Power system dispatching method and program product based on source load joint probability modeling and adaptive opportunity constraint

Abstract

The application discloses a power system scheduling method and program product based on source-load joint probability modeling and adaptive opportunity constraint, and comprises the following steps: obtaining historical operation data in a power system, constructing a source-load joint probability distribution model among wind power output, photovoltaic output and load demand, and calculating the correlation parameters of each time in the source-load joint probability distribution model; predicting the correlation parameters of a future time according to the correlation parameter values of each time in the past, and then predicting the source-load comprehensive correlation index of the future time, determining the default probability of each time in the future, and constructing an adaptive opportunity constraint which dynamically adjusts with the change of the default probability; establishing an optimization scheduling model with constraint conditions at least including the adaptive opportunity constraint, and solving the optimization scheduling model to obtain an optimization scheduling strategy. The application introduces an adaptive opportunity constraint mechanism, dynamically adjusts the constraint confidence level and the scheduling strategy, and effectively improves the economy and safety and reliability of the power system operation.

Description

Technical Field

[0001] This invention relates to the field of power system optimization scheduling technology, and in particular to a power system scheduling method and program product based on source-load joint probability modeling and adaptive chance constraints. Background Technology

[0002] With the rapid development of new energy power generation technologies, the penetration rate of renewable energy sources such as wind power and photovoltaics in the power system is constantly increasing. The power system is gradually shifting from a structure dominated by traditional controllable power sources to a structure characterized by a high proportion of uncertain power sources. Compared with conventional units, wind power and photovoltaic output have obvious randomness and volatility, while load demand also exhibits complex time-varying characteristics. This makes the power system face greater uncertainty challenges during operation. Against this backdrop, how to accurately characterize source and load uncertainties and effectively respond to them during dispatching has become an important research direction in the field of power system optimization operation.

[0003] Existing technologies for handling uncertainty mainly include deterministic methods, robust optimization methods, and stochastic optimization methods. Among these, chance-constrained programming (CCP) is widely used in power system dispatching problems because it can control the probability of constraint default under certain confidence levels. However, traditional CCP methods are usually based on independence or simple correlation assumptions, which are insufficient to characterize the complex relationships between wind power, photovoltaic power, and loads, and fail to reflect the coupling characteristics between multiple variables. Furthermore, traditional methods often use fixed confidence levels to control risk. This static setting cannot adapt to the dynamic changes in system operating conditions, and it is easy to make it difficult to achieve a balance between safety and economy in dispatching results.

[0004] On the other hand, the Copula function provides an effective tool for characterizing the nonlinear correlation structure among multiple variables. It can flexibly construct joint distribution models while preserving the marginal distribution characteristics of each variable, and has been increasingly applied in the field of power system uncertainty modeling in recent years. However, existing Copula-based research mostly focuses on static correlation analysis, lacking in-depth consideration of the time-varying characteristics of correlation, and rarely embedding joint distribution modeling results directly into scheduling optimization constraints for unified modeling. Therefore, how to introduce a dynamic correlation mechanism based on characterizing the source-load joint distribution, and construct an opportunity-constrained scheduling method that can adaptively adjust the risk level according to the system state, still requires further research and improvement. Summary of the Invention

[0005] To address the problems existing in the prior art, the purpose of this invention is to provide a scheduling method that, based on characterizing the joint distribution of source and load, introduces a dynamic correlation mechanism and constructs a scheduling method that can adaptively adjust the risk level and opportunity constraints according to the system state, thereby improving the safety and economy of the scheduling method.

[0006] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0007] A power system dispatching method based on joint source-load probabilistic modeling and adaptive chance constraints includes the following steps:

[0008] S1. Obtain historical operating data of wind power output, photovoltaic power output and load demand in the power system, and establish marginal probability distribution models for each type of historical operating data.

[0009] S2. Based on the marginal probability distribution model of various types of historical operation data, construct a source-load joint probability distribution model between wind power output, photovoltaic power output and load demand, and calculate the correlation parameters of each historical moment in the source-load joint probability distribution model.

