High-proportion new energy access-oriented main and distribution network power balance capability evaluation method and system

Abstract

The invention discloses a high-proportion new energy access-oriented main and distribution network power balance capability evaluation method and system, and the method comprises the steps: accurately describing an energy two-way transmission relation between a main network and a power distribution network at different time scales and different levels through constructing a multi-time scale balance capability analysis architecture and a main and distribution multi-level coupling operation model; and a typical operation scene set capable of reflecting a macroscopic statistical rule and local aggregation characteristics is generated, and uncertainty brought by high-proportion new energy access is quantified. Based on an affine form power feasible region reported by a power distribution network side, a robust feasibility judgment method based on affine arithmetic is adopted, whether main network key node voltage meets safety constraints under all uncertain fluctuations or not is verified through interval operation, and finally a robust feasible region of system balance capacity is obtained. And rapid and accurate evaluation of the cooperative balance capability of the main and distribution networks in a high-proportion new energy environment is realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power system safety analysis, in particular to a main and distribution network power balance capability evaluation method and system for high-proportion new energy access. BACKGROUND

[0002] With large-scale access of high-proportion new energy such as wind power and photovoltaic power in a power system, the strong randomness and volatility of both the source and the load are significantly intensified, which poses a severe challenge to the power balance capability of the coordinated operation of the main network and the distribution network. The traditional power grid safety analysis method based on a deterministic model cannot effectively cope with such uncertainty, and although the existing research proposes to use demand side response and active distribution network technology to improve new energy consumption capacity, it is also subject to the uncertainty of user behavior, the limitation of prediction accuracy and high coordination cost.

[0003] The existing evaluation method often separately processes or simplifies the power interaction and constraints between different levels of power grids, resulting in a large deviation between the evaluation results and the actual operation conditions. In addition, in the face of the complexity and real-time requirements brought by the access of a large number of distributed resources to the distribution network, the traditional evaluation method that relies on fine models for long-time simulation cannot meet the urgent needs of speed and adaptability of dispatching decisions.

[0004] Therefore, it is urgent to develop a method and system that can quickly and accurately evaluate the overall power balance capability of the main and distribution network under high-proportion new energy access, and provide efficient and reliable quantitative support for the safe and stable operation and optimized dispatching of the power grid. SUMMARY

[0005] The application aims to solve the problem that the power balance evaluation in the prior art relies on the Newton-Raphson method and other nonlinear power flow calculation iterations, which are time-consuming and cannot meet the demand for rapid evaluation of the safety margin of the main and distribution network under the high-proportion new energy scenario. The application discloses a main and distribution network power balance capability evaluation method and system for high-proportion new energy access.

[0006] Technical scheme

[0007] The application proposes a main and distribution network power balance capability evaluation method for high-proportion new energy access, which comprises the following steps:

[0008] S1: Construct a multi-time scale balance capability analysis architecture, wherein the time scale comprises a day-ahead stage and an intra-day rolling stage, and the balance capability architecture comprises: the distribution network side reports the power adjustable range of the distribution network side to the main network side, and the main network side analyzes and issues instructions according to the information reported by the distribution network side and safety constraints;

[0009] S2: constructing a main grid operation model and a distribution grid operation model according to the balance capability analysis framework, and predicting the operation of the power system, wherein the main grid operation model is connected to the distribution grid operation model, and a coupling constraint is applied to a boundary node of the connection;

[0010] S3: collecting historical operation data of variables of the main grid and the distribution grid, wherein the variables include new energy output and load, and extracting time sequence features of each variable; performing probability estimation on the new energy output and the load by using kernel density estimation, introducing a Copula function to represent the probability estimation result, and generating a typical scenario set;

[0011] S4: based on the typical scenario set, combining power flow equations, voltage safety constraints, line capacity limits and distributed resource operation constraints of the distribution grid, and calculating a power adjustable range of the distribution grid;

[0012] S5: using an affine function to describe the power adjustable range of the distribution grid, and reporting to the main grid side according to the balance capability analysis framework, and performing safety checking by calculating a voltage sensitivity coefficient to realize power balance capability evaluation of the main grid and the distribution grid.

[0013] Further, the day-ahead stage and the intra-day rolling stage include:

[0014] In the day-ahead stage, each distribution grid side performs balance capability calculation according to operation characteristics with an hour resolution, and reports analysis results and a power adjustable range of the distribution grid side to the main grid side;

[0015] In the intra-day rolling stage, the distribution grid side calculates a power adjustable range of the distribution grid based on ultra-short-term prediction data with a 15-minute cycle, and reports a robust feasible region obtained based on a robust optimization method to the main grid side.

[0016] Further, the power adjustable range of the distribution grid side is represented as:

[0017]

[0018] wherein is a net power of the connection point, is a lower power limit, is an upper power limit, and the upper and lower power limits are determined by internal resource constraints of the distribution grid;

[0019] The main grid side issues a decision according to a system power balance condition, and the condition is represented as:

[0020]

[0021] wherein is a main grid generator output, is a main grid load, For network loss.

