve the utilization of computing resources. Each computing node independently processes the corresponding scene calculation according to the task allocation, avoiding the bottleneck caused by serial computing and accelerating the overall computing process.

[0091] According to the functions and usage of the DistRun module, in order to utilize the distributed cluster scheduling tasks, the multi-scenario deduction tasks containing new energy forecast data can also be packaged into images, created into batch tasks, and the clearing preparation, file generation, optimization calculation, and post-statistics clearing tasks of multiple scenarios can be processed in parallel to improve the deduction efficiency and obtain the deduction results faster. In practical applications, when packaging into images, containerization technology (such as Docker) can be used to achieve this.

[0092] Summarize the methods of the above embodiments of the present application, as follows: Figure 4 The figure shows the process principle diagram of the multi-scenario generation method for power market simulation in this application. Based on the probability distribution model of new energy and load forecast errors, the Latin hypercube sampling method is used to generate forecast error simulation data, and the validity is verified based on the probability density function of the simulation data and historical data. The whole process of parallel clearing of multiple scenarios based on the DistRun module is completed to generate multiple simulation scenarios from load forecast data and new energy forecast data.

[0093] As an example, a subsequent analysis is performed based on the multiple power market deduction scenarios generated by the above-mentioned power market deduction multiple scenario generation method to further discuss the role of this application: After completing the parallel clearing of the "Extremely Hot and No Wind" and other case groups, we conducted an in-depth analysis of key statistical indicators such as market electricity price information, market entity revenue, unit output, and market supply and demand. By carefully combing these data and calculating the expected value of each indicator, we revealed the average performance of these indicators in the deduction case group. At the same time, calculating the variance can quantitatively measure market volatility and help identify key factors that may affect market stability. Through further analysis, the stability of market electricity prices is closely related to the actual output of new energy units, while the volatility of thermal power unit revenue may be affected by the supply-demand ratio and market bidding space. Sensitivity analysis and risk assessment provide us with deep insights into potential risks in market operations and guide us to formulate more robust market strategies. In practical applications, a Gaussian distribution is fitted based on the sample characteristics of new energy prediction errors in different forms such as extreme heat and no wind, and Latin hypercube sampling is used to generate multiple sets of prediction error data for this scenario. Then, multiple sets of case data are generated in combination with other invariants in the case data. This constitutes a complete set of multiple sets of deduction case data based on the characteristics of the specified scenarios. Mindopt is used to call multiple resources to realize time-series parallel calculation of multiple sets of deduction scenarios, improve computing efficiency, and obtain the final statistical analysis data of the case based on expectation and variance analysis.

[0094] In the electricity market, variance statistics and expectation statistics are key tools for evaluating market performance and supporting decision-making. Variance analysis is used to compare significant differences in price fluctuations and unit output under different time periods or market conditions, while expectation statistics focus on average market performance, such as average market price, average unit output, etc. Through in-depth analysis of the clearing results of the "Extremely Hot and No Wind" case group, we collected key statistical indicator data including market price, unit revenue, output, and supply-demand ratio. Calculating the expected value of these indicators provides the central trend of market performance, while variance measures the volatility of the deduction result data. Smaller price variance indicates that the market price is relatively stable, while larger unit output variance may indicate that the performance of a specific unit in different cases fluctuates greatly. The application of these statistical results not only helps quantify market performance, but also can evaluate the power market performance of the "Extremely Hot and No Wind" case group, providing quantitative support for research on high-proportion renewable energy participation in power market operations. According to the sample characteristics of new energy prediction errors in different forms such as extreme heat and no wind, a Gaussian distribution is fitted, and multiple groups of prediction error data under this scenario are generated through sampling. Then, multiple groups of case data are generated by combining other invariants in the case data, thus forming a complete set of multiple groups of deduction case data based on the characteristics of the specified scenario.

[0095] In order to solve the limitations of the existing technology in the simulation and deduction of multi-scenario electricity markets, this application proposes a method for generating multi-scenario electricity market deduction and related devices. This method can generate output curves of new energy units according to different target deduction scenario requirements, and combine these data with known quantities as boundary conditions for electricity market simulation and deduction. Through this technology, continuous simulation and deduction of multi-scenario electricity markets can be supported, providing comprehensive quantitative evaluation support for the analysis and verification of policies, rules and operating mechanisms.

