CN-121983968-A - Wind-fire bundling system adjustable margin prediction method suitable for uncertain working conditions

CN121983968ACN 121983968 ACN121983968 ACN 121983968ACN-121983968-A

Abstract

The invention relates to the technical field of operation and optimization of an electric power system, in particular to a wind-fire bundling system adjustable margin prediction method suitable for uncertain working conditions, which comprises the following steps of firstly, establishing a power model; the method comprises the steps of parameterizing and expressing, establishing a robust optimization problem, determining candidate upper and lower bounds, solving a sub-problem, determining an interface power bound, outputting a prediction result, establishing a multi-period power model of an outgoing interface point for an 'wind-fire bundling outgoing' scene, uniformly integrating regulation constraints such as wind power fluctuation, thermal power climbing/minimum power output and the like into an interface adjustable margin depiction, and explicitly integrating mechanism conditions such as wind power, thermal power, energy storage, climbing constraint and the like through establishing a uniform aggregation model of wind power, thermal power energy storage, outgoing point interface power, so that the predicted interface power adjustable region is guaranteed to have decomposable and executable scheduling significance, and meanwhile, the method is lighter in weight and reduces support control complexity.

Inventors

LIU JIANZHE
LI CANBING
YOU MEI
LI FENG
WANG YANFENG
PENG SUI
ZHANG MINCHANG

Assignees

上海交通大学

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (8)