[0022] Further, the constraint conditions of the main grid operation model include: power balance constraint, generator active power / reactive power constraint, node voltage constraint, branch power constraint, energy storage device operation constraint, thermal power unit output constraint, and the constraint conditions of the distribution grid operation model include: power flow distribution constraint, voltage quality constraint and device safety constraint.

[0023] Further, the coupling constraint is expressed as:

[0024]

[0025] Wherein represents the set of all distribution grids, represents any distribution grid k in the set; U represents voltage, P represents active power, Q represents reactive power, subscript B,k represents the node connected between the main grid and the distribution grid k, and subscript k,r represents the root node in the distribution grid k.

[0026] Further, the introduction of the Copula function represents the probability estimation result, including:

[0027] Using the fitted Copula model, a large number of correlated uniform random vectors are generated by Monte Carlo sampling , a large number of samples are randomly extracted from the constructed cumulative distribution function, and each sample represents a source load power combination;

[0028] Through the inverse transformation of the marginal distribution The source load power combination is restored to the original power value scenario to form an initial scenario set , wherein each scenario contains the source load power values of all nodes of the main grid and the distribution grid;

[0029] The net fluctuation intensity index is introduced to quantify the comprehensive risk of the scenario, and the initial scenario meeting is recorded as a typical scenario; the formula of the net fluctuation intensity index is:

[0030]

[0031] Wherein is the net fluctuation intensity of the scenario s, , is a weight coefficient, reflecting the influence degree of the fluctuation on the system balance on the generation side and the load side, is the system net power fluctuation value under the scenario s, is the standard deviation of the load fluctuation of the scenario s; the threshold value is set, wherein , For all scenarios The mean and standard deviation, This is an adjustable robustness coefficient.

[0032] Furthermore, the calculated adjustable power range of the distribution network is as follows:

[0033] Under each scenario s, the adjustable upper limit of power reported by the distribution network to the main grid side and lower limit The calculation formula is expressed as:

[0034] (1) Calculate the minimum interaction power:

[0035]

[0036] in This refers to the net power at the connection point between the distribution network and the main grid. Let be the minimum interaction power in scenario s. For the current state variable, To control variables, and These are the feasible regions for the state variables and the control variables, respectively.

[0037] (2) Calculate the maximum interaction power:

[0038]

[0039] in This represents the maximum interaction power in scenario s.

[0040] Furthermore, S5 specifically includes:

[0041] Step 1: The distribution network side reports the robust feasible region. Each distribution network calculates the power adjustable range under each typical scenario s based on the balance capacity analysis model described in S4, and reports it to the main grid side in affine form.

[0042] Step 2: Perform robust security verification on the main network side. After receiving the affine power expression reported by each distribution network, the main network side calculates the voltage sensitivity coefficient for each scenario s and each key node r, and verifies whether the voltage upper and lower limit constraints are met.

[0043] Step 3: Perform a global feasibility assessment. If all key nodes meet the voltage upper and lower limit constraints and safety constraints in all typical scenarios, then the main network determines that the system has the ability to balance within the current robust feasible domain.

[0044] Furthermore, the security constraints include:

[0045] For each typical scenario s, the linearized relationship between the main grid node voltage and power injection is as follows:

[0046]

[0047] in Let be the voltage value of node r at time t in scenario s. Let be the voltage value of node r at time t under the baseline state. Let be the partial derivative of the voltage at node r with respect to the power injected into the distribution network at node k. The planned power injection into the distribution network k at time t under the baseline condition;

[0048] Voltage safety constraints require that, for all typical scenarios s, the voltage be maintained within a safe range, expressed as:

[0049]

[0050] in Let be the lower limit of the voltage safety for node r at time t. To determine the safe voltage limit for node r at time t, substituting the voltage expression into the safety constraint yields:

[0051] .

[0052] The present invention also proposes a power balance capability assessment system for main and distribution networks with high proportion of new energy access, including a memory, a processor and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the steps of any of the aforementioned methods.

[0053] Beneficial effects:

[0054] This invention addresses the challenges of computational efficiency and accuracy in power balance assessment under high-proportion renewable energy access. By constructing a multi-timescale balance capability analysis architecture integrating main and distribution systems, it achieves collaborative optimization at the day-ahead hourly and intraday 15-minute levels, fully considering the time response differences of various types of controllable resources and the uncertainty of renewable energy output.

[0055] This invention proposes a method for generating typical scenarios of source load uncertainty based on kernel density estimation and Copula function. It generates a set of typical operating scenarios that can reflect macroscopic statistical laws and local clustering characteristics. Key scenarios are screened by the net fluctuation intensity index, and a robust evaluation framework is constructed, which significantly improves the ability to handle source load uncertainty and the reliability of evaluation results.

[0056] This invention establishes a multi-level coupled operation model between the main grid and the distribution network to accurately describe the bidirectional energy transmission relationship between the main grid and the distribution network. It also adopts a linearized voltage-power mapping model to transform the complex process of traditional nonlinear power flow calculation into efficient matrix operations, breaking through the computational bottleneck of the original method and achieving millisecond-level rapid evaluation capability.

