CN113033003B

CN113033003B - Load recovery model generation method and device, computer equipment and storage medium

Info

Publication number: CN113033003B
Application number: CN202110338005.5A
Authority: CN
Inventors: 顾雪平; 周光奇; 李少岩; 许浩波
Original assignee: North China Electric Power University
Current assignee: North China Electric Power University
Priority date: 2021-03-30
Filing date: 2021-03-30
Publication date: 2022-04-05
Anticipated expiration: 2041-03-30
Also published as: CN113033003A

Abstract

The application provides a load recovery model generation method, a load recovery model generation device, computer equipment and a storage medium, and relates to the technical field of power automation. The method comprises the following steps: establishing a load recovery model according to the access state parameters, the predicted yield value, the recovery state parameters of the power loss load and the predicted active power of the wind power plant to be recovered; configuring source load multiple uncertain factors for the load recovery model, and obtaining a main model in a prediction scene and a sub model in an error scene through decoupling; and solving the main model according to the predicted force value and the predicted active power to obtain a load recovery scheme under a prediction scene, acquiring a first prediction error uncertain set and a second prediction error uncertain set, solving the sub-model according to the load recovery scheme, the first prediction error uncertain set and the second prediction error uncertain set, and outputting the load recovery scheme if the solving result of the sub-model meets the preset recovery condition. The load recovery process can be safer through the application.

Description

Load recovery model generation method and device, computer equipment and storage medium

Technical Field

The invention relates to the technical field of power automation, in particular to a load recovery model generation method and device, computer equipment and a storage medium.

Background

With the rapid development of the manufacturing and control technology of the wind turbine generator, the wind power grid-connected capacity is promoted year by year, and when a large-scale power failure accident occurs in a high wind power permeability system, how to arrange the safe and rapid recovery of the load is a new problem which needs to be researched and solved urgently.

According to the existing scheme, a source load coordination recovery method considering that a wind power plant participates in load recovery is provided, a wind turbine generator has the advantages of being small in starting power requirement, high in starting speed and the like, and timely access is facilitated to accelerate the system recovery process.

However, under the restriction of random variation of wind speed, the wind turbine generator cannot continuously and stably output power, the output power of the wind turbine generator has the characteristics of uncertainty, poor schedulability and the like, the frequency modulation and voltage regulation capability of the wind turbine generator is poor, and more uncertainty factors are brought to system recovery decision and control due to participation of a wind power plant.

Disclosure of Invention

The present invention provides a method, an apparatus, a computer device and a storage medium for generating a load recovery model to ensure the security of the source load recovery process and accelerate the load recovery process.

In order to achieve the above purpose, the technical solutions adopted in the embodiments of the present application are as follows:

in a first aspect, an embodiment of the present application provides a load recovery model generation method, including:

establishing a load recovery model according to an access state parameter of a wind power plant to be recovered, a predicted power output value of the wind power plant to be recovered, a recovery state parameter of a power loss load and predicted active power of the power loss load;

configuring a source load multiple uncertainty factor in the load recovery model, the source load multiple uncertainty comprising: at least two uncertain factors of wind power plant and load;

decoupling the load recovery model after the uncertain factors are configured to obtain a main model in a prediction scene and a sub model in an error scene;

according to the predicted power output value of the wind power plant to be restored and the predicted active power of the power loss load, solving the main model to obtain a load restoration scheme under the prediction scene, wherein the load restoration scheme comprises the following steps: solving results of the access state parameters of the wind power plant to be recovered and solving results of the recovery state parameters of the power loss load;

acquiring a first prediction error uncertain set of the wind power plant to be recovered and a second prediction error uncertain set of the power loss load;

solving the submodel according to the load recovery scheme, the first uncertain prediction error set and the second uncertain prediction error set;

and if the solving result of the submodel meets the preset recovery condition, outputting the load recovery scheme.

Optionally, before the load recovery model configured with the uncertain factor is decoupled to obtain a main model in a prediction scene and a sub model in an error scene, the method further includes:

carrying out linearization processing on the load recovery model after the uncertain factors are configured by using a preset linearization method to obtain a target load recovery model;

the decoupling of the load recovery model after the uncertain factors are configured to obtain a main model in a prediction scene and a sub model in an error scene comprises the following steps:

and decoupling the target load recovery model to obtain the main model and the sub model.

Optionally, the establishing a load recovery model according to the access state parameter of the wind farm to be recovered, the predicted power output value of the wind farm to be recovered, the recovery state parameter of the power loss load, and the predicted active power of the power loss load includes:

and establishing the load recovery model according to the access state parameter of the wind power plant to be recovered, the first predicted power output value of the wind power plant to be recovered, the recovery state parameter of the power loss load, the predicted active power of the power loss load, the second predicted power output value of the recovered wind power plant and a preset parameter vector.

Optionally, the solving the main model according to the predicted power output value of the wind farm to be restored and the predicted active power of the power loss load to obtain the load restoration scheme in the prediction scene includes:

and solving the main model according to the first prediction force value, the prediction active power and the second prediction force value to obtain the load recovery scheme.

Optionally, the obtaining a first uncertain prediction error set of the wind farm to be restored and a second uncertain prediction error set of the power loss load includes:

according to the historical wind power prediction error, constructing a first ellipsoid uncertain set of the historical wind power prediction error by adopting a generalized ellipsoid set method;

converting the first ellipsoid uncertainty set by adopting a linear polyhedron set method to obtain a first prediction error uncertainty set;

and constructing the second prediction error uncertainty set by utilizing a box type constraint mode according to the historical power loss load prediction error.

