Multi-region frequency safety unit scheduling method and device based on Bernstein polynomial

Abstract

The invention discloses a Bernstein polynomial-based multi-region frequency safety unit scheduling method and device, and the method comprises the steps: constructing a multi-region frequency response model based on the dynamic primary frequency modulation response of a synchronous generator, and the optimal distribution of a speed regulator dead zone and frequency modulation reserve; carrying out approximation on an integral-differential algebraic equation of the multi-region frequency response model through a Bernstein polynomial to obtain a dynamic algebraic expression of the multi-region frequency, and constructing a multi-region frequency security constraint according to the dynamic algebraic expression of the multi-region frequency; embedding the multi-region frequency security constraint into a unit commitment problem, and constructing a multi-region frequency security constraint unit commitment model; decomposing a mixed integer programming problem of the multi-region frequency security constraint unit commitment model into a main problem and a sub-problem through an adaptive double-layer decomposition algorithm; and solving to obtain a multi-region frequency safety unit scheduling scheme. According to the method, the frequency dynamic difference between the regions is accurately captured, and the frequency safety of the high-proportion new energy power system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system dispatching and frequency security, and particularly relates to a multi-region frequency security unit dispatching method and device based on Bernstein polynomials. Background Art

[0002] With the large-scale access of renewable energy (VRE) mainly composed of wind and light to the power grid, the inertia level and primary frequency response (PFR) ability of the power system have been significantly reduced, threatening frequency security. Most VRE units are far from the load center, resulting in regional differences in the distribution of system inertia and prone to inter-regional frequency oscillations and over-limit problems after faults. Therefore, it is urgent to optimize the unit start-stop and PFR reserve in unit commitment (UC) overall to ensure the frequency security of the multi-region system.

[0003] Currently, existing research coordinates optimization and frequency stability by introducing frequency security constraints into UC. The main methods include: 1) Based on the equivalent single-machine model, aggregating all system synchronous machines into a single machine and deriving the frequency dynamics of the center of inertia (COI), and generating constraints by combining piecewise linearization. However, this method assumes that the unit response characteristics are the same, ignores regional frequency differences, has large linearization errors and complex variable scales, and cannot obtain the optimal configuration of unit-level PFR; 2) Based on the linear ramp model, assuming that the PFR dynamics is a linear function, but ignoring the actual response process and relying on transfer time parameters that are difficult to accurately obtain, resulting in conservative results; 3) Using machine learning to approximate frequency dynamics, although reducing the computational complexity, the model effect depends on sample data and has insufficient scalability. In addition, when analyzing multi-region frequency security UC, there are still some limitations in existing technologies, and it is difficult to balance the solution efficiency and accuracy of complex integral-differential equations.

[0004] Therefore, there is an urgent need for a multi-region frequency security constraint optimization method that takes into account both accuracy and efficiency to support the safe and stable operation of the power system under high-proportion VRE access. Summary of the Invention

[0005] To achieve the above object, the present invention provides the following technical solutions: A multi-region frequency security unit dispatching method based on Bernstein polynomials, including:

[0006] To achieve the above object, the present invention provides the following technical solutions: A multi-region frequency security unit dispatching method based on Bernstein polynomials, including:

[0007] Based on the optimal allocation of the dynamic primary frequency regulation response of synchronous generators, governor dead zones, and frequency regulation reserves, a multi-area frequency response model is constructed;

[0008] The integral-differential algebraic equations of the multi-area frequency response model are approximated by Bernstein polynomials to obtain an algebraic expression for the multi-area frequency dynamics;

[0009] Based on the algebraic expression of the multi-area frequency dynamics, multi-area frequency security constraints are constructed;

[0010] The multi-area frequency security constraints are embedded into the unit commitment problem to construct a multi-area frequency security constrained unit commitment model;

[0011] Through the adaptive two-layer decomposition algorithm, the mixed-integer programming problem of the multi-area frequency security constrained unit commitment model is decomposed into a master problem and a sub-problem; by solving the master problem and the sub-problem, a multi-area frequency security unit scheduling scheme is obtained.

[0012] As a preferred scheme of the multi-area frequency security unit scheduling method based on Bernstein polynomials, the expression of the multi-area frequency response model is:

[0013] When the frequency deviation is less than the governor dead zone:

[0014]

[0015] When the frequency deviation is greater than the governor dead zone:

[0016]

[0017]

[0018] In the formula, H i is the total inertia of area i; Δf i (τ) is the frequency deviation at time τ in area i; k i L is the load damping ratio of area i; is the load power of area i; is the total primary frequency regulation response power at time τ in area i; is the disturbance power of area i; Ψ i is the set of areas connected to area i; X ij is the reactance of the tie line between area i and area j; η is the integration variable; Δf i DB is the action dead zone value of area i; is the critical point time of the governor dead zone; is the set of synchronous generators in area i; Primary frequency regulation power of generator g at time τ; T g Response constant of generator g; K g Droop coefficient of generator g; U g Is a 0-1 variable indicating whether unit g is online, and is 1 if so.