[0057] This invention utilizes affine functions to characterize power interaction uncertainty and constructs an integrated analysis model of main grid balancing capability based on a robust feasible region. This model ensures that the voltage of key nodes in the main grid always meets safety constraints under boundary power interaction uncertainty fluctuations. Based on the affine power feasible region reported by the distribution network side, a robust feasibility judgment method based on affine arithmetic is adopted. Interval calculations are used to verify whether the voltage of key nodes in the main grid meets safety constraints under all uncertain fluctuations, ultimately obtaining the system's balancing capability assessment result. Attached Figure Description

[0058] Figure 1 This is a flowchart of the method of the present invention;

[0059] Figure 2 This is a structural diagram of the system of the present invention. Detailed Implementation

[0060] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0061] Example 1

[0062] The present invention provides a method for assessing the power balance capability of main and distribution networks for high-proportion renewable energy access, as follows: Figure 1 As shown, it includes the following steps:

[0063] S1: Constructing a multi-timescale balancing capability analysis framework integrating primary and secondary systems.

[0064] To ensure the safe and stable operation of the power system under a high proportion of renewable energy access, this invention first constructs a multi-timescale balance capability analysis architecture integrating main and distribution systems.

[0065] The architecture fully considers the differences in time response of various types of controllable resources such as thermal power units, energy storage systems, and flexible loads in the main grid and distribution network, as well as the uncertainty and rapid fluctuation of the output of new energy sources such as wind power and photovoltaics. The balance analysis is divided into two scales: day-ahead hour level and intraday 15-minute level.

[0066] In the current phase, each distribution network performs preliminary balancing capacity calculations on an hourly basis based on its own operating characteristics, and reports the analysis results and power adjustment range to the main grid. The main grid then integrates the information reported by all distribution networks, the adjustment capacity of its own large generating units, and safety constraints such as transmission section limits, to conduct a global optimization analysis, and issues power balance demand decisions for each time section for the next day.

[0067] During the intraday rolling phase, the distribution network side updates its boundary adjustable power range and corresponding robust feasible region in a rolling manner with a 15-minute cycle based on ultra-short-term forecast data.

[0068] Active power variables are injected into the boundaries of each distribution network at the main and distribution common coupling point. The adjustable power range for time period t is expressed as follows:

[0069]

[0070] in and These represent the lower and upper limits of the boundary power that distribution network k can absorb or send to the main grid at time t, respectively.

[0071] Meanwhile, when issuing decisions from the main network side, system power balance constraints must be met:

[0072]

[0073] in Power output from the main grid generator. Main network load, This architecture enables bidirectional interaction and collaborative balancing between the primary and distribution networks, moving from slow to fast speeds and from planned to real-time operations. This provides a framework for accurately assessing the overall system balance capability.

[0074] S2: Construct a multi-level coupled operation model of primary and secondary components.

[0075] To accurately predict the operation of the power system, this invention constructs a multi-level coupled operation model based on a multi-timescale balancing capacity analysis architecture integrating the main grid and distribution network. This model effectively reflects the interaction between the main grid and the distribution network. The model fully considers the topological differences between the main grid and the distribution network, the distribution and characteristics of adjustable resources within their respective jurisdictions, and the power transmission capacity of key transmission sections, among other key operational characteristics. Based on this, by accurately describing the bidirectional energy transmission relationship between the main grid and the distribution network at different time scales and levels, a unified and coordinated main-distribution collaborative optimization framework is formed.

[0076] Mainnet operation model:

[0077] As the backbone network of the power system, the main grid model aims to ensure the security and economy of large-scale power transmission, and mainly includes the following constraints:

[0078] 1) Power balance constraints

[0079]

[0080]

[0081] in and They are nodes The active and reactive power generated by the generator. and They are nodes Active and reactive loads, and They are nodes and nodes The conductance and susceptance of the branches between them. This represents the set of nodes in the main network. This constraint ensures that the sum of the power injected and outflowed by any node is zero, which is the foundation for the stable operation of the system.

[0082] 2) Constraints on generator active and reactive power output

[0083]

[0084]

[0085] in and Representing nodes respectively The upper and lower limits of the active power output of the generator. and Representing nodes respectively The upper and lower limits of the reactive power output of the generator.

[0086] 3) Node voltage constraints

[0087]

[0088] in and They are nodes The upper and lower voltage limits at the specified location.

[0089] 4) Branch power constraints

[0090]

[0091] in and They are nodes , Apparent power and upper limit of apparent power of the branch between them. This is the set of all branches in the main network.

[0092] 5) Operational constraints of energy storage devices

[0093] Energy storage devices must meet operational constraints such as charging and discharging power constraints and state of charge constraints.