Optionally, the converting the first ellipsoid uncertainty set by using a linear polyhedron set method to obtain the first prediction error uncertainty set includes:

performing rotational translation on the first ellipsoid uncertainty set to obtain a second ellipsoid uncertainty set after the rotational translation;

calculating a first vertex coordinate of the second ellipsoid uncertainty set, and constructing an initial linear polyhedron uncertainty set according to the first vertex coordinate;

inverting the initial linear polyhedron uncertain set according to the inversion relation of the first ellipsoid uncertain set and the second ellipsoid uncertain set to obtain an inverted initial linear polyhedron uncertain set;

and scaling the inverted initial linear polyhedron uncertain set to obtain the first prediction error uncertain set.

Optionally, the method further includes:

if the solution result of the submodel does not meet the preset recovery condition, adding an infeasible scene which does not meet the preset recovery condition in the solution result into the main model for iteration until the solution result of the submodel meets the preset recovery condition, wherein the infeasible scene comprises: an unrecoverable wind farm and an unrecoverable power loss load.

In a second aspect, an embodiment of the present application further provides a load recovery model generation apparatus, where the apparatus includes:

the recovery model establishing module is used for establishing a load recovery model according to the access state parameter of the wind power plant to be recovered, the predicted power output value of the wind power plant to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load;

an uncertainty factor configuration module, configured to configure a source load multiple uncertainty factor in the load recovery model, where the source load multiple uncertainty includes: at least two uncertain factors of wind power plant and load;

the decoupling module is used for decoupling the load recovery model after the uncertain factors are configured to obtain a main model in a prediction scene and a sub model in an error scene;

the main model solving module is used for solving the main model according to the predicted power output value of the wind power plant to be restored and the predicted active power of the power loss load to obtain a load restoration scheme under the prediction scene, and the load restoration scheme comprises the following steps: solving results of the access state parameters of the wind power plant to be recovered and solving results of the recovery state parameters of the power loss load;

the uncertain set acquisition module is used for acquiring a first uncertain set of prediction errors of the wind power plant to be recovered and a second uncertain set of prediction errors of the power-loss load;

the submodel solving module is used for solving the submodel according to the load recovery scheme, the first prediction error uncertainty set and the second prediction error uncertainty set;

and the output module is used for outputting the load recovery scheme if the solving result of the submodel meets the preset recovery condition.

Optionally, before the decoupling module, the apparatus further comprises:

the linearization processing module is used for carrying out linearization processing on the load recovery model after the uncertain factors are configured by utilizing a preset linearization method to obtain a target load recovery model;

the decoupling module is used for decoupling the target load recovery model to obtain the main model and the sub model.

Optionally, the recovery model establishing module is configured to establish the load recovery model according to the access state parameter of the wind farm to be recovered, the first predicted power output value of the wind farm to be recovered, the recovery state parameter of the power loss load, the predicted active power of the power loss load, the second predicted power output value of the recovered wind farm, and a preset parameter vector.

Optionally, the main model solving module is configured to solve the main model according to the first predicted power value, the predicted active power, and the second predicted power value, so as to obtain the load recovery scheme.

Optionally, the indeterminate set obtaining module includes:

the device comprises an ellipsoid uncertain set acquisition unit, a first ellipsoid uncertain set generation unit and a second ellipsoid set generation unit, wherein the ellipsoid uncertain set acquisition unit is used for constructing a first ellipsoid uncertain set of historical wind power prediction errors by adopting a generalized ellipsoid set method according to the historical wind power prediction errors;

a first uncertainty set obtaining unit, configured to convert the first ellipsoid uncertainty set by using a linear polyhedron set method to obtain the first prediction error uncertainty set;

and the second uncertain set acquisition unit is used for constructing a second uncertain set of prediction errors by using a box type constraint mode according to the prediction errors of the historical power-loss load.

Optionally, the first uncertain set obtaining unit includes:

the rotation translation subunit is configured to perform rotation translation on the first ellipsoid uncertainty set to obtain a second ellipsoid uncertainty set after the rotation translation;

the initial linear polyhedron uncertain set constructing subunit is used for calculating a first vertex coordinate of the second ellipsoid uncertain set and constructing an initial linear polyhedron uncertain set according to the first vertex coordinate;

the inversion subunit is configured to invert the initial linear polyhedron indeterminate set according to the inversion relationship between the first ellipsoid indeterminate set and the second ellipsoid indeterminate set, so as to obtain an inverted initial linear polyhedron indeterminate set;

and the scaling subunit is used for scaling the inverted initial linear polyhedron indeterminate set to obtain the first prediction error indeterminate set.

Optionally, the apparatus further comprises:

an iteration module, configured to, if the solution result of the sub-model does not satisfy the preset recovery condition, add an infeasible scenario that does not satisfy the preset recovery condition in the solution result to the main model for iteration until the solution result of the sub-model satisfies the preset recovery condition, where the infeasible scenario includes: an unrecoverable wind farm and an unrecoverable power loss load.

In a third aspect, an embodiment of the present application further provides a computer device, including: a processor, a storage medium and a bus, wherein the storage medium stores program instructions executable by the processor, when the computer device runs, the processor communicates with the storage medium through the bus, and the processor executes the program instructions to execute the steps of the load recovery model generation method according to any one of the above embodiments.

In a fourth aspect, the present application further provides a computer-readable storage medium, where a computer program is stored on the storage medium, and when the computer program is executed by a processor, the method for generating a load recovery model according to any one of the foregoing embodiments is performed.