[0019] As an optimal solution of the multi-area frequency security unit scheduling method based on Bernstein polynomials, the algebraic expression of the multi-area frequency dynamics is:

[0020]

[0021] In the formula, l s Divides the time range L of the frequency dynamic response into multiple segments, and the length of each segment; both i and j are regions; Is The column vector composed of; Is the Bernstein spline coefficient corresponding to the time-domain frequency; B m,k (τ) is the mth-degree Bernstein polynomial; Is The initial value of; J T Is the coefficient matrix of the integral property of the Bernstein polynomial; Is the column vector of Bernstein spline coefficients corresponding to the total frequency regulation power in the time domain; Is the column vector of Bernstein spline coefficients corresponding to the disturbance power; Is the column vector of Bernstein spline coefficients corresponding to the frequency regulation power of unit g in the time domain; Is The initial value of; Δf i DB,B Is the dead zone value Δf i DB The vector composed of; m is the order of the Bernstein polynomial; k = 0, 1…, m.

[0022] As an optimal solution of the multi-area frequency security unit scheduling method based on Bernstein polynomials, the multi-area frequency security constraints include: initial moment frequency change rate constraint, frequency lowest point deviation constraint, quasi-steady state frequency deviation constraint and primary frequency regulation reserve constraint;

[0023] The expression of the initial moment frequency change rate constraint is:

[0024]

[0025] In the formula, RoCoF max Is the limit value of the initial moment frequency change rate;

[0026] The expression of the frequency lowest point deviation constraint is:

[0027]

[0028] Wherein, V is the enhancement matrix for the transformation of the inequality equation and is a constant matrix; Δf max is the deviation limit of the lowest frequency point;

[0029] The expression of the quasi-steady-state frequency deviation constraint is:

[0030]

[0031] Wherein, Δf ss,max is the quasi-steady-state frequency deviation limit;

[0032] The expression of the primary frequency regulation reserve constraint is:

[0033]

[0034] Wherein, R g is the primary frequency regulation reserve power of unit g.

[0035] As an optimal solution of the multi-area frequency security unit scheduling method based on Bernstein polynomials, the objective function expression of the multi-area frequency security constrained unit commitment model is:

[0036]

[0037] Wherein, Ω G 、Ω T 、Ω W 、Ω B are the sets of thermal power units, scheduling periods, wind turbines, and buses respectively; U gt is a 0-1 variable indicating whether unit g is online in period t, 1 means online, and 0 means offline; R gt is the primary frequency regulation reserve power of unit g in period t; are the start-stop cost and primary frequency regulation reserve cost of unit g respectively; is the fuel cost of unit g in period t; c W and c B are the unit wind curtailment cost and load shedding cost respectively; ΔP wt is the wind curtailment power of wind turbine w in period t; is the load shedding power of bus b in period t.

[0038] As an optimal solution of the multi - area frequency - secure unit commitment method based on Bernstein polynomials, in the process of decomposing the mixed - integer programming problem of the multi - area frequency - secure constrained unit commitment model into the master problem and the sub - problem through the adaptive two - layer decomposition algorithm, the mixed - integer programming problem is decomposed into the master problem and the sub - problem for iterative calculation through the Benders decomposition strategy and the adaptive low - frequency moment estimation model; through solving the master problem, the unit start - stop and frequency regulation reserve allocation strategies are obtained; through solving the sub - problem, the multi - area frequency - secure constraints are verified.

[0039] The present invention also provides a multi - area frequency - secure unit commitment device based on Bernstein polynomials. Based on the above - mentioned multi - area frequency - secure unit commitment method based on Bernstein polynomials, it includes:

[0040] A multi - area frequency response model construction module, which is used to construct a multi - area frequency response model based on the dynamic primary frequency regulation response of synchronous generators, the governor dead zone, and the optimal allocation of frequency regulation reserves;

[0041] A multi - area frequency - dynamic algebraic expression acquisition module, which is used to approximate the integral - differential algebraic equation of the multi - area frequency response model through Bernstein polynomials to obtain an algebraic expression of multi - area frequency dynamics;

[0042] A multi - area frequency - secure constraint construction module, which is used to construct multi - area frequency - secure constraints according to the algebraic expression of the multi - area frequency dynamics;

[0043] A multi - area frequency - secure constrained unit commitment model construction module, which is used to embed the multi - area frequency - secure constraints into the unit commitment problem to construct a multi - area frequency - secure constrained unit commitment model;

[0044] A multi - area frequency - secure constrained unit commitment model solution module, which is used to decompose the mixed - integer programming problem of the multi - area frequency - secure constrained unit commitment model into a master problem and a sub - problem through an adaptive two - layer decomposition algorithm; by solving the master problem and the sub - problem, a multi - area frequency - secure unit commitment solution is obtained.