[0010] S3. Based on the correlation parameter values ​​at each historical moment, a correlation parameter estimation model is learned, and the correlation parameters at future moments are predicted based on the correlation parameter estimation model, thereby predicting the source-load integrated correlation index at future moments; wherein, in the correlation parameter estimation model, the correlation parameter value at each moment is a weighted sum of the correlation parameter value at the previous moment and the local observation value of the correlation parameter calculated based on the local operating data at the current moment; the larger the correlation parameter value, the larger the absolute value of the source-load integrated correlation index;

[0011] S4. Based on the comprehensive correlation index of source and load, determine the probability of default at each future time and construct an adaptive opportunity constraint that considers the random fluctuation of source and load and dynamically adjusts with the change of default probability.

[0012] S5. Establish an optimized scheduling model, wherein the decision variable of the optimized scheduling model is the output of conventional units at each future time, the optimization objective is to minimize the system operating cost within the scheduling time, and the constraints include at least the adaptive opportunity constraints.

[0013] S6. Solve the optimal scheduling model to obtain the optimal scheduling strategy.

[0014] Furthermore, step S1 specifically includes the following steps:

[0015] S11. Obtain historical operating data reflecting the operating status of the power system. Collect wind power output, photovoltaic power output and load demand in the historical operating data according to time. After abnormal data processing and time scale unification, organize them into a historical source-load dataset reflecting the historical source-load combination operating status.

[0016] S12. Based on the historical source-load dataset, calculate the probability density functions of wind power output, photovoltaic power output and load demand respectively, and establish marginal probability distribution models of wind power output, photovoltaic power output and load demand based on the probability density functions respectively.

[0017] Furthermore, step S2 specifically includes the following steps:

[0018] S21. Based on the edge probability distribution model of wind power output, photovoltaic power output and load demand, a source-load joint probability distribution model is constructed using the Copula function;

[0019] S22. Establish a log-likelihood function based on the source-load joint probability distribution model, and solve for the correlation parameter values ​​with the objective of maximizing the log-likelihood function.

[0020] Furthermore, step S3 specifically includes the following steps:

[0021] S31. For historical operating data of wind power output, photovoltaic power output and load demand, data is collected using a preset sliding window before each time point, so as to extract a historical data segment at each time point.

[0022] S32. Based on the historical data segment at each time point, calculate the correlation parameter at each time point to obtain the local observed value of the correlation parameter;

[0023] S33. Based on the correlation parameter values ​​at each historical moment, a correlation parameter estimation model is learned, wherein the correlation parameter value at each moment in the correlation parameter estimation model is a weighted sum of the correlation parameter value at the previous moment and the local observation value of the correlation parameter at the current moment.

[0024] S34. Based on historical correlation parameter values, a correlation parameter estimation model is used to predict the correlation parameter values ​​at future times.

[0025] S35. Based on the correlation parameter values ​​at future times, determine the comprehensive source-load correlation index at future times. The larger the correlation parameter value, the larger the absolute value of the comprehensive source-load correlation index.

[0026] Furthermore, step S4 specifically includes the following steps:

[0027] S41. Construct an opportunity constraint, wherein the opportunity constraint is the difference between the probability of random fluctuation of source load being greater than or equal to 1 and the probability of default.

[0028] S42. Based on the comprehensive correlation index of source and load, the probability of default is dynamically determined. The larger the absolute value of the comprehensive correlation index of source and load, the smaller the probability of default.

[0029] S43. Based on the dynamic default probability, the opportunity constraint is transformed into a deterministic constraint that changes with the comprehensive correlation index of source and load, and serves as an adaptive opportunity constraint.

[0030] Furthermore, the adaptive opportunity constraint is:

[0031]

[0032]

[0033] Where g(x,t) represents the adaptive chance constraint, x represents the decision variable, and t represents time t. , Indicates the standard deviation and variance of the joint uncertainty variables; The inverse cumulative distribution function of the standard normal distribution; Indicates the quantile corresponding to the confidence level. Let represent the probability of default at time t. This represents the uncertainty variable caused by random fluctuations in the source load. This indicates the expectation, which is calculated using the source-load joint probability distribution model.

[0034] Furthermore, the optimized scheduling model described in step S5 is as follows:

[0035]

[0036] st

[0037] Where T represents the set of scheduling times; This represents the output of the conventional generating unit at time t; Indicates the amount of wind and solar power curtailed; This represents the power generation cost function of a conventional generating unit; This represents the cost function for penalties related to wind and solar power curtailment.