[0096] In the prediction and uncertainty modeling of renewable energy output, the prediction error is usually concentrated near the predicted value, with certain peak characteristics, and may show large deviations due to randomness and volatility. To this end, this application uses Gaussian distribution to statistically model the renewable energy prediction error, and accurately characterizes the probability distribution characteristics of the error through mean and standard deviation parameters. Gaussian distribution has a clear theoretical basis and flexible applicability, which can effectively describe the main characteristics of renewable energy prediction errors, and is convenient for mathematical processing and sampling generation.

[0097] In order to generate diverse scenario data, this application combines Latin hypercube sampling technology to sample from the fitted Gaussian distribution to generate new energy output data and system load forecast data covering multiple scenarios. Latin hypercube sampling can evenly cover the sample space within the uncertainty range, ensuring that the deduced data includes both conventional scenarios and low-probability extreme scenarios, thereby improving the adaptability and comprehensiveness of power market simulation deduction to complex operating conditions. Through the modeling and data generation method that combines Gaussian distribution with Latin hypercube sampling, this application can accurately characterize the uncertainty of new energy output, provide a scientific and comprehensive input basis for power market simulation deduction, and effectively support continuous simulation and analysis verification of multiple scenarios under different deduction goals and requirements.

[0098] like Figure 5 As shown, it is a schematic diagram of a multi-scenario generation system for power market simulation of the present application, which may include: Data module, used to obtain historical operation data; An error preprocessing module, used for performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data with a histogram to obtain an error histogram; A Gaussian fitting module, used to superimpose a Gaussian distribution fitting curve on the error histogram, and obtain a corresponding error probability density function according to the fitting result; A statistical module, used for calculating a cumulative distribution function corresponding to a Gaussian distribution according to the error probability density function; A sampling module, used for sampling the cumulative distribution function by using a Latin hypercube probability sampling method to obtain a corresponding combined sample; A superposition module, used for superimposing the combined sample onto the historical operation data to obtain forecast data for power market deduction; A deduction module is used to generate multiple scenarios of power market deduction based on the forecast data.

[0099] In some embodiments of the multi-scenario generation system for power market simulation of the present application, the error preprocessing module includes an abnormal data correction unit, a missing value supplement unit, a data normalization processing unit and a histogram visualization unit; the abnormal data correction unit is used to perform abnormal data correction on the historical operation data; the missing value supplement unit is used to supplement missing values ​​for the data after abnormal data correction; the data normalization processing unit is used to perform data normalization on the data after missing value supplementation; the histogram visualization unit is used to visualize the preprocessed operation data with a histogram to obtain an error histogram.

[0100] In some embodiments of the multi-scenario generation system for power market simulation of the present application, the abnormal data correction unit is used to screen outliers using the Raida criterion; the abnormal data correction unit is also used to correct the abnormal data of the outliers using at least one of the following methods: filling the outliers with the mean, correcting the missing values ​​by treating the outliers as missing data, replacing the outliers with similar days, and deleting the outliers.

[0101] In some embodiments of the multi-scenario generation system for power market simulation of the present application, the missing value supplement unit is used to supplement the missing values ​​using the Lagrange interpolation method.

[0102] In some embodiments of the power market deduction multi-scenario generation system of the present application, the power market includes wind power and photovoltaic power; The Gaussian fitting module uses the following formula to calculate the corresponding error probability density function according to the fitting result:

[0103] in, is the error probability density function of wind power, for t The error value of photovoltaic prediction at the moment, for t The standard deviation of the photovoltaic prediction error at the moment, for t The expected value of the photovoltaic prediction error at time is the error probability density function of photovoltaic, for t The error value of wind power forecast at the moment, for t The standard deviation of wind power forecast error at each moment, for tThe expected value of wind power forecast error at that moment.

[0104] In some embodiments of the power market simulation multi-scenario generation system of the present application, the statistical module uses the following formula to calculate the cumulative distribution function corresponding to the Gaussian distribution: .

[0105] In some embodiments of the multi-scenario generation system for power market simulation of the present application, the sampling module includes an equal division unit, a value selection unit, an inverse transformation sampling unit and an arrangement unit; The dividing unit is used to divide the cumulative distribution function into n intervals to obtain a set T of all intervals; wherein n is the number of preset generated scenes; The value selection unit is used to randomly select a uniformly distributed value in each of the n intervals; The inverse transform sampling unit is used to map all the selected uniformly distributed values ​​to the Gaussian distribution through the inverse transform sampling method to obtain sample points of the Gaussian distribution; The arrangement unit is used to arrange the prediction errors of wind power and photovoltaic power according to the sample points of Gaussian distribution and according to the time period to form a combined sample.