[0094]

[0095] in and They represent the main network nodes at time t, respectively. The charging and discharging power of the energy storage device. , Mainnet nodes The capacity and power of the energy storage device , Mainnet nodes for time period t The charging and discharging states of energy storage devices are 0-1 variables. , This indicates that the energy storage device is in a charging state. , This indicates that the energy storage device is in a discharging state. For the mainnet node in time period t In a state of energy storage and charge, , These refer to the energy storage charging and discharging efficiency, respectively. , These represent the maximum and minimum states of charge of the main grid energy storage, respectively.

[0096] 6) Output constraints of thermal power units

[0097]

[0098] in Mainnet nodes for time period t The upper and lower limits of the output of thermal power units.

[0099] 7) Other constraints

[0100]

[0101]

[0102] in , and Representing nodes respectively , The tap positions of the transformers on the branch lines, and their upper and lower limits. , and Representing nodes respectively The value of the capacitor connected at that point, and its upper and lower limits. It is a set of positive integers.

[0103] Distribution network operation model:

[0104] The distribution network model focuses on power flow distribution, voltage quality, and equipment safety under a radial network structure, and its operating constraints are as follows:

[0105] st

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112] in , They represent the nodes at time t, respectively. , The square of the voltage amplitude, , and Represents a node , The square of the current in the branch, the square of the resistance in that branch, and the square of the reactance. and This indicates the active and reactive power on that branch. This represents the set of all branches in the distribution network. This represents the set of all nodes in a power distribution network. This represents the set of nodes other than the root node (the substation node connected to the main network). This means that at any point in time t, , and They represent the nodes at time t, respectively. The active power, reactive power, and apparent power generated by renewable energy sources (such as photovoltaics). , They represent the nodes at time t, respectively. Active and reactive loads, Represents the node at time t The reactive power provided by the capacitor bank. , and They represent the nodes at time t, respectively. Upper and lower limits of voltage amplitude and nodes , The upper limit of the current in the branch between them.

[0113] Principal-partner coupling model:

[0114] The core of multi-level coupled operation of the main network and distribution networks lies in the consistency of boundary information. To achieve effective coordination between the main network sub-problems and the various distribution network sub-problems, the following coupling constraints need to be applied at their connecting boundary nodes:

[0115]

[0116] in , and These represent the node voltage, active power, and reactive power connected between the main grid and the distribution network k, respectively. , and Let these represent the voltage, active power, and reactive power injection at the root node in distribution network k, respectively. Represents the set of all distribution networks. This indicates that for each distribution network k.

[0117] S3: Propose a method for generating typical scenarios of source-charge uncertainty based on kernel density estimation and Copula function.

[0118] Probabilistic modeling is performed on the power output and load fluctuations of renewable energy sources on both the main grid and distribution grid sides. Due to the strong randomness and intermittency of renewable energy output and load, their probability distributions often do not conform to a standard normal distribution, making it difficult for traditional distribution models based on parameters (such as mean and variance) to accurately describe their actual statistical patterns. Therefore, this invention employs kernel density estimation (KDE) for nonparametric probability density estimation. KDE is a data-driven smoothing estimation method that approximates the true probability density function by placing a kernel function (such as a Gaussian kernel) at each historical data point and performing a weighted summation, without requiring prior assumptions about the distribution form. For a one-dimensional random variable X (such as the renewable energy output at a certain node), the kernel density estimation formula is:

[0119]

[0120] in For the point The probability density estimate at that location, For the number of historical samples, For the first Each sample value The bandwidth parameter controls the smoothness, and its value is determined through cross-validation to balance the bias and variance of the estimation. For the kernel function, this invention uses the Gaussian kernel function, that is:

[0121] .

[0122] To characterize the complex spatiotemporal correlations between multivariable source and load power in the main grid and distribution network, the Copula function is introduced. The Copula function is a tool that connects marginal distributions and joint distributions, capable of separating the marginal distributions of individual variables from the dependency structure between variables, flexibly describing nonlinear and asymmetric correlations. Let d-dimensional random variables... These represent the renewable energy output or load at different nodes in the main grid and distribution network, respectively, and their edge distribution functions are: Then the joint distribution function This can be represented by the Copula function C as follows:

[0123]

[0124] in For Copula functions, The parameter vector of a Copula can be fitted from historical data using maximum likelihood estimation. Maximum likelihood estimation is a commonly used parameter estimation method; its core idea is to find a set of parameter values ​​that maximizes the probability of observing existing historical data. This invention selects an appropriate Copula type based on data characteristics; for example, Gaussian Copula is suitable for symmetric correlation structures, while t-Copula is suitable for scenarios with strong tail correlations.

[0125] Based on the above probability model, the steps for generating a typical operating scenario are as follows:

[0126] Step 1: Data Collection and Feature Extraction. Collect historical operational data from the main grid and distribution network, including renewable energy output (wind power, photovoltaic) and load data, covering different time segments (e.g., 96 points / day) and spatial nodes. Extract the temporal features of each variable and construct a feature vector containing spatiotemporal correlation information, such as the output values ​​of multiple nodes at the same time and time lag correlation.