The beneficial effect of this application is:

according to the load recovery model generation method, the load recovery model generation device, the computer equipment and the storage medium, a load recovery model is established according to the access state parameter of the wind power plant to be recovered, the predicted output value of the wind power plant to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load; configuring multiple uncertain factors of source load in a load recovery model, decoupling the load recovery model after the uncertain factors are configured, and obtaining a main model in a prediction scene and a sub model in an error scene; solving the main model according to the predicted output value of the wind power plant to be recovered and the predicted active power of the power loss load to obtain a load recovery scheme under a prediction scene, obtaining a first prediction error uncertain set and a second prediction error uncertain set of the power loss load of the wind power plant to be recovered, solving the sub-model according to the load recovery scheme, the first prediction error uncertain set and the second prediction error uncertain set, and outputting the load recovery scheme if the solving result of the sub-model meets preset recovery conditions. According to the scheme provided by the application, multiple uncertainty factors of the source load of the wind power plant and the power loss load are considered, the first uncertain prediction error set and the second uncertain prediction error set of the power loss load of the wind power plant are combined, the sub-model is utilized to perform safety verification on the load recovery scheme of the main model, the safety of the load recovery scheme obtained through the main model is higher, and the recovery process is safer.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.

Fig. 1 is a schematic flowchart of a first load recovery model generation method according to an embodiment of the present disclosure;

fig. 2 is a schematic flowchart of a second load recovery model generation method according to an embodiment of the present application;

fig. 3 is a schematic flowchart of a third load recovery model generation method according to an embodiment of the present application;

FIG. 4 is a block diagram of a first prediction error uncertainty set generation process provided by an embodiment of the present application;

fig. 5 is a schematic structural diagram of a load recovery model generation apparatus according to an embodiment of the present application;

fig. 6 is a schematic diagram of a computer device provided in an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention.

Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

In the description of the present application, it should be noted that if the terms "upper", "lower", etc. are used for indicating the orientation or positional relationship based on the orientation or positional relationship shown in the drawings or the orientation or positional relationship which is usually arranged when the product of the application is used, the description is only for convenience of describing the application and simplifying the description, but the indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation and operation, and thus, cannot be understood as the limitation of the application.

Furthermore, the terms "first," "second," and the like in the description and in the claims, as well as in the drawings, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be noted that the features of the embodiments of the present application may be combined with each other without conflict.

After a large-scale power failure accident occurs in a modern power system, the power supply recovery of the power system is usually realized from three stages of black start, grid reconstruction and load comprehensive recovery. Due to the randomness of the output of the wind power plant, the wind power plant is accessed in the grid frame reconstruction stage to be not beneficial to system power restoration, power failure can be caused again, and the main task of the load comprehensive restoration stage is to realize safe and quick input of the load in a wide area range and restore the power system to a normal operation state as soon as possible. At this stage, the power system is relatively strong and can be accessed to a wind farm to improve the efficiency of load recovery. The load recovery stage generally disperses the load recovery stage into a plurality of time steps with fixed duration, and the load is progressively recovered in time-sharing steps.

According to the load recovery model generation method provided by the embodiment of the application, the execution main body is computer equipment with the load recovery model generation function, a modeling system is installed on the computer equipment, and the load recovery model and the establishment and solving processes are completed in the modeling system. The computer device may be configured as an Intel (R) core (TM) i5 CPU with 8.00GB installed memory and The Modeling System may be The General Algebra Modeling System (GAMS).

Fig. 1 is a schematic flowchart of a first load recovery model generation method according to an embodiment of the present disclosure; as shown in fig. 1, the method includes:

s100: and establishing a load recovery model according to the access state parameter of the wind power plant to be recovered, the predicted power output value of the wind power plant to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load.

Specifically, the wind power plant to be recovered is the wind power plant which needs to be confirmed whether to be recovered or not at the current time step, and the access state parameter x_wsAccess status parameter x_wsA binary decision variable for representing whether to select to recover the wind power plant s at the current time step is represented by a value of 1, namely the wind power plant is selected to be recovered, and a predicted force value P of the wind power plant to be recovered_w,sAnd (4) predicting a force value after the wind power plant to be recovered is connected to the power system for power supply. The power-off load is the load which needs to be confirmed whether to be recovered or not at the current time step, and the recovery state parameter z_ikA binary decision variable of the current time step load recovery state, the value of which is 1 represents that the power-losing load on the feeder line k connected with the selected recovery node i and the predicted active power P of the power-losing load_dikAnd (4) predicting active power after power loss load power supply is restored. According to the access state parameter x of the wind power plant to be recovered_wsAnd the predicted output value P of the wind power plant to be recovered_w,sRecovery state parameter z of power loss load_ikPredicted active power P of power loss load_dikAnd establishing a load recovery model.

In an alternative embodiment, the S100 includes:

and establishing a load recovery model according to the access state parameter of the wind power plant to be recovered, the first predicted power output value of the wind power plant to be recovered, the recovery state parameter of the power loss load, the predicted active power of the power loss load, the second predicted power output value of the recovered wind power plant and a preset parameter vector.

Specifically, the recovered wind farm is the wind farm recovered before the current time step, and the recovered wind farm has an influence on the recovery of the power loss load at the current time step, so that a second predicted power output value P of the recovered wind farm needs to be considered when a load recovery model is established_wres,s. In addition, for protectionThe accuracy of the load recovery model is proved, the influence of other factors on the power loss load recovery is also considered, and the other factors are represented by the preset parameter vector psi.