1. A wind and fire bundling system adjustable margin prediction method suitable for uncertain working conditions comprises the steps of establishing a power model, establishing parameterized expression, establishing a robust optimization problem, determining candidate upper and lower bounds, solving a sub-problem, determining an interface power boundary, outputting a prediction result and the like, and is characterized in that: In the first step, a wind-fire bundling and sending interface point aggregation power model is established; in the second step, an uncertain set is constructed, and parameterization expression is carried out on the adjustable margin; In the third step, a two-stage adaptive robust optimization problem is established; in the fourth step, an uncertain scene is initialized, and a main problem is solved to obtain upper and lower boundaries of candidates; in the fifth step, solving the sub-problem, verifying the robustness of the problem by using the candidate boundary and searching the most unfavorable scene; in the sixth step, the main-sub problem iterates until a robust and flexible interface power boundary is found; in the seventh step, the adjustable margin prediction result is output and used for upper layer control.
2. The method for predicting adjustable margin of fire and wind bundling system according to claim 1, wherein in said step one, send out point is marked as node The rest node set except the interface node in the system is marked as N\ { k }, the power change of multiple time periods is considered, and the time period is marked as The aggregate active power of the interface node is defined as: (1) Wherein the method comprises the steps of In order to be the interface voltage, For the voltage at the other nodes of the system, For the mutual resistance between the interface node k and the two nodes of the distributed node j, the power of the interface node can be written as follows by the product of first-order Taylor expansion and increment in a certain period: (2) Wherein P k 0 、I k 0 、V k 0 represents the power, current and voltage of the interface node at the linearization base point, respectively, deltaV k and DeltaV j represent the voltage increment of node k and node j, respectively, g kj represents the equivalent conductance between node k and node j, and g kj =1/R kj ; when all equipment of the current system is connected with equivalent resistance from a node to an outgoing point, the wind power cluster W, the thermal power unit output U, the energy storage injection E, the station power or auxiliary power, the cable and the converter loss are mapped to the node in sequence The net injection of (2) is: (3) Wherein the method comprises the steps of Represent the first Desk fan or the first The wind fields are converged to the equivalent active power of the direct current system, and the available power and the limited-power constraint are as follows: (4) (5) the available active power of the fan/wind field g, representing the period t, is affected by the uncertainty amount ζ g,t defined by step two, The power limit of the fan/wind field g in the period t is represented, the upper limit of the available power determined by the fact that the output of the fan is limited by the wind condition and the state of the unit in the same period is reflected, and the power limit is possibly carried out for the purpose of balancing and stably running the system in relation to the scheduling requirement of the system; for adjustable output of the thermal power unit in bundling and sending, setting the time period t thermal power as the power meeting the upper limit and the lower limit of output and climbing constraint: (6) (7) Wherein, the And Respectively representing the minimum and maximum active output of the thermal power unit u; and The method is characterized by respectively representing the limit of downward climbing and upward climbing of the thermal power unit u, wherein the limit of climbing reflects the regulation hysteresis characteristic of the thermal power unit and is used for forming complementary regulation with the fluctuation of wind power; For the injection power of the distributed energy storage equipment to the system, the upper and lower limit constraint of the power is satisfied: and device SOC constraints across time scales: (8) (9) Wherein, the And SOCe, t represents the charge state of the energy storage unit e in a period t; and Respectively representing charging power and discharging power; and Respectively representing charging efficiency and discharging efficiency; Adopting a sign convention of discharging to be positive and charging to be negative; For station electricity or auxiliary power, The system loss is introduced as a given constant, and modeling can be performed by using a fixed efficiency or a quadratic model in practice; Unifying the power models into the power of the interface through a formula (3), and restraining the interface power boundary and the interface voltage boundary, wherein the method comprises the following steps: (10) (11) the upper and lower limits of the formula (10) are unknown, namely the boundary of the adjustable margin is obtained by solving the following problems; And establishing a unified multi-period model for the wind fire bundling cluster at a power transmission node of the wind fire bundling cluster, wherein the unified multi-period model comprises available power and limited power constraint of a fan, output and climbing constraint of a thermal power unit, energy storage charge state and charge and discharge constraint and interface unified constraint, and a computable multi-period constraint set is formed.
3. The method for predicting adjustable margin of wind-fire bundling system for uncertain conditions according to claim 1, wherein in step two, the power boundary of the interface power is used as an unknown quantity for predicting the adjustable margin problem, and the interface power is obtained by subsequent solution to simplify the model The definition is as follows: (12) wherein xi 0 is an indefinite amount, For the lower limit of the controllable power of node k at time t, For the upper limit of the controllable power of node k at time t, For normalizing the uncertainty amount and realizing the normalization of the interface power, it is clear that the interface power is defined as the uncertainty amount, and the meaning is that any power realization track in the required adjustable margin can find the self-adaptive energy storage and fan limit power generation strategy aiming at the working condition, namely modeling the energy storage and fan limit power generation strategy as an uncertainty amount, and finding the boundary of the uncertainty amount when all constraints of the system are met; Meanwhile, in order to ensure the robustness of the model, the randomness of the fan output is considered, and the uncertainty is defined and normalized; (13) Wherein, the And The wind power unit g respectively represents a lower limit and an upper limit of the prediction of available active power in a period t, an uncertainty set can be defined as a box shape, an ellipsoid shape and a polyhedron shape according to actual requirements, the historical data of a direct current interface and a fan are analyzed to be generalized into the box-shaped uncertainty set which is randomly valued in the upper and lower limits, and cost limitation is considered, so that the wind power unit g needs to be reset according to problems.
4. The method for predicting adjustable margin of wind-fire bundling system for uncertain conditions according to claim 1, wherein in step three, an interface power adjustable margin boundary is maximized as an optimization target, and the method comprises the following steps of Representing two-stage adjustable decisions, uniformly writing constraint columns in the first step and the second step into compact matrix form equality and inequality constraint, and describing an interface adjustable margin prediction problem as the following adaptive robust optimization problem: (14) Wherein the method comprises the steps of And Respectively representing a lower boundary vector and an upper boundary vector of aggregate active power at an interface of all time periods, A and E are constraint coefficient matrixes, a and E are constant vectors, B and F are uncertainty mapping matrixes, C and G are interface power mapping matrixes, p c (ξ 0 ) is an interface power track vector given by a formula (12), an objective function represents the maximization of the adjustable margin width of each time period, the problem is a two-stage robust optimization problem, ζ is a random variable, and the data are collected The data, representing the fluctuation of the fan, and y (ζ) is the self-adaptive quantity realized after the system looks at ζ, and the formula (14) can not be directly solved due to the random variation of the uncertain quantity, so a main-sub problem compression solving method is designed.
5. The method for predicting adjustable margin of fire bundling system according to claim 1, wherein in step four, firstly, initializing some uncertain scenes, solving optimization problem under fixed scene, and collecting the uncertain scenes during the mth round of iteration as follows At this time, the main problem only needs to ensure that the decomposition is feasible under the scenes, namely: (15) main problem output candidate boundary , For the upper bound of the adjustable margin to be optimized, For the lower bound of the adjustable margin to be optimized, The vector is a decision variable vector in the s-th scene, A, B, C, E, F and G are coefficient matrixes, and a and e are constant vectors.
6. The method for predicting adjustable margin of fire and wind bundling system according to claim 1, wherein in step five, in order to check whether the candidate boundary is decomposable and feasible under all uncertain implementations, introducing maximum constraint violation z not less than 0, and fixing the candidate boundary in step four Step five is then used to find the scene that makes the constraint most vulnerable and characterized by the maximum violation z: (16) If z Ε, which is a tolerance, typically 0 or a very small positive number, the candidate boundaries are considered to meet the robustness feasibility; If z Epsilon, then the sub-problem gives the worst scenario For updating the master question, wherein, For the dc interface power lower bound A, B, C is the inequality constraint coefficient matrix and E, F, G is the equality constraint coefficient matrix.
7. The method for predicting the adjustable margin of the fire bundling system according to claim 1, wherein in the sixth step, the main-sub problem iteratively updates the uncertain set scene set: (17) Repeating the steps four to five until z is satisfied Finding robust viable interface power boundaries less than or equal to ε Consider interface nodes and different times, i.e 。
8. The method for predicting the adjustable margin of the wind-fire bundling system for adapting to uncertain conditions according to claim 1, wherein in the step seven, the final interface adjustable margin boundary of each period is output: And can get up/down margin, relative to some reference power pc 0 : (18) constraint boundaries for outbound plan tracking, roll optimization, and frequency support control.