[0038] Furthermore, step S6 specifically includes the following steps:

[0039] S61. Unify the objective function and constraints of the optimization scheduling model into a standard optimization form, call the optimization solver to perform calculations, and obtain the optimal solution for the decision variables;

[0040] S62. Transform the optimal solution of the decision variables into an executable scheduling strategy for the power system.

[0041] A computer program product includes a computer program that, when executed by a processor, implements the above-described method.

[0042] A computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the above-described method.

[0043] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention innovatively addresses the problems of strong source-load uncertainty, insufficient correlation characterization, and fixed risk levels in traditional opportunity constraint methods in high-proportion renewable energy power systems. It effectively solves the problems of difficulty in unified modeling of multi-source uncertainty and the difficulty in balancing the conservatism and security of scheduling constraints, achieving significant results. Specifically, to address the difficulty in accurately describing the randomness and correlation of wind power, photovoltaics, and loads, this invention first proposes a source-load joint distribution modeling method based on Copula functions. While maintaining the marginal distribution characteristics of each variable, it achieves accurate characterization of the multivariate correlation structure. Second, to address the problem that traditional methods ignore the time-varying characteristics of correlation, a dynamic correlation modeling mechanism is constructed. By introducing time-varying Copula correlation parameters, the dynamic evolution of source-load correlation relationships is described. Furthermore, this invention proposes an adaptive opportunity constraint method based on dynamic correlation. It introduces correlation indicators into the default probability design to achieve dynamic adjustment of risk levels and quantifies uncertainty by combining joint distribution, transforming probabilistic constraints into a solvable deterministic form. Finally, it is embedded in the scheduling optimization model for solution, thereby significantly improving scheduling economy and renewable energy absorption capacity while ensuring system security. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating a power system scheduling method based on source-load joint probability modeling and adaptive chance constraints provided in an embodiment of the present invention.

[0045] Figure 2 This is a schematic diagram of the scheduling logic of a power system scheduling method based on source-load joint probability modeling and adaptive chance constraints provided by an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0047] Example 1

[0048] This invention provides a power system scheduling method based on joint source-load probabilistic modeling and adaptive chance constraints, such as... Figure 1 and Figure 2 As shown, it includes the following steps:

[0049] S1. Obtain historical operating data of wind power output, photovoltaic power output and load demand in the power system, and establish marginal probability distribution models for each type of historical operating data.

[0050] Step S1 specifically includes the following steps:

[0051] S11. Obtain historical operating data reflecting the power system's operating status. Collect wind power output, photovoltaic power output, and load demand data from the historical operating data according to time, including wind power output at time t. Photovoltaic power output Load demand After anomaly processing and time-scale unification, the data was compiled into a historical source-load dataset reflecting the historical source-load combination operation status. , where i represents the i-th historical sample, N represents the total number of samples, and each sample represents the source-load combination operation status at the same moment.

[0052] S12. Based on the historical source-load dataset, calculate the probability density functions of wind power output, photovoltaic power output and load demand respectively, and establish marginal probability distribution models of wind power output, photovoltaic power output and load demand based on the probability density functions respectively.

[0053] Specifically, marginal probability distribution models are performed for wind power output, photovoltaic power output, and load demand to describe the uncertainty characteristics of each variable. For any variable X (X=w,pv,L, where w,pv,L represent wind power, photovoltaic power, and load, respectively), its marginal cumulative distribution function is defined as:

[0054]

[0055] in, It represents the probability that variable X is less than or equal to x; Represents probability operators.

[0056] In this invention, the empirical distribution function method or the kernel density estimation method is preferably used for marginal distribution fitting. This embodiment takes the kernel density estimation method as an example; the probability density function of variable X is estimated as follows:

[0057]

[0058] in, The estimated probability density function is given by: h represents the bandwidth parameter, used to control the smoothness of the distribution curve; X i This represents the value of the i-th sample. The kernel function is usually chosen, and the Gaussian kernel function is typically selected.

[0059] Furthermore, the marginal cumulative distribution function can be obtained by integration:

[0060]

[0061] After obtaining the marginal distribution function, a probability integral transformation is performed on the original variables to map them to the standard uniform distribution space:

[0062]

[0063]

[0064]

[0065] in, , where represent the probability values ​​corresponding to wind power, photovoltaic, and load samples, respectively.