[0106] It should be noted that in the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the system embodiments described above are merely schematic. For example, the division of each module is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules can be combined or integrated into another device, or some features can be ignored or not executed. The module described as a separate component may or may not be physically separated. The component displayed as a module may be a physical unit or multiple physical units, that is, it may be located in one place, or it may be distributed in multiple different places. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0107] In addition, each module in each embodiment of the present invention may be integrated into a processing unit, each module may exist physically separately, or two or more modules may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0108] An embodiment of the present application also provides an electronic device, which may include one or more processors, a memory, and a communication interface.

[0109] The memory, the communication interface and the processor are coupled, for example, the memory, the communication interface and the processor may be coupled together via a bus.

[0110] The communication interface is used for data transmission with other devices. The memory stores computer program code. The computer program code includes computer instructions, and when the computer instructions are executed by the processor, the electronic device executes the steps of the above-mentioned method for generating multiple scenarios for power market simulation.

[0111] Wherein, the processor can be a processor or a controller, for example, a central processing unit (CPU), a general processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the present disclosure. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of DSP and microprocessors, and the like. The processor can be used to support electronic devices to execute the method steps provided in the above embodiments.

[0112] The bus may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The above bus may be divided into an address bus, a data bus, a control bus, etc.

[0113] An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for generating multiple scenarios for power market simulation are implemented.

[0114] The computer-readable storage medium involved in the present application includes random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the technical field.

[0115] For the description of the relevant parts of the power market deduction multi-scenario generation system, electronic device and computer-readable storage medium provided in the embodiment of the present application, please refer to the detailed description of the corresponding parts in the power market deduction multi-scenario generation method provided in the embodiment of the present application, which will not be repeated here. In addition, the parts of the above-mentioned technical solutions provided in the embodiment of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive elaboration.

[0116] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for generating multiple scenarios for power market simulation, characterized in that: include: Get historical operation data; Performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data using a histogram to obtain an error histogram; Superimposing a Gaussian distribution fitting curve on the error histogram, and obtaining a corresponding error probability density function according to the fitting result; Calculate the cumulative distribution function corresponding to the Gaussian distribution according to the error probability density function; Using Latin hypercube probability sampling method, sampling the cumulative distribution function to obtain corresponding combined samples; Superimposing the combined sample on the historical operation data to obtain forecast data for power market deduction; Multiple scenarios for power market simulation are generated based on the forecast data.

2. The method for generating multiple scenarios for power market simulation according to claim 1 is characterized in that: The error preprocessing of the historical operation data includes abnormal data correction, missing value supplementation and data normalization processing.

3. The method for generating multiple scenarios for power market simulation according to claim 2 is characterized in that: The abnormal data correction includes: Raida criterion was used to screen outliers; Use at least one of the following methods to correct outliers: fill in the mean of outliers, correct missing values ​​by treating outliers as missing data, and correct outliers. Replace similar days and delete outliers.

4. The method for generating multiple scenarios for power market simulation according to claim 2 is characterized in that: The missing value supplementation includes: using Lagrange interpolation method to supplement the missing values.

5. The method for generating multiple scenarios for power market simulation according to claim 1 is characterized in that: The electricity market includes wind power and photovoltaics; The formula for obtaining the corresponding error probability density function according to the fitting result is as follows: in, is the error probability density function of wind power, for t The error value of photovoltaic prediction at the moment, for t The standard deviation of the photovoltaic prediction error at the moment, for t The expected value of the photovoltaic prediction error at time is the error probability density function of photovoltaic, for t The error value of wind power forecast at the moment, for t The standard deviation of wind power forecast error at each moment, for t The expected value of wind power forecast error at that moment.

6. The method for generating multiple scenarios for power market simulation according to claim 5 is characterized in that: The cumulative distribution function corresponding to the Gaussian distribution is calculated using the following formula: in, is the cumulative distribution function corresponding to the Gaussian distribution obeyed by the wind power prediction error, is the cumulative distribution function corresponding to the Gaussian distribution obeyed by the photovoltaic prediction error.

7. The method for generating multiple scenarios for power market simulation according to claim 6 is characterized in that: The method of using Latin hypercube probability sampling to sample the cumulative distribution function to obtain a corresponding combined sample includes: Divide the cumulative distribution function into n intervals, and obtain a set T of all intervals; where n is the number of preset generated scenes; Randomly select a uniformly distributed value in each of the n intervals; All the uniformly distributed values ​​selected are mapped to the Gaussian distribution through the inverse transform sampling method to obtain the sample points of the Gaussian distribution; According to the sample points of Gaussian distribution, the prediction errors of wind power and photovoltaic power are arranged according to the time period to form a combined sample.