[0127] Step 2: Marginal Distribution Estimation. For each variable, such as grid wind farm output, distribution grid photovoltaic output, and load, kernel density estimation is applied to obtain its marginal probability density function. and cumulative distribution function ,in .

[0128] Step 3: Copula Modeling and Parameter Fitting. The marginal distribution values... Mapping to a uniformly distributed space yields Based on Given the sample data, select an appropriate Copula function form and estimate the parameters. This is used to capture the dependency structure between variables.

[0129] Step 4: Joint Scene Generation. Using the fitted Copula model, a large number of correlated uniform random vectors are generated through Monte Carlo sampling. Monte Carlo sampling is a numerical computation method based on random numbers. Its core idea is to simulate the possible states of a complex system through repeated random sampling. In this method, a large number of samples are randomly drawn from a pre-constructed joint probability distribution, with each sample representing a possible source-load power combination. Then, an inverse marginal distribution transformation is applied. Restore the scene to its original power value to form an initial scene set. Each of the scenes It includes the source load power values ​​of all nodes in the main grid and distribution network.

[0130] Step 5: Scenario Reduction and Key Scenario Selection. To improve computational efficiency, a net volatility intensity index is introduced to quantify the overall risk of each scenario:

[0131]

[0132] in Let be the net fluctuation intensity of scenario s. , These are weighting coefficients that reflect the degree to which fluctuations on the generation and load sides affect the system balance. The net power fluctuation value of the system under scenario s. Let the standard deviation of the load fluctuation in scenario s be defined. Set a threshold. ,in , For all scenarios The mean and standard deviation, The robustness coefficient is adjustable. Screening is performed to ensure compliance. The key scenarios are used for subsequent evaluation, which significantly reduces the computational scale.

[0133] The typical scenario set generated through the above process can not only accurately reflect the probabilistic characteristics of source-load uncertainty, but also effectively capture its spatiotemporal correlation, providing a targeted and highly representative input scenario for the assessment of the power balance capability of the main and distribution networks.

[0134] S4: Establish a distribution network balancing capacity analysis model

[0135] Based on the generated typical scenarios of uncertain source and load on the distribution network side, and considering the probability distribution of its new energy sources, a distribution network balancing capacity analysis model is constructed. This model aims to quantify the balancing capacity of the distribution network under uncertain conditions and report key boundary information to the main grid side for integrated analysis of the main and distribution networks.

[0136] Considering the power flow equations, voltage safety constraints, line capacity limitations, and distributed resource operation constraints of the distribution network, and linearizing the nonlinear constraints at the reference operating point, the voltage constraint linearization is as follows:

[0137]

[0138] The power balance constraint is linearized as follows:

[0139]

[0140] in , This is a node-branch association matrix. This represents the net change in power.

[0141] All constraints are represented in matrix form:

[0142]

[0143] in This is the state variable coefficient matrix, which includes linearization coefficients such as voltage sensitivity and power balance. The coefficient matrix of the control variables characterizes the ability of controllable resources to regulate the system state. The historical state correlation matrix describes the temporal coupling relationships. The coefficient matrix of the scenario variables quantifies the impact of uncertainty. To constrain the boundary vectors, various safety operation limits are included. Based on the above constraint system, a balancing capacity assessment and optimization problem is constructed to determine the feasible range of power interaction with the main grid.

[0144] (1) Calculate the minimum interaction power:

[0145]

[0146] in This refers to the net power at the connection point between the distribution network and the main grid. Let be the minimum interaction power in scenario s. For the current state variable, To control variables, and These are the feasible regions of state variables and control variables, respectively, and are composed of physical constraints such as power flow equations, voltage safety constraints, line capacity limits, and distributed generation regulation capabilities of the distribution network.

[0147] (2) Calculate the maximum interaction power:

[0148]

[0149] in This represents the maximum interaction power in scenario s.

[0150] During the day-ahead rolling phase, the photovoltaic system adjusts its power output based on short-term forecasts of load and renewable energy fluctuations. Through short-term forecasting, the photovoltaic system can respond in real-time to load and renewable energy fluctuations, thus providing timely power regulation support to the distribution network. During the day-ahead phase, on-load tap-changing transformers optimize voltage and power transmission by adjusting transformer taps.

[0151] Ultimately, the adjustable power range reported by the distribution network to the main grid under each scenario s is obtained as follows: Upper limit and lower limit This range refers to the robust feasible region within which the distribution network can operate safely and interact with the main grid under current renewable energy output and load conditions. The main grid will then use this range, following the multi-timescale balancing capacity analysis architecture constructed in S1, to conduct a global balancing capacity assessment and decision-making process.

[0152] S5: Integrated Analysis Model and Solution for Main-Support Balance Capability Based on Robust Feasible Domain

[0153] For typical uncertain scenarios generated on the main grid side, considering the information interaction and energy transmission on the distribution network side, the upper and lower limits of the balancing capacity provided by the distribution network side to the main grid side are characterized by affine functions. Then, by integrating affine mathematics and the linearized voltage-power mapping model, an integrated analysis model of the main grid balancing capacity based on the robust feasible region is proposed.