[0066] S2. Based on the marginal probability distribution model of various types of historical operating data, construct a source-load joint probability distribution model between wind power output, photovoltaic power output and load demand, and calculate the correlation parameters of each historical moment in the source-load joint probability distribution model.

[0067] Step S2 specifically includes the following steps:

[0068] S21. Based on the edge probability distribution model of wind power output, photovoltaic power output and load demand, a source-load joint probability distribution model is constructed using the Copula function;

[0069] Specifically, a multivariate joint distribution model of source and load is constructed based on the Copula function to achieve decoupled modeling of marginal distributions and related structures. According to Sklar's theorem, for any multidimensional joint distribution function... Both can be represented as:

[0070]

[0071] in, This represents the joint cumulative distribution function of wind power, solar power, and load. This represents the joint cumulative distribution function of wind power, solar power, and load. The correlation parameter represents the Copula function; , , These are the marginal distribution function values ​​obtained in step S1.

[0072] The variable after the probability integral transformation is defined as:

[0073]

[0074]

[0075]

[0076] in, This means mapping the original physical quantities to a unified probability space.

[0077] The joint probability density function can be expressed as:

[0078]

[0079] in, Denotes the joint probability density function; Denotes the Copula density function, and ; , , This represents the marginal probability density function.

[0080] In this invention, the Copula function is preferably a Gaussian Copula function, the expression of which is:

[0081]

[0082] in, The inverse cumulative distribution function of the standard normal distribution; The covariance matrix is ​​represented as The multivariate normal distribution function; This is the correlation coefficient matrix, used to describe the linear correlation between variables. When using a Gaussian Copula, the correlation parameter... equal to the correlation coefficient matrix ,and

[0083]

[0084] in, This represents the correlation coefficient between wind power and photovoltaic power. This represents the correlation coefficient between wind power and photovoltaic power. This represents the correlation coefficient between wind power and load.

[0085] S22. Establish a log-likelihood function based on the source-load joint probability distribution model, and solve for the correlation parameter values ​​with the objective of maximizing the log-likelihood function.

[0086] Specifically, after constructing the Copula model, it is necessary to analyze the correlation parameters within it. Estimation is performed to ensure the model closely matches historical data, i.e., the parameters in the Copula function are determined. This minimizes the deviation between the joint distribution and the actual observed data. Construct the log-likelihood function:

[0087]

[0088] The parameter estimation problem is transformed into:

[0089]

[0090] in, Represents the log-likelihood function; This represents the Copula density function; N is the number of samples; It is an estimate of the correlation parameter; This indicates the parameter that makes the function take its maximum value.

[0091] S3. Based on the correlation parameter values ​​at each historical moment, learn the correlation parameter estimation model, and predict the correlation parameters at future moments based on the correlation parameter estimation model, thereby predicting the source-load comprehensive correlation index at future moments.

[0092] To better reflect the temporal correlation of the correlation parameters, the Copula correlation parameters at various historical moments are... The static form is extended to a time-dependent form to reflect the characteristics of source-load correlation changing over time. The correlation parameters are... Represented as a time function That is, the Copula correlation parameter at time t is When using the Gaussian Copula function, ,and

[0093]

[0094] in, This represents the correlation coefficient between wind power and photovoltaic power at time t; This represents the correlation coefficient between wind power and photovoltaic power at time t; This represents the correlation coefficient between wind power and load at time t.

[0095] Step S3 specifically includes the following steps:

[0096] S31. For historical operating data of wind power output, photovoltaic power output and load demand, data is collected using a preset sliding window before each time point, so as to extract a historical data segment at each time point.

[0097] S32. Based on the historical data segments at each time point, calculate the correlation parameter for each time point to obtain the local observed value of the correlation parameter. ;

[0098] Specifically, the calculation of the correlation parameter is the same as in step S22, also using the log-likelihood function, only the data segment is different. The specific formula is as follows:

[0099]

[0100] in, This represents a sliding time window centered at the current time t, indicating that only the most recent data is used to estimate the current correlation.

[0101] S33. Based on the correlation parameter values ​​at each historical moment, a correlation parameter estimation model is learned.