According to the access state parameter x of the wind power plant to be recovered_wsFirst prediction power value P of wind power plant to be recovered_w,sRecovery state parameter z of power loss load_ikPredicted active power P of power loss load_dikAnd a second predicted power value P of the recovered wind power plant_wres,sAnd a predetermined parameter vector Ψ, the established load recovery model can be represented as:

wherein, P_dsumAnd g is an inequality constraint set in the load recovery model, and h is an equality constraint set in the load recovery model.

S200: and configuring source load multiple uncertainty factors in the load recovery model.

Specifically, the source load multiple uncertainty includes: at least two uncertainty factors of the wind farm and the load. In the load recovery stage, the predicted output value P of the wind power plant to be recovered in the load recovery model_w,sPredicted active power P of power loss load_dikAll the predicted values are predicted values and are difficult to accurately obtain in advance, so the predicted output value P of the wind power plant to be recovered needs to be predicted_w,sPredicted active power P of power loss load_dikAnd (6) optimizing.

The method and the device optimize a load recovery model based on a robust optimization framework and combined with multiple uncertain factors of source load, and predict a force value P of a wind power plant to be recovered in the load recovery model_w,sConfiguring uncertainty errors of predicted force values

The output value of the wind power field to be recovered is expressed as

Predicted active power P for power loss load_dikConfiguration of uncertainty error of predicted active power

So that the active power of the power-loss load is expressed as

And obtaining the load recovery model after the uncertain parameters are configured.

In an optional embodiment, a second predicted stress value P of the recovered wind power plant is introduced when the load recovery model is established_wres,sAnd a predetermined parameter vector Ψ, due to a second predicted power value P of the restored wind farm_wres,sIs also a predicted value, therefore, in the load recovery model, a second predicted power value P for the recovered wind farm_wres,sConfiguring uncertainty errors of predicted force values

The output value of the recovered wind power plant is expressed as

According to the formula (1), the load recovery model after the uncertain factors are configured can be represented as follows:

and phi represents an uncertain set and comprises a first uncertain set of prediction errors of the wind power plant to be recovered and a second uncertain set of prediction errors of the power-losing load.

S300: and decoupling the load recovery model after the uncertain factors are configured to obtain a main model in a prediction scene and a sub model in an error scene.

Specifically, because the load recovery model comprises the predicted value and the uncertain error value of the output value of the wind power plant to be recovered and the predicted value and the uncertain error value of the active power of the power-off load, in order to reduce the solving difficulty of the load recovery model, the load recovery model and the error scene can be decoupled in the prediction scene, and a main model and a sub model in the error scene in the prediction scene are constructed.

For example, the main model and the sub model obtained by decoupling the load recovery model according to the formula (2) can be respectively expressed as:

wherein a, A, B, C, D and E are corresponding coefficient matrixes,

representing a Hadamard product of a matrix; II (-) is all linear infeasible cut constraints generated by the previous L iterations, Φ^KAnd returning an infeasible scene for the neutron model in the K iteration.

Wherein eta is a vector consisting of relaxation variables, and is a non-negative variable; μ 1 and μ 2 are dual variables corresponding to inequality constraints and equality constraints, respectively.

S400: and solving the main model according to the predicted power output value of the wind power plant to be recovered and the predicted active power of the power loss load to obtain a load recovery scheme under the prediction scene.

Specifically, in the load recovery model, the predicted power value of the wind farm to be recovered and the predicted active power of the power loss load are known quantities, the access state parameter of the wind farm to be recovered and the recovery state parameter of the power loss load are unknown parameters, the load recovery model is decoupled into a main model in a prediction scene and a sub model in an error scene, and then the main model can be solved in a commercial solver according to the predicted power value of the wind farm to be recovered and the predicted active power of the power loss load, so that a load recovery scheme in the prediction scene is obtained. The load recovery scheme obtained by solving comprises the following steps: solving results of the access state parameters of the wind power plant to be recovered and solving results of the recovery state parameters of the power-losing load. For example, a commercial solver may be a CPLEX solver.

In an alternative embodiment, the force value P is predicted according to the first prediction_w,sPredicting active power P_dikAnd a second predicted force value P_wres,sAnd solving the main model of the formula (3) to obtain a load recovery scheme. The load recovery scheme includes: solving result x of access state parameter of wind power plant to be recovered_wAnd the solution result z, x of the recovery state parameter of the power loss load_wAccess status parameter x for a plurality of wind farms to be restored_wsThe value in the solving result vector is 1, which indicates that the wind power plant to be recovered corresponding to the '1' is selected and recovered at the current time step; and the value in the solving result vector is 0, which indicates that the wind power plant to be recovered corresponding to the '0' is not recovered at the current time step.

z is a recovery state parameter z of a plurality of power loss loads_ikThe value in the solving result vector is 1, which indicates that the power loss load corresponding to 1 is selected to be recovered at the current time step; and the value in the solving result vector is 0, which indicates that the power loss load corresponding to 0 is not recovered in the current time step selection.

S500: and acquiring a first prediction error uncertainty set of the wind power plant to be recovered and a second prediction error uncertainty set of the power loss load.

Specifically, a first prediction error uncertainty set of the wind farm to be recovered is a set of historical errors affecting the output value of the wind farm, and the output value historical errors within a certain time range can be selected by constructing the first prediction error uncertainty set. And the second prediction error uncertainty set of the power loss load is a set of historical errors according to the active power influencing the load, and the construction of the second prediction error uncertainty set can select the historical errors of the active power within a certain time range.

S600: and solving the submodel according to the load recovery scheme, the first prediction error uncertainty set and the second prediction error uncertainty set.