[0154] The integrated analysis model of main and auxiliary power supply balancing capability is the core method for achieving the ultimate goal of this invention. The final evaluation result is determined based on whether power balance is satisfied under all disturbance conditions; the evaluation result indicates either balancing capability or non-balancing capability. The judgment steps of the integrated analysis model of main and auxiliary power supply balancing capability are as follows:

[0155] Step 1: Distribution network side reports robust feasible region. Based on the balance capacity analysis model described in S4, each distribution network calculates the power adjustable range under each typical scenario s and reports it to the main grid side in affine form.

[0156] The main power grid and multiple distribution networks exchange power through connection points. The uncertainty of power exchange at these connection points is mathematically characterized, and the connection points between the distribution networks and the main grid are referred to as boundary nodes. The feasible range of interactive power corresponding to each typical scenario s reported by the distribution network is mapped to boundary nodes. The affine form of the net injected power. For scene s, the boundary nodes... Active injection It can be represented as:

[0157]

[0158] in This is the center value of power injection, i.e., the expected power exchange point in this scenario. This is the amplitude coefficient for power fluctuations. It is an independent noise element with a value range of [−1,1], used to characterize the power fluctuation caused by the uncertainty of new energy sources and load in this specific scenario.

[0159] Assuming the active distribution network operates in a certain state (p0, q0, V0, θ0), when a disturbance occurs in the injected power at a node, its operating state changes accordingly, and the changed operating state becomes (p, q, V, θ). Then, the following holds:

[0160]

[0161] Where p, q, V, θ represent the active power injection vector, reactive power injection vector, voltage magnitude vector, and voltage phase angle vector at the node, respectively, and Δp, Δq, ΔV, Δθ represent the changes in the active power injection vector, reactive power injection vector, voltage magnitude vector, and voltage phase angle vector at the node, respectively.

[0162] Step 2: Perform robust security verification on the main network side. After receiving the affine power expressions reported by each distribution network, the main network side calculates the voltage sensitivity coefficient for each scenario s and each critical node r to evaluate the degree of response of voltage changes to power fluctuations and verify whether the upper and lower voltage limits are met.

[0163] The corrected equations based on the Newton-Raphson power flow calculations are shown below:

[0164]

[0165] By performing an equivalent transformation on the above equation, we can obtain:

[0166]

[0167]

[0168] Where J represents the Jacobian matrix, S p θ S represents the "voltage phase angle - nodal active power injection" sensitivity matrix. q θ S represents the "voltage phase angle - nodal reactive power injection" sensitivity matrix. p V S represents the "voltage amplitude - nodal active power injection" sensitivity matrix. q V This represents the sensitivity matrix of "voltage amplitude - nodal reactive power injection".

[0169] For node r in the main network that needs to be monitored for safe operation, its voltage amplitude Change relative to the reference operating point Approximately expressed as:

[0170]

[0171] in and These are the sensitivity coefficients of voltage to active and reactive power, respectively. Since the sensitivity relationship is linear, the voltage deviation at node r... The affine form that inherits power injection is represented as:

[0172]

[0173]

[0174] Therefore, the voltage affine expression for node r is obtained as follows:

[0175]

[0176] in Let be the initial voltage of node r.

[0177] Step 3: Perform a global feasibility assessment. If all critical nodes meet the above constraints in all typical scenarios, the main network determines that the system has balancing capabilities within the current robust feasible domain, and the power adjustment range reported by the distribution network can be accepted by the main network.

[0178] The robust feasible region requirement for the mainnet is that for each reported typical scenario s, when the boundary interaction power fluctuates arbitrarily within the interval described by its affine form, i.e., the noise element... When ∈[−1,1], the voltage amplitude of all critical nodes in the main network must always meet the upper and lower limits of safe operation constraints:

[0179]

[0180] in , These represent the lower and upper limits of the node voltage amplitude, respectively.

[0181] To ensure this robustness constraint holds, the voltage deviation under worst-case conditions must be considered. According to the interval operation rules of affine arithmetic, the above constraints can be transformed into a set of deterministic inequality constraints without uncertain variables. For the upper voltage limit constraint:

[0182]

[0183] Since the expression is linear, the maximum value is at the endpoints. Therefore, it is equivalent to:

[0184]

[0185] Similarly, for the lower voltage limit constraint:

[0186]

[0187] The integrated analysis model of main grid balancing capacity can be expressed as a main grid optimization problem considering multiple scenarios. The objective function can be to maximize the main grid's capacity to accommodate new energy sources, minimize total operating costs, or maximize system balance margin, etc.

[0188] For each typical scenario *s*, the main grid node voltage must meet safety constraints to ensure system operational feasibility under that scenario. The linearized relationship between voltage and power injection is as follows:

[0189]

[0190] in Let be the voltage value of node r at time t in scenario s. Let be the voltage value of node r at time t under the baseline state. Let be the partial derivative of the voltage at node r with respect to the power injected into the distribution network at node k. Let be the planned power injection into the distribution network k at time t under baseline conditions. Voltage safety constraints require that the voltage remain within safe limits for all typical scenarios s:

[0191]

[0192] in Let be the lower limit of the voltage safety for node r at time t. Let be the upper limit of the safe voltage of node r at time t.