[0102] For time-varying parameters The modeling of change patterns describes how correlation parameters evolve from historical states to the current state, enabling them to be updated recursively over time, achieving dynamic updates. The recursive correlation parameter estimation model is as follows:

[0103]

[0104] in, This indicates the correlation parameter at the current time. This represents the correlation parameter at the previous time step; These are observation parameters estimated based on the current window data. Indicates the update coefficient, when When the data is larger, it relies more on current data and responds more sensitively. When the data is smaller, it relies more on historical data and the changes are more stable.

[0105] S34. Based on historical correlation parameter values, a correlation parameter estimation model is used to predict the correlation parameter values ​​at future times.

[0106] S35. Determine the source-load comprehensive correlation index for future time periods based on the correlation parameter values ​​at future time periods.

[0107] Specifically, the comprehensive correlation index of source and load Represented as ,in This represents the mapping function. The value range of this index is typically [-1, 1]. The larger the correlation parameter value, the larger the absolute value of the source-load comprehensive correlation index. When... The time source and the load change in the same direction, when The time source and the load change in opposite directions. The larger the value, the stronger the source-load correlation.

[0108] S4. Based on the comprehensive correlation index of source and load, determine the probability of default at each future time and construct an adaptive opportunity constraint that considers the random fluctuation of source and load and dynamically adjusts with the change of default probability.

[0109] S4 specifically includes the following steps:

[0110] S41. Construct an opportunity constraint, wherein the opportunity constraint is the difference between the probability of random fluctuations in the source load being greater than or equal to 1 and the probability of default; specifically:

[0111]

[0112] in, Represents probability operators; Represents constraint functions, used to describe system operating constraints, such as power balance and line power flow limitations; This represents scheduling decision variables, such as unit output; It is an uncertain variable, representing random fluctuations in the source load; This represents the probability of default, and is usually a fixed constant.

[0113] In traditional methods, Using pre-set fixed values ​​cannot reflect the dynamic changes in source-load correlation during actual operation, and has the following shortcomings: First, when source-load correlation is strong, the actual uncertainty of the system is low, but the fixed values... This can lead to overly conservative constraints; secondly, when the source and load are weakly or negatively correlated, the system faces greater volatility risk, but the fixed... This could also lead to insufficient safety margins. Therefore, fixed-chance constraints make it difficult to balance the economy and safety of a system.

[0114] S42. Based on the comprehensive correlation index of source and load, the probability of default is dynamically determined. The larger the absolute value of the comprehensive correlation index of source and load, the smaller the probability of default.

[0115] Specifically, based on the strength of the correlation between source and payload, a dynamic confidence level is set to dynamically adjust the probability of default. This allows the conservatism of the constraints to automatically adjust as the system state changes. The time-dependent default probability... Defined as:

[0116]

[0117] in, Let represent the dynamic default probability at time t; Indicates the baseline probability of default; It is an adjustment coefficient used to control the degree of influence of correlation on risk; This represents a comprehensive correlation index between source and load. This represents the absolute value of the correlation, indicating the strength of the correlation. When... hour, This indicates that the system has low uncertainty, which can reduce the degree of conservatism and improve economic efficiency; when hour, This indicates that the system has high uncertainty and the safety margin should be increased.

[0118] S43. Based on the dynamic default probability, the opportunity constraint is transformed into a deterministic constraint that changes with the comprehensive correlation index of source and load, and serves as an adaptive opportunity constraint.

[0119] Specifically, after obtaining the dynamic default probability, the probabilistic constraints are transformed into deterministic constraints, allowing them to be directly embedded into the scheduling optimization model for solution. First, the uncertain variables are... Defined as:

[0120]

[0121] in, , , Let represent the wind power output, photovoltaic power output, and load demand at time t, respectively. A positive value indicates that power generation exceeds load. A negative value indicates insufficient power supply.

[0122] In joint distribution Next, define the expectation operator. , indicating that for the joint density function The integral operation. Then the expectation... and variance The calculation is as follows:

[0123]

[0124]

[0125] The variance can be expanded as follows:

[0126]

[0127] in, Indicates in joint distribution The variance below; This represents the covariance under the same joint distribution, which is determined by the correlation parameter in the Copula model.

[0128] Transform the opportunity constraint into:

[0129]

[0130] in, Indicates the standard deviation of joint uncertainty; The inverse cumulative distribution function of the standard normal distribution; This indicates the quantile corresponding to the confidence level.