Specifically, a solving result x of an access state parameter of the wind power plant to be recovered is output in the main model_wAnd after the solution result z of the recovery state parameters of the power-losing load is obtained, verifying whether the load recovery scheme output by the main model can meet the safe operation constraint condition of the sub model according to the first uncertain prediction error set of the wind power plant to be recovered and the second uncertain prediction error set of the power-losing load.

In an optional implementation mode, the submodel of the formula (4) is solved, and since the submodel of the formula (4) is a max-min multilayer optimization model and cannot be directly solved by using the existing commercial solver, the inner layer min problem of the submodel is converted into a max problem by using a preset conversion method, so that the submodel is converted into a single-layer optimization model. For example, the preset conversion method is a strong dual theory of linear optimization, and the obtained single-layer optimized submodel may be expressed as:

s700: and if the solving result of the sub-model meets the preset recovery condition, outputting a load recovery scheme.

Specifically, according to the solution result of the submodel, if the solution result of the submodel meets the preset recovery condition, it is indicated that the load recovery scheme output by the main model meets the safe operation constraint condition of the submodel, and the stability and the safety of the power system are not affected by the load recovery according to the load recovery scheme, so that the load recovery scheme can be output for load recovery.

In an alternative embodiment, according to the solution result of the above equation (5), the preset recovery condition is that all the results in the vector η composed of the slack variables are 0, and if all the elements in η are 0, it indicates that the load recovery scheme output by the main model meets the safe operation constraint condition, and the load recovery scheme is output.

According to the load recovery model generation method provided by the embodiment of the application, a load recovery model is established according to the access state parameter of the wind power plant to be recovered, the predicted output value of the wind power plant to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load; configuring multiple uncertain factors of source load in a load recovery model, decoupling the load recovery model after the uncertain factors are configured, and obtaining a main model in a prediction scene and a sub model in an error scene; solving the main model according to the predicted output value of the wind power plant to be recovered and the predicted active power of the power loss load to obtain a load recovery scheme under a prediction scene, obtaining a first prediction error uncertain set and a second prediction error uncertain set of the power loss load of the wind power plant to be recovered, solving the sub-model according to the load recovery scheme, the first prediction error uncertain set and the second prediction error uncertain set, and outputting the load recovery scheme if the solving result of the sub-model meets preset recovery conditions. According to the method provided by the embodiment of the application, multiple uncertainty factors of the source load of the wind power plant and the power loss load are considered, the first uncertain prediction error set of the wind power plant and the second uncertain prediction error set of the power loss load are combined, the sub-model is utilized to perform safety check on the load recovery scheme of the main model, the safety of the load recovery scheme obtained through the main model is higher, and the recovery process is safer.

On the basis of the foregoing embodiment, an embodiment of the present application further provides a load recovery model generation method, where before the foregoing S300, the method further includes:

and carrying out linearization processing on the load recovery model after the uncertain factors are configured by using a preset linearization method to obtain a target load recovery model.

Specifically, because the load recovery model includes discrete variables and continuous variables and also includes highly nonlinear equation constraints, the load recovery model belongs to a typical non-convex nonlinear Mixed integer Programming model, the solution of the model is very difficult, in order to improve the calculation efficiency of the model and ensure the accuracy of the calculation result of the model, the load recovery model configured with uncertain factors is linearized by using a preset linearization method, and the model is converted into a Mixed Integer Linear Programming (MILP) model to obtain a target load recovery model. As an example, the preset linearization method is an approximately linearization trend method (Linear Programming adaptation of AC Power Flows, LPAC).

The above S300 includes:

and decoupling the target load recovery model to obtain a main model and a sub model.

Specifically, the linearized target recovery model is decoupled to obtain a main model and a sub model. The decoupling method is the same as S300, and is not described herein.

According to the load recovery model generation method provided by the embodiment of the application, the load recovery model with uncertain factors is subjected to linearization processing by using a preset linearization method to obtain a target load recovery model, and the target load recovery model is decoupled to obtain a main model and a sub model. By the method provided by the embodiment of the application, the load recovery model is subjected to linearization processing, so that the calculation efficiency of the model is higher, the accuracy of the calculation result of the model is ensured, and the main model can output an accurate load recovery scheme conveniently.

On the basis of the foregoing embodiment, an embodiment of the present application further provides a load recovery model generation method, fig. 2 is a flowchart illustrating a second load recovery model generation method provided in the embodiment of the present application, and as shown in fig. 2, the foregoing S500 includes:

s501: and constructing a first ellipsoid uncertain set of the historical wind power prediction error by adopting a generalized ellipsoid set method according to the historical wind power prediction error.

Specifically, because the prediction errors among the wind power plants have correlation, the uncertainty of the historical wind power prediction errors can be solved by adopting a generalized ellipsoid set method.

In an alternative embodiment, an ellipsoid uncertainty set which is minimum in volume and can cover all historical wind power prediction errors in a preset time range is constructed. Exemplary, ellipsoid uncertainty set Φ_w1Can be expressed as:

Φ_w1:＝{δ:(δ-c)^TQ(δ-c)≤1} (6)

wherein, δ represents the prediction error of the historical wind power output value, Q is a positive definite matrix representing the deviation of the symmetry axis of the ellipsoid from the coordinate axis direction, and c represents the central coordinate value of the ellipsoid.

S502: and converting the first ellipsoid uncertainty set by adopting a linear polyhedron set method to obtain a first prediction error uncertainty set.