Description

Wind-fire bundling system adjustable margin prediction method suitable for uncertain working conditions Technical Field The invention belongs to the technical field of operation and optimization of power systems, and particularly relates to an adjustable margin prediction method of a wind-fire bundling system suitable for uncertain working conditions. Background With the large-scale development of offshore wind power, the offshore decentralized access of a wind power plant is gradually changed to open sea cluster development and centralized delivery, and an operation mode of 'wind fire bundling and centralized delivery' is gradually formed. The fluctuation of wind power output is strong, and constraint such as slow climbing, minimum output of a thermal power unit is strict, and the wind power fluctuation cannot be responded timely, so that the adjustable margin of an output interface is changed rapidly along with working conditions and is difficult to determine. Considering the problems of weight and space limitation of the offshore platform, long-distance power transmission loss and the like, how to construct a low-cost and light-weight control method and platform has important significance for further developing long-distance and large-capacity power transmission. Under the above scenario, the current research does not consider the necessary supporting effect of the adjustable margin of the outgoing interface on the stable operation of the wind-fire bundling system, so that the calculation of the safety boundary of the interface power robustness under the uncertain change of the working condition is lacking, namely, the interface power plan or the local margin under the single working condition is usually obtained based on the deterministic prediction/optimal power flow calculation of each equipment unit, or the interface power feasible interval can be statistically estimated based on the feasibility test of multiple scenes/samples, but the problem that the on-line availability and accuracy, robustness and flexibility are difficult to be considered in the prediction result of the adjustable margin of the interface under the wind-fire bundling power transmission scenario in the prior art can be specifically stated as follows: (1) Modeling limitation is that the adjustable margin of a bundling power transmission interface is not uniformly considered, but discrete calculation is directly carried out on a thermal power, a fan, a circuit and a converter, so that the complexity of control is high, the dependence on synchronization of a pair of interface protocols and communication is high, the light weight and centralized processing are difficult to achieve, and online updating is difficult. (2) The promise risk is that the promise of the adjustable margin at the point-sending interface is not considered by the system uncertainty, the accuracy and the robustness of the forecasting result cannot be guaranteed, the influence of uncertain factors such as wind speed fluctuation, fan availability change, limited sending strategy, sea cable parameter deviation and temperature rise environment change is obvious, the existing method estimates the upper and lower power limits under a limited scene, and the promise margin is difficult to ensure to still meet the constraints of voltage, current and equipment capability under any implementation in an uncertain set, so that the promise margin has failure risk. (3) The method conservation is that the conventional robust method is easy to keep, the 'up-down adjustment margin' is not explicitly described, the uncertainty is deducted according to the worst case, the adjustable decision which is adjusted along with the uncertain realization is not introduced, the adjustable margin is obviously reduced, or the output is a single-point plan, and the 'up-down adjustment margin' which can be called to the outside is difficult to directly form. Disclosure of Invention The invention aims to solve the problems and provide the adjustable margin prediction method of the wind-fire bundling system, which is simple in structure and reasonable in design and is suitable for uncertain working conditions. The invention realizes the above purpose through the following technical scheme: A wind-fire bundling system adjustable margin prediction method suitable for uncertain working conditions comprises the steps of firstly establishing a power model, secondly parameterizing and expressing, thirdly establishing a robust optimization problem, fourthly determining candidate upper and lower bounds, fifth solving a sub-problem, sixth determining an interface power boundary, seventh outputting a prediction result; In the first step, a wind-fire bundling and sending interface point aggregation power model is established; in the second step, an uncertain set is constructed, and parameterization expression is carried out on the adjustable margin; In the third step, a two-stage adaptive robust optimization p