[0131] S5. Establish an optimized scheduling model, wherein the decision variables of the optimized scheduling model are the output of conventional units at each future time, the optimization objective is to minimize the system operating cost within the scheduling time, and the constraints include at least the adaptive opportunity constraints.

[0132] The optimized scheduling model is as follows:

[0133]

[0134] st

[0135] Where T represents the set of scheduling times; This represents the output of the conventional generating unit at time t; Indicates the amount of wind and solar power curtailed; This represents the power generation cost function of a conventional generating unit; This represents the cost function for penalizing wind and solar power curtailment.

[0136] Electricity generation cost function It can be represented as:

[0137]

[0138] Where a, b, and c represent the unit cost coefficients.

[0139] Costs of wind and solar power curtailment It can be represented as:

[0140]

[0141] in, This represents the curtailment penalty coefficient, used to reflect the principle of prioritizing the consumption of new energy sources.

[0142] The power balance constraint is expressed as:

[0143]

[0144] in, , , Let represent the wind power output, photovoltaic power output, and load demand at time t, respectively.

[0145] The unit output constraint is expressed as:

[0146]

[0147] in, This indicates the unit's minimum output. This indicates the maximum output of the generator unit.

[0148] The unit ramp-up constraint is expressed as:

[0149]

[0150] in, This indicates the unit's maximum climbing ability.

[0151] The adaptive chance constraint is:

[0152]

[0153] S6. Solve the optimal scheduling model to obtain the optimal scheduling strategy.

[0154] Step S6 specifically includes the following steps:

[0155] S61. Consolidate the objective function and constraints of the optimization scheduling model into a standard optimization form, call the optimization solver to perform calculations, and obtain the optimal solution for the decision variables. ;in, , These represent the optimal unit output and the optimal amount of wind and solar curtailment, respectively.

[0156] S62. Transform the optimal solution of the decision variables into an executable scheduling strategy for the power system.

[0157] Analysis of power allocation results for each time period and To determine the output arrangement of each unit and the level of renewable energy consumption; to verify whether dynamic constraints are met. To determine whether the scheduling plan meets the expected risk level; to calculate It is used to assess the size of the system's safety margin.

[0158] Control commands are generated based on the optimization results, including commands to increase unit output. Send the command to each generator set for execution; when there is power wastage... If so, the issuance of instructions will be limited and the strategy adjustment plan will be prioritized for consumption.

[0159] Example 2

[0160] This invention also provides a computer program product, such as an app on a mobile phone or tablet, or an installer on a computer. This product includes a computer program / instructions that, when executed by a processor, implement the method described in Embodiment 1. The code for the computer-executable program used to perform the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0161] Example 3

[0162] This invention provides a storage medium containing a computer-executable program, which, when executed by a computer processor, is used to perform the method of Embodiment 1.

[0163] The storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0164] Of course, the computer-executable program in the storage medium provided in the embodiments of the present invention is not limited to the above-described method operations, but can also perform related operations in the methods provided in any embodiment of the present invention.

[0165] It should be understood that the embodiments and descriptions above are only the principles, main features and advantages of the present invention. Various changes and modifications can be made to the present invention without departing from the spirit and scope of the invention, and all such changes and modifications fall within the protection scope of the present invention.