[0045] As an optimal solution of the multi - area frequency - secure unit commitment device based on Bernstein polynomials, in the multi - area frequency response model construction module, the expression of the multi - area frequency response model is:

[0046] When the frequency deviation is less than the governor dead zone:

[0047]

[0048] When the frequency deviation is greater than the governor dead zone:

[0049]

[0050] In the formula, H i is the total inertia of area i; Δf i (τ) is the frequency deviation at time τ in area i; is the load damping rate of area i; is the load power of area i; is the total primary frequency regulation response power at time τ in area i; is the disturbance power of area i; Ψ i is the set of areas connected to area i; X ij is the reactance of the tie line between area i and area j; η is the integration variable; Δf i DB is the action dead zone value of area i; is the critical point time of the governor dead zone; is the set of synchronous generators in area i; is the primary frequency regulation power of generator g at time τ; T g is the response constant of generator g; K g is the droop coefficient of generator g; U g is a 0-1 variable indicating whether unit g is online, and it is 1 if so.

[0051] As an optimal solution of the multi-area frequency security unit scheduling device based on Bernstein polynomials, in the multi-area frequency dynamics algebraic expression acquisition module, the algebraic expression of the multi-area frequency dynamics is:

[0052]

[0053] In the formula, l s is the length of each segment when dividing the time range L of the frequency dynamic response into multiple segments; both i and j are areas; is the column vector composed of; is the Bernstein spline coefficient corresponding to the time-domain frequency; B m,k (τ) is the m-th Bernstein polynomial; is the initial value of; J T is the coefficient matrix of the integral property of the Bernstein polynomial; is the column vector of Bernstein spline coefficients corresponding to the time-domain total frequency regulation power; ΔP i L,B is the column vector of Bernstein spline coefficients corresponding to the disturbance power; is the column vector of Bernstein spline coefficients corresponding to the frequency regulation power of unit g in the time domain; is the initial value of; Δfi DB,B is the dead - zone value Δf i DB is a column vector composed of; m is the order of the Bernstein polynomial; k = 0, 1…, m.

[0054] As an optimal solution of the multi - area frequency - secure unit scheduling device based on the Bernstein polynomial, in the multi - area frequency - security constraint construction module, the multi - area frequency - security constraints include: initial - time frequency change rate constraint, frequency lowest - point deviation constraint, quasi - steady - state frequency deviation constraint, and primary frequency regulation reserve constraint;

[0055] The expression of the initial - time frequency change rate constraint is:

[0056]

[0057] In the formula, RoCoF max is the limit value of the initial - time frequency change rate;

[0058] The expression of the frequency lowest - point deviation constraint is:

[0059]

[0060] In the formula, V is the augmented matrix for the transformation of the inequality equation, which is a constant matrix; Δf max is the limit value of the frequency lowest - point deviation;

[0061] The expression of the quasi - steady - state frequency deviation constraint is:

[0062]

[0063] In the formula, Δf ss,max is the limit value of the quasi - steady - state frequency deviation;

[0064] The expression of the primary frequency regulation reserve constraint is:

[0065]

[0066] In the formula, R g is the frequency regulation reserve power of unit g.

[0067] As an optimal solution of the multi - area frequency - secure unit scheduling device based on the Bernstein polynomial, in the multi - area frequency - security constraint unit commitment model construction module, the objective - function expression of the multi - area frequency - security constraint unit commitment model is:

[0068]

[0069] In the formula, Ω G 、Ω T 、ΩW , Ω B are the sets of thermal power units, scheduling periods, wind turbines, and buses, respectively; U gt is a 0-1 variable indicating whether unit g is online in period t, where 1 means online and 0 means offline; R gt is the primary frequency regulation reserve power of unit g in period t; are the start-up and shut-down costs and the primary frequency regulation reserve costs of unit g, respectively; is the fuel cost of unit g in period t; c W and c B are the unit wind power curtailment cost and the load shedding cost, respectively; ΔP wt is the wind power curtailment power of wind turbine w in period t; is the load shedding power of bus b in period t.