8. The method for generating multiple scenarios for power market simulation according to claim 1, characterized in that: The generating multiple power market simulation scenarios according to the forecast data includes: A distributed computing module is used to perform parallel calculations on each set of forecast data to generate corresponding power market simulation scenarios.

9. The method for generating multiple scenarios for power market simulation according to claim 8, characterized in that: Before using the distributed computing module to perform calculations on each set of prediction data in parallel, the method further includes: packaging the prediction data and the tasks for performing the calculations into a mirror image.

10. A multi-scenario generation system for power market simulation, characterized in that: include: Data module, used to obtain historical operation data; An error preprocessing module, used for performing error preprocessing on the historical operation data to obtain preprocessed operation data, and visualizing the preprocessed operation data with a histogram to obtain an error histogram; A Gaussian fitting module, used to superimpose a Gaussian distribution fitting curve on the error histogram, and obtain a corresponding error probability density function according to the fitting result; A statistical module, used for calculating a cumulative distribution function corresponding to a Gaussian distribution according to the error probability density function; A sampling module, used for sampling the cumulative distribution function by using a Latin hypercube probability sampling method to obtain a corresponding combined sample; A superposition module, used for superimposing the combined sample onto the historical operation data to obtain forecast data for power market deduction; A deduction module is used to generate multiple scenarios of power market deduction based on the forecast data.

11. The power market simulation multi-scenario generation system according to claim 10, characterized in that: The error preprocessing module includes an abnormal data correction unit, a missing value supplement unit, a data normalization processing unit and a histogram visualization unit; The abnormal data correction unit is used to correct the abnormal data of the historical operation data; The missing value supplement unit is used to supplement the missing values ​​of the data after the abnormal data is corrected; The data normalization processing unit is used to perform data normalization processing on the data after missing value supplementation; The histogram visualization unit is used to visualize the preprocessed operation data using a histogram to obtain an error histogram.

12. The power market simulation multi-scenario generation system according to claim 11, characterized in that: The abnormal data correction unit is used to filter out abnormal values ​​using the Raida criterion; The abnormal data correction unit is also used to Use at least one of the following methods to correct outliers: Fill the outliers with the mean, treat the outliers as missing data for missing value correction, replace the outliers with similar days, and delete the outliers.

13. The power market simulation multi-scenario generation system according to claim 11, characterized in that: The missing value supplement unit is used to supplement the missing values ​​by using the Lagrange interpolation method.

14. The power market simulation multi-scenario generation system according to claim 10, characterized in that: The electricity market includes wind power and photovoltaics; The Gaussian fitting module uses the following formula to calculate the corresponding error probability density function according to the fitting result: in, is the error probability density function of wind power, for t The error value of photovoltaic prediction at the moment, for t The standard deviation of the photovoltaic prediction error at the moment, for t The expected value of the photovoltaic prediction error at time is the error probability density function of photovoltaic, for t The error value of wind power forecast at the moment, for t The standard deviation of wind power forecast error at each moment, for t The expected value of wind power forecast error at that moment.

15. The power market simulation multi-scenario generation system according to claim 14, characterized in that: The statistical module uses the following formula to calculate the cumulative distribution function corresponding to the Gaussian distribution: in, is the cumulative distribution function corresponding to the Gaussian distribution obeyed by the wind power prediction error, is the cumulative distribution function corresponding to the Gaussian distribution obeyed by the photovoltaic prediction error.

16. The power market simulation multi-scenario generation system according to claim 15, characterized in that: The sampling module includes an equal division unit, a value selection unit, an inverse transformation sampling unit and an arrangement unit; The dividing unit is used to divide the cumulative distribution function into n intervals to obtain a set T of all intervals; wherein n is the number of preset generated scenes; The value selection unit is used to randomly select a uniformly distributed value in each of the n intervals; The inverse transform sampling unit is used to map all the selected uniformly distributed values ​​to the Gaussian distribution through the inverse transform sampling method to obtain sample points of the Gaussian distribution; The arrangement unit is used to arrange the prediction errors of wind power and photovoltaic power according to the sample points of Gaussian distribution and according to the time period to form a combined sample.

17. An electronic device, characterized in that: include: A memory and one or more processors; the memory is coupled to the processor; wherein the memory stores computer program code, the computer program code includes computer instructions, and when the computer instructions are executed by the processor, the electronic device executes the steps of the method for generating multiple scenarios for power market simulation as described in any one of claims 1 to 9.

18. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for generating multiple scenarios for power market simulation as described in any one of claims 1 to 9.

Images