[0193] Substituting the voltage expression into the safety constraints, we get:

[0194]

[0195] This embodiment tests the load during peak hours (14:00-16:00, wind power fluctuation ±30%) of a high-proportion renewable energy demonstration zone in a provincial power grid on a certain day. The results comparing the traditional optimal power flow method with the method of this invention are shown in Table 1.

[0196] Table 1. Comparison of the present invention with the traditional main-distribution low-power balance capability evaluation method

[0197] Evaluation method Prediction confidence interval Single scenario solve time (seconds) Kilo scenario solve time (minutes) Conventional method 75%~85% 3.5 58 Inventive method 95% or more 0.18 0.75

[0198] As shown in Table 1, the optimized prediction confidence interval, scenario solution time, and other indicators of the method of the present invention are effectively improved, thereby increasing the operating efficiency of the system.

[0199] Example 2

[0200] The second embodiment of the present invention provides a power balance capability assessment system for main and distribution networks, the structure of which is as follows: Figure 2 As shown, it includes: an initialization module, a main-distribution integrated multi-timescale balance capacity analysis architecture construction module, a main-distribution multi-level coupled operation model construction module, a source-load uncertainty typical scenario generation module, a distribution network balance capacity analysis module, and a main-distribution balance capacity integrated analysis module based on robust feasible domain.

[0201] The initialization module is used to collect data on the structure and equipment parameters of the main power distribution network, providing basic support for the rapid assessment of subsequent power balance capabilities.

[0202] The main and distribution integrated multi-timescale balancing capacity analysis architecture construction module is used to build a main and distribution collaborative analysis framework with dual timescales of day-ahead hourly level and intraday 15-minute level. It fully considers the time response differences of various types of controllable resources such as thermal power units, energy storage systems, and flexible loads, and realizes two-way interaction and collaborative balancing from slow to fast and from planned to real-time, providing a time-series basis for subsequent accurate assessment.

[0203] The main and distribution multi-level coupled operation model construction module is used to establish a main grid model that includes main grid power balance constraints, generator output constraints, node voltage constraints, branch power constraints, and energy storage device operation constraints, as well as a distribution network model that includes radial network power flow equations, voltage quality constraints, and equipment safety constraints. It also achieves effective coordination between the main and distribution networks through boundary coupling constraints, forming a unified main and distribution collaborative optimization framework.

[0204] The module for generating typical scenarios of source-load uncertainty is used to perform probabilistic modeling of the output of new energy sources and load fluctuations in the main grid and distribution network using kernel density estimation and Copula function. It generates a set of typical scenarios that reflect spatiotemporal correlations and selects key scenarios through net fluctuation intensity index, providing a structured quantitative method and probabilistic basis for uncertainty in the assessment of the power balance capability of the main grid and distribution network.

[0205] The distribution network balancing capacity analysis module is used to verify the operational feasibility based on typical scenarios. By solving an optimization problem with the goal of maximizing slack variables, it quantifies the balancing capacity of the distribution network under uncertain environments, calculates the feasible range of power interaction with the main grid, and reports key boundary information to the main grid side.

[0206] The integrated analysis module for main and distribution network balancing capacity based on robust feasible region is used to characterize the upper and lower limits of balancing capacity reported by the distribution network side using affine functions. It integrates affine mathematics and linearized voltage-power mapping model to construct a robust feasible region evaluation model for main and distribution network coordination, ensuring that key operating parameters of the main network meet safety constraints under uncertain fluctuations in boundary interactive power.

Claims

1. A method for assessing the power balance capability of a main and distribution network for high-proportion renewable energy access, characterized in that, include: S1: Construct a multi-timescale balancing capacity analysis architecture, wherein the timescale includes the day-ahead stage and the intraday rolling stage, and the balancing capacity architecture includes: the distribution network side reporting the adjustable power range of the distribution network side to the main grid side, and the main grid side analyzing and issuing instructions based on the information reported by the distribution network side and the safety constraints. S2: Based on the balance capacity analysis architecture, construct the main grid operation model and the distribution network operation model respectively to predict the operation of the power system. The main grid operation model is connected to the distribution network operation model, and the boundary nodes of the connection are subject to coupling constraints. S3: Collect historical operating data of variables in the main grid and distribution network, including the output and load of new energy sources, and extract the time-series characteristics of each variable; use kernel density estimation to perform probability estimation of the output and load of the new energy sources, introduce a Copula function to represent the probability estimation results, and generate a set of typical scenarios; S4: Based on the aforementioned typical scenario set, and combined with the power flow equations, voltage security constraints, line capacity limitations, and distributed resource operation constraints of the distribution network, the adjustable power range of the distribution network is calculated. S5: The adjustable power range of the distribution network is characterized by affine functions, and the data is reported to the main grid side based on the balance capacity analysis architecture. The main grid side performs safety verification by calculating the voltage sensitivity coefficient, thereby realizing the power balance capacity assessment of the main and distribution networks.