In particular, due to the indeterminate set of ellipsoids Φ_w1The applicable model range is narrow, so in order to make up for the deficiency of the ellipsoid uncertainty set, a linear polyhedron combination can be constructed by adopting a linear polyhedron set method on the basis of the ellipsoid uncertainty set to obtain a first prediction error uncertainty set phi of the wind power plant to be recovered_w. The first set of prediction error uncertainties Φ_wThe method not only can highlight the correlation among the prediction errors of the wind power plant, but also can not change the form of the sub-model when the sub-model is solved.

S503: and constructing a second prediction error uncertainty set by utilizing a box type constraint mode according to the historical power loss load prediction error.

Specifically, the box constraint method is a common expression for describing parameter uncertainty, has a linear structure, and is convenient for solving the submodel. Due to the influence of factors such as load starting characteristics and the like, the difference of load characteristics before and after recovery is large, and the prediction errors of the historical power-loss load have no correlation, so that a second uncertain set phi of the prediction errors of the power-loss load can be constructed by using a box type constraint method_d。

For example, the second prediction error uncertainty set may be expressed as:

Φ_dan indeterminate set representing the power loss load prediction error,

and

respectively representing load prediction errors

Beta is an adjustment coefficient, and the value range is (0, 1)]To adjust for the conservation of the box uncertainty set.

According to the load recovery model generation method provided by the embodiment of the application, a first ellipsoid uncertain set of historical wind power prediction errors is constructed by adopting a generalized ellipsoid set method according to the historical wind power prediction errors; converting the first ellipsoid uncertainty set by adopting a linear polyhedron set method to obtain a first prediction error uncertainty set; and constructing a second prediction error uncertainty set by utilizing a box type constraint mode according to the historical power loss load prediction error. According to the method provided by the embodiment of the application, the relevance between the wind power plant prediction errors and the irrelevance between the power-off load prediction errors can be considered, the corresponding first prediction error uncertain set and the second prediction error uncertain set are respectively constructed, the sub-model can be ensured to accurately judge the safe operation condition of the load recovery scheme, the relevance between the wind power plant prediction errors can be fully considered by the linear polyhedron set method, the conservative model of the sub-model judgment is reduced, and reference is provided for an operator to balance the system recovery safety and rapidity.

On the basis of the foregoing embodiments, an embodiment of the present application further provides a load recovery model generation method, fig. 3 is a schematic flow chart of a third load recovery model generation method provided in the embodiment of the present application, and fig. 4 is a block diagram of a first prediction error uncertainty set generation process provided in the embodiment of the present application; as shown in fig. 3, S502 includes:

s5021: and rotationally translating the first ellipsoid uncertainty set to obtain a rotationally translated second ellipsoid uncertainty set.

Specifically, as shown in (1) in fig. 4, the first ellipsoid uncertainty set Φ obtained in S501 is_w1To facilitate the construction of the first prediction error uncertainty set, the first ellipsoid uncertainty set Φ is used_w1Rotating and translating to make the first ellipsoid uncertain set phi_w1The symmetric axis and the coordinate axis are superposed to obtain a second ellipsoid uncertainty set phi after the rotation translation_w2As shown in (2) of fig. 4. Exemplary, the second set of ellipsoid uncertainties Φ_w2Can be expressed as:

wherein δ' is a coordinate value of δ in the rotated coordinate axis, P and D are matrices obtained by orthogonal decomposition of Q, P is a transformation matrix, and D is a diagonal matrix with positive diagonal elements in the form of a second ellipsoid uncertainty set Φ_w2P and D satisfy the following relationship.

Q＝P^TDP＝P^-1DP (9)

S5022: and calculating a first vertex coordinate of the second ellipsoid uncertainty set, and constructing an initial linear polyhedron uncertainty set according to the first vertex coordinate.

Specifically, according to the solving method of the vertex coordinates of the ellipsoid, a second ellipsoid uncertainty set phi with the symmetric axis coincident with the coordinate axis after the rotation translation is calculated_w2And constructing an initial linear polyhedron uncertainty set based on the first vertex coordinates, as shown in (3) of fig. 4. For example, let D ═ diag (λ)₁,λ₂,…，λ_nw) The second ellipsoid uncertainty set phi can be known by the analysis formula (8)_w22n of_wThe coordinates of each vertex are:

s5023: and inverting the initial linear polyhedron uncertain set according to the inversion relation of the first ellipsoid uncertain set and the second ellipsoid uncertain set to obtain the inverted initial linear polyhedron uncertain set.

In particular, due to the indeterminate set Φ of the second ellipsoid_w2Is determined by the first set of ellipsoid uncertainty Φ_w1Obtained by rotation and translation on coordinate axes, and therefore can be determined according to the first ellipsoid uncertainty set phi_w1And a second set of uncertainty ellipsoids Φ_w2Inverse relationship betweenInverting the initial linear polyhedron uncertain set to make the symmetrical axis of the inverted initial linear polyhedron uncertain set parallel to the symmetrical axis of the first ellipsoid uncertain set to obtain the inverted initial linear polyhedron uncertain set phi_w3As shown in (4) in fig. 4. Exemplary, initial linear polyhedron uncertainty set Φ_w3Can be expressed as:

s5024: and scaling the inverted initial linear polyhedron uncertain set to obtain a first prediction error uncertain set.

Specifically, since the initial linear polyhedron indeterminate set obtained by conversion according to the ellipsoid indeterminate set cannot completely cover the historical wind power prediction error, the coverage degree of the historical wind power prediction error needs to be adjusted by scaling the inverted initial linear polyhedron indeterminate set to obtain the first prediction error indeterminate set Φ_w. Illustratively, the first set of prediction error uncertainties Φ_wCan be expressed as:

wherein gamma is a robust adjustment coefficient, the larger the value of gamma is, the larger the coverage rate of the historical scene is, and the value range of gamma is (0, gamma)_min]There is a minimum value for γ.