Claims

1. A power system dispatching method based on joint source-load probabilistic modeling and adaptive chance constraints, characterized in that, Includes the following steps: S1. Obtain historical operating data of wind power output, photovoltaic power output and load demand in the power system, and establish marginal probability distribution models for each type of historical operating data. S2. Based on the marginal probability distribution model of various types of historical operation data, construct a source-load joint probability distribution model between wind power output, photovoltaic power output and load demand, and calculate the correlation parameters of each historical moment in the source-load joint probability distribution model. S3. Based on the correlation parameter values ​​at each historical moment, a correlation parameter estimation model is learned, and the correlation parameters at future moments are predicted based on the correlation parameter estimation model, thereby predicting the source-load integrated correlation index at future moments; wherein, in the correlation parameter estimation model, the correlation parameter value at each moment is a weighted sum of the correlation parameter value at the previous moment and the local observation value of the correlation parameter calculated based on the local operating data at the current moment; the larger the correlation parameter value, the larger the absolute value of the source-load integrated correlation index; S4. Based on the comprehensive correlation index of source and load, determine the probability of default at each future time and construct an adaptive opportunity constraint that considers the random fluctuation of source and load and dynamically adjusts with the change of default probability. S5. Establish an optimized scheduling model, wherein the decision variable of the optimized scheduling model is the output of conventional units at each future time, the optimization objective is to minimize the system operating cost within the scheduling time, and the constraints include at least the adaptive opportunity constraints. S6. Solve the optimal scheduling model to obtain the optimal scheduling strategy; Step S3 specifically includes the following steps: S31. For historical operating data of wind power output, photovoltaic power output and load demand, data is collected using a preset sliding window before each time point, so as to extract a historical data segment at each time point. S32. Based on the historical data segment at each time point, calculate the correlation parameter at each time point to obtain the local observed value of the correlation parameter; S33. Based on the correlation parameter values ​​at each historical moment, a correlation parameter estimation model is learned, wherein the correlation parameter value at each moment in the correlation parameter estimation model is a weighted sum of the correlation parameter value at the previous moment and the local observation value of the correlation parameter at the current moment. S34. Based on historical correlation parameter values, a correlation parameter estimation model is used to predict the correlation parameter values ​​at future times. S35. Determine the source-load comprehensive correlation index for future time based on the correlation parameter values ​​at future time. Step S4 specifically includes the following steps: S41. Construct an opportunity constraint, wherein the opportunity constraint is the difference between the probability of random fluctuation of source load being greater than or equal to 1 and the probability of default. S42. Based on the comprehensive correlation index of source and load, the probability of default is dynamically determined. The larger the absolute value of the comprehensive correlation index of source and load, the smaller the probability of default. S43. Based on the dynamic default probability, the opportunity constraint is transformed into a deterministic constraint that changes with the comprehensive correlation index of source and load, serving as an adaptive opportunity constraint. The adaptive opportunity constraint is as follows: , , Where g(x,t) represents the adaptive chance constraint, x represents the decision variable, and t represents time t. , Indicates the standard deviation and variance of the joint uncertainty variables; The inverse cumulative distribution function of the standard normal distribution; Indicates the quantile corresponding to the confidence level. Let represent the probability of default at time t. This represents the uncertainty variable caused by random fluctuations in the source load. This indicates the expectation, which is calculated using the source-load joint probability distribution model.

2. The power system dispatching method based on joint source-load probabilistic modeling and adaptive chance constraints according to claim 1, characterized in that, Step S1 specifically includes the following steps: S11. Obtain historical operating data reflecting the operating status of the power system. Collect wind power output, photovoltaic power output and load demand in the historical operating data according to time. After abnormal data processing and time scale unification, organize them into a historical source-load dataset reflecting the historical source-load combination operating status. S12. Based on the historical source-load dataset, calculate the probability density functions of wind power output, photovoltaic power output and load demand respectively, and establish marginal probability distribution models of wind power output, photovoltaic power output and load demand based on the probability density functions respectively.

3. The power system dispatching method based on source-load joint probabilistic modeling and adaptive chance constraints according to claim 1, characterized in that, Step S2 specifically includes the following steps: S21. Based on the edge probability distribution model of wind power output, photovoltaic power output and load demand, a source-load joint probability distribution model is constructed using the Copula function; S22. Establish a log-likelihood function based on the source-load joint probability distribution model, and solve for the correlation parameter values ​​with the objective of maximizing the log-likelihood function.

4. The power system dispatching method based on joint source-load probabilistic modeling and adaptive chance constraints according to claim 1, characterized in that, The optimized scheduling model described in step S5 is as follows: , s.t. , Where T represents the set of scheduling times; This represents the output of the conventional generating unit at time t; Indicates the amount of wind and solar power curtailed; This represents the power generation cost function of a conventional generating unit; This represents the cost function for penalties related to wind and solar power curtailment.

5. The power system dispatching method based on joint source-load probabilistic modeling and adaptive chance constraints according to claim 1, characterized in that, Step S6 specifically includes the following steps: S61. Unify the objective function and constraints of the optimization scheduling model into a standard optimization form, call the optimization solver to perform calculations, and obtain the optimal solution for the decision variables; S62. Transform the optimal solution of the decision variables into an executable scheduling strategy for the power system.

6. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, it implements the method of any one of claims 1-5.

7. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that: The computer program / instructions, when executed by a processor, implement the method of any one of claims 1-5.

Images