2. The power balance capability assessment method according to claim 1, characterized in that, The day-ahead phase and intraday rolling phase include: In the current phase, each distribution network side performs balancing capacity calculations based on its operating characteristics at an hourly resolution, and reports the analysis results and the adjustable power range of the distribution network side to the main grid side. During the intraday rolling phase, the distribution network side calculates the adjustable power range of the distribution network based on ultra-short-term forecast data with a period of 15 minutes, and reports the robust feasible region obtained based on the robust optimization method to the main grid side.

3. The power balance capability assessment method according to claim 2, characterized in that, The adjustable power range on the distribution network side is expressed as follows: in This is the net power of the connection point. This is the lower limit of power. This is the upper limit of power; the upper and lower limits of power are determined by the resource constraints within the distribution network. The main grid side makes decisions based on the condition of system power balance, which is expressed as: in Power output from the main grid generator. Main network load, This is due to network loss.

4. The power balance capability assessment method according to claim 3, characterized in that, The constraints of the main grid operation model include: power balance constraints, generator active / reactive power output constraints, node voltage constraints, branch power constraints, energy storage device operation constraints, and thermal power unit output constraints. The constraints of the distribution network operation model include: power flow distribution constraints, voltage quality constraints, and equipment safety constraints.

5. The power balance capability assessment method according to claim 4, characterized in that, The coupling constraint is expressed as follows: in Represents the set of all distribution networks. Let k represent any distribution network in the set; U represents voltage, P represents active power, Q represents reactive power, subscript B,k represents the node connecting the main network and distribution network k, and subscript k,r represents the root node in distribution network k.

6. The power balance capability assessment method according to claim 5, characterized in that, The introduction of the Copula function represents the probability estimation result, including: Using the fitted Copula model, a large number of correlated uniform random vectors are generated through Monte Carlo sampling. A large number of samples are randomly drawn from the constructed cumulative distribution function, and each sample represents a source-load power combination. Inverse transformation of edge distribution The source-load power combination is restored to the original power value scenario to form an initial scenario set. Each of the scenes Includes source load power values ​​for all nodes in the main grid and distribution network; Introducing the net volatility index to quantify the comprehensive risk of the scenario, and screening for scenarios that meet the criteria. The initial scenario is denoted as the typical scenario; the formula for the net volatility intensity index is: in Let be the net fluctuation intensity of scenario s. , These are weighting coefficients that reflect the degree to which fluctuations on the generation and load sides affect the system balance. The net power fluctuation value of the system under scenario s. Define the standard deviation of load fluctuation in scenario s; set a threshold. ,in , For all scenarios The mean and standard deviation, This is an adjustable robustness coefficient.

7. The power balance capability assessment method according to claim 6, characterized in that, The calculated adjustable power range of the distribution network is as follows: Under each scenario s, the adjustable upper limit of power reported by the distribution network to the main grid side and lower limit The calculation formula is expressed as: (1) Calculate the minimum interaction power: in This refers to the net power at the connection point between the distribution network and the main grid. Let be the minimum interaction power in scenario s. For the current state variable, To control variables, and These are the feasible regions for the state variables and the control variables, respectively. (2) Calculate the maximum interaction power: in This represents the maximum interaction power in scenario s.

8. The power balance capability assessment method according to claim 7, characterized in that, S5 specifically includes: Step 1: The distribution network side reports the robust feasible region. Each distribution network calculates the power adjustable range under each typical scenario s based on the balance capacity analysis model described in S4, and reports it to the main grid side in affine form. Step 2: Perform robust security verification on the main network side. After receiving the affine power expression reported by each distribution network, the main network side calculates the voltage sensitivity coefficient for each scenario s and each key node r, and verifies whether the voltage upper and lower limit constraints are met. Step 3: Perform a global feasibility assessment. If all key nodes meet the voltage upper and lower limit constraints and safety constraints in all typical scenarios, then the main network determines that the system has the ability to balance within the current robust feasible domain.

9. The power balance capability assessment method according to claim 8, characterized in that, The security constraints include: For each typical scenario s, the linearized relationship between the main grid node voltage and power injection is as follows: in Let be the voltage value of node r at time t in scenario s. Let be the voltage value of node r at time t under the baseline state. Let be the partial derivative of the voltage at node r with respect to the power injected into the distribution network at node k. The planned power injection into the distribution network k at time t under the baseline condition; Voltage safety constraints require that, for all typical scenarios s, the voltage be maintained within a safe range, expressed as: in Let be the lower limit of the voltage safety for node r at time t. To determine the safe voltage limit for node r at time t, substituting the voltage expression into the safety constraint yields: 。 10. A power balance capability assessment system for main and distribution networks with high proportion of renewable energy access, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of any of the methods of claims 1 to 9.

Images