In an alternative embodiment, formula (5) includes a non-linear term, and if the solution to the submodel of formula (5) is to be implemented, the non-linear term needs to be converted into a linear term, so as to convert the submodel of formula (5) into a linear optimization model. Because the optimal solution of the linear optimization model is on the boundary of the random variable uncertainty set, the continuous random variable can be first predicted error uncertainty set phi_wAnd a second set of prediction error uncertainties Φ_dUsing a finite discrete extreme scene representation, i.e. uncertainty set phi of the second prediction error_dIs further shown as：

First set of uncertainty of prediction error Φ_wFurther expressed as:

wherein ξ is an indication

The variable 0-1 of the lower bound is taken.

As an example, table 1 shows the prediction error coverage of the first uncertainty set of prediction errors under different robust adjustment coefficients γ according to the embodiment of the present application.

TABLE 1 prediction error coverage for a first prediction error uncertainty set under different robust adjustment coefficients γ

According to the results in table 1, the influence of the robust adjustment coefficient γ on the prediction error coverage rate is known, and the larger the prediction error coverage rate is, the more conservative the uncertain set is.

According to the load recovery model generation method provided by the embodiment of the application, the first ellipsoid uncertainty set is subjected to rotational translation to obtain a second ellipsoid uncertainty set subjected to rotational translation; calculating a first vertex coordinate of the second ellipsoid uncertainty set, and constructing an initial linear polyhedron uncertainty set according to the first vertex coordinate; inverting the initial linear polyhedron uncertain set according to the inversion relation of the first ellipsoid uncertain set and the second ellipsoid uncertain set to obtain an inverted initial linear polyhedron uncertain set; and scaling the inverted initial linear polyhedron uncertain set to obtain a first prediction error uncertain set. According to the method provided by the embodiment of the application, the ellipsoid uncertainty set is converted into the linear polyhedron form uncertainty set, so that the first prediction error uncertainty set can highlight the correlation among the wind power plant prediction errors, the form of the sub-model cannot be changed when the sub-model is solved, and the verification accuracy of the sub-model on the load recovery scheme is ensured.

On the basis of the foregoing embodiment, an embodiment of the present application further provides a load recovery model generation method, where the method may further include:

and if the solution result of the sub-model does not meet the preset recovery condition, adding the infeasible scene which does not meet the preset recovery condition in the solution result into the main model for iteration until the solution result of the sub-model meets the preset recovery condition.

Specifically, according to the solution result of the submodel, if the solution result of the submodel does not satisfy the preset recovery condition, it indicates that the load recovery scheme output by the main model does not satisfy the safe operation constraint condition of the submodel, and an infeasible scene exists in the load recovery scheme output by the main model, where once the load recovery scheme is recovered, the infeasible scene will affect the stability and safety of the power system, and the infeasible scene includes: an unrecoverable wind farm and an unrecoverable power loss load. And returning the infeasible scene to the main model for iteration, recalculating the load recovery scheme by the main model, verifying by the sub model until the solution result of the sub model meets the preset recovery condition, completing iteration, and outputting the load recovery scheme.

In an alternative embodiment, according to the solution result of the above equation (5), if there is an element not equal to 0 in the vector η formed by the relaxation variables, the element not equal to 0 corresponds to an unrecoverable wind farm and/or an unrecoverable power loss load in the load recovery scheme.

By parameterizing infeasible scenes

And

characterizing and returning to the main model of formula (3), and iterating the main model. Exemplary, iterative Main model TableShown as follows:

according to the load recovery model generation method provided by the embodiment of the application, if the solution result of the sub-model does not meet the preset recovery condition, the infeasible scene which does not meet the preset recovery condition in the solution result is added into the main model for iteration until the solution result of the sub-model meets the preset recovery condition. According to the scheme provided by the embodiment of the application, the main model outputs the load recovery scheme meeting the safe operation constraint condition of the sub-model in an iteration mode, so that the load recovery is carried out according to the load recovery scheme output by the main model, the recovery process can be ensured to be safer, and the power failure fault is avoided in the load recovery process.

Table 2 shows a comparison between the load recovery scheme of the present application and other load recovery schemes

As can be seen from table 2, the scheme provided by the present application may lose a certain amount of load recovery, but from the analysis of the load recovery model, the problem that the line transmission capacity limits the backup transmission does not occur in the load recovery scheme output by the model, and the recovery process can be made safer. The comparison scheme is taken as an example for brief explanation, and the actual load value is assumed to be 1.3 times of the predicted value, namely the prediction error is the upper boundary value, the prediction error is obtained according to the table 1 and the wind power plant prediction error uncertain set calculation formula, the wind power output prediction error is the limit scene [ -13.4MW, -15.3MW ], it is seen that the actual load ratio is larger than the predicted value, the wind power output ratio is smaller than the predicted value, the system needs to call the reserved upper spare capacity, at this moment, the active power transmission capacity of the line exceeds the maximum transmission capacity of the line, and the safety of the recovery process is influenced.

Based on the analysis, the load recovery model generation method considering the source load uncertainty factor can bear the prediction error in a certain range, obtain a safer load recovery scheme, and avoid potential safety hazards such as unavailable spare capacity and the like although part of load recovery quantity may be lost, so that the recovery process is safer.

On the basis of the foregoing embodiments, an embodiment of the present application further provides a load recovery model generation apparatus, and fig. 5 is a schematic structural diagram of the load recovery model generation apparatus provided in the embodiment of the present application, and as shown in fig. 5, the apparatus includes:

the recovery model establishing module 100 is used for establishing a load recovery model according to the access state parameter of the wind power plant to be recovered, the predicted power output value of the wind power plant to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load;

an uncertainty factor configuration module 200, configured to configure source load multiple uncertainty factors in a load recovery model, where the source load multiple uncertainty includes: at least two uncertain factors of wind power plant and load;

the decoupling module 300 is configured to decouple the load recovery model configured with the uncertain factors to obtain a main model in a prediction scene and a sub model in an error scene;

the main model solving module 400 is configured to solve the main model according to the predicted power output value of the wind farm to be restored and the predicted active power of the power loss load to obtain a load restoration scheme in the prediction scenario, where the load restoration scheme includes: solving results of access state parameters of the wind power plant to be recovered and solving results of recovery state parameters of the power-losing load;

the uncertain set obtaining module 500 is used for obtaining a first uncertain set of prediction errors of the wind power plant to be recovered and a second uncertain set of prediction errors of the power-losing load;

the submodel solving module 600 is used for solving the submodel according to the load recovery scheme, the first prediction error uncertainty set and the second prediction error uncertainty set;

and the output module 700 is configured to output the load recovery scheme if the solution result of the sub-model meets the preset recovery condition.

Optionally, before the decoupling module 300, the apparatus further comprises:

the decoupling module 300 is configured to decouple the target load recovery model to obtain a main model and a sub model.

Optionally, the recovery model establishing module 100 is configured to establish a load recovery model according to an access state parameter of the wind farm to be recovered, a first predicted power output value of the wind farm to be recovered, a recovery state parameter of the power loss load, a predicted active power of the power loss load, a second predicted power output value of the recovered wind farm, and a preset parameter vector.

Optionally, the main model solving module 400 is configured to solve the main model according to the first predicted power value, the predicted active power, and the second predicted power value, so as to obtain the load recovery scheme.

Optionally, the indeterminate set obtaining module 500 includes:

the first uncertain set acquisition unit is used for converting the first ellipsoid uncertain set by adopting a linear polyhedron set method to obtain a first prediction error uncertain set;

and the second uncertain set acquisition unit is used for constructing a second uncertain set of prediction errors by using a box type constraint mode according to the historical power-loss load prediction errors.

Optionally, the first uncertainty set obtaining unit includes:

the rotation translation subunit is used for performing rotation translation on the first ellipsoid uncertainty set to obtain a second ellipsoid uncertainty set after the rotation translation;

the inversion subunit is used for inverting the initial linear polyhedron uncertain set according to the inversion relation between the first ellipsoid uncertain set and the second ellipsoid uncertain set to obtain an inverted initial linear polyhedron uncertain set;

and the scaling subunit is used for scaling the inverted initial linear polyhedron uncertain set to obtain a first prediction error uncertain set.

Optionally, the apparatus further comprises:

the iteration module is used for adding an infeasible scene which does not meet the preset recovery condition in the solution result into the main model for iteration if the solution result of the sub-model does not meet the preset recovery condition until the solution result of the sub-model meets the preset recovery condition, wherein the infeasible scene comprises: an unrecoverable wind farm and an unrecoverable power loss load.

The above-mentioned apparatus is used for executing the method provided by the foregoing embodiment, and the implementation principle and technical effect are similar, which are not described herein again.

These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

Fig. 6 is a schematic diagram of a computer device provided in an embodiment of the present application, where the computer device 800 includes: a processor 801, a storage medium 802 and a bus, wherein the storage medium 802 stores program instructions executable by the processor 801, when the computer device 800 runs, the processor 801 communicates with the storage medium 802 through the bus, and the processor 801 executes the program instructions to execute the method embodiment. The specific implementation and technical effects are similar, and are not described herein again.

Optionally, the invention also provides a program product, for example a computer-readable storage medium, comprising a program which, when being executed by a processor, is adapted to carry out the above-mentioned method embodiments.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.

The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and shall be covered by the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for generating a load recovery model, comprising:

2. The method of claim 1, wherein before decoupling the load recovery model after configuration of the uncertainty factor to obtain the main model in the prediction scenario and the sub model in the error scenario, the method further comprises:

3. The method according to claim 1, wherein the building of the load recovery model according to the access state parameter of the wind farm to be recovered, the predicted power output value of the wind farm to be recovered, the recovery state parameter of the power loss load and the predicted active power of the power loss load comprises:

4. The method according to claim 3, wherein the step of solving the main model according to the predicted power value of the wind farm to be restored and the predicted active power of the power loss load to obtain the load restoration scheme under the prediction scene comprises the steps of:

5. The method of claim 1, wherein obtaining a first uncertainty set of prediction errors for the wind farm to be restored and a second uncertainty set of prediction errors for the power loss load comprises:

6. The method of claim 5, wherein the transforming the first ellipsoid uncertainty set using a linear polyhedron aggregation method to obtain the first prediction error uncertainty set comprises:

7. The method of claim 1, further comprising:

8. A load recovery model generation apparatus, characterized in that the apparatus comprises:

9. A computer device, comprising: a processor, a storage medium and a bus, the storage medium storing program instructions executable by the processor, the processor and the storage medium communicating via the bus when the computer device is running, the processor executing the program instructions to perform the steps of the load recovery model generation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the load recovery model generation method according to any one of claims 1 to 7.