[0070] As an optimal solution of the multi-region frequency security unit scheduling device based on Bernstein polynomials, in the multi-region frequency security constrained unit commitment model solving module, in the process of decomposing the mixed integer programming problem of the multi-region frequency security constrained unit commitment model into the master problem and the sub-problem by the adaptive two-layer decomposition algorithm, through the benders decomposition strategy and the adaptive low frequency point moment estimation model, the mixed integer programming problem is decomposed into the master problem and the sub-problem for iterative calculation; through the solution of the master problem, the unit start-up and shut-down and frequency regulation reserve allocation strategies are obtained; through the solution of the sub-problem, the multi-region frequency security constraints are verified.

[0071] The present invention has the following advantages: Based on the dynamic primary frequency regulation response of a synchronous generator, the dead zone of a governor, and the optimal allocation of frequency regulation reserve, the present invention constructs a multi-area frequency response model; approximates the integral-differential algebraic equations of the multi-area frequency response model through Bernstein polynomials to obtain an algebraic expression of multi-area frequency dynamics; constructs multi-area frequency security constraints according to the algebraic expression of multi-area frequency dynamics; embeds the multi-area frequency security constraints into the unit commitment problem to construct a multi-area frequency security constraint unit commitment model; decomposes the mixed-integer programming problem of the multi-area frequency security constraint unit commitment model into a master problem and a sub-problem through an adaptive two-layer decomposition algorithm; and obtains a multi-area frequency security unit scheduling scheme by solving the master problem and the sub-problem. The present invention simultaneously considers the differential dynamic frequency response of units, the action dead zone, and the optimal allocation of frequency regulation reserve in the multi-area frequency response model, and uses Bernstein polynomials to model the integral-differential algebraic equations in the time domain. On this basis, multi-area frequency security constraints based on Bernstein polynomial approximation are obtained. In addition, an adaptive two-layer decomposition algorithm is proposed to decompose the original problem into a master problem and a sub-problem, and at the same time, the problem scale is adaptively reduced by estimating the moment of the lowest frequency point in the sub-problem, greatly reducing the optimization dimension and time of the problem. The multi-area frequency response model dynamics based on Bernstein polynomial approximation of the present invention is tight and can more accurately simulate multi-area frequency dynamics compared with the prior art. Compared with the traditional single-area model based on the center of inertia, the frequency index values of the proposed method more accurately reflect the actual situation of regional frequency response differences. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, other implementation drawings can be obtained by extension according to the provided drawings without creative efforts.

[0073] The structures, ratios, sizes, etc. illustrated in this specification are only used to cooperate with the content disclosed in the specification for those who are familiar with this technology to understand and read, and are not used to limit the limited conditions under which the present invention can be implemented. Therefore, they do not have a substantial technical meaning. Any modification of the structure, change of the proportional relationship, or adjustment of the size should still fall within the scope covered by the technical content disclosed in the present invention without affecting the effects that the present invention can produce and the purposes that can be achieved.

[0074] Figure 1 It is a schematic flow chart of the multi-area frequency security unit scheduling method based on Bernstein polynomials provided in Embodiment 1 of the present invention;

[0075] Figure 2 Schematic diagram of a two - area 12 - node system in a possible embodiment provided in Embodiment 1 of the present invention;

[0076] Figure 3 Schematic diagram of the comparison result between the multi - area frequency dynamics of the present invention and Simulink numerical simulation in a possible embodiment provided in Embodiment 1 of the present invention;

[0077] Figure 4 Schematic diagram of the deviation result of the lowest frequency point between the present invention and the inertia - center - based model in a possible embodiment provided in Embodiment 1 of the present invention;

[0078] Figure 5 Schematic diagram of the initial frequency change rate result between the present invention and the inertia - center - based model in a possible embodiment provided in Embodiment 1 of the present invention;

[0079] Figure 6 Schematic diagram of the architecture of a multi - area frequency - secure unit scheduling device based on Bernstein polynomials provided in Embodiment 2 of the present invention. Specific implementation manners

[0080] The following specific embodiments illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0081] Embodiment 1

[0082] See Figure 1 , Embodiment 1 of the present invention provides a multi - area frequency - secure unit scheduling method based on Bernstein polynomials, including the following steps:

[0083] S1. Based on the dynamic primary frequency regulation response of synchronous generators, the dead - band of governors, and the optimal allocation of frequency regulation reserves, construct a multi - area frequency response model;

[0084] S2. Approximate the integral - differential algebraic equations of the multi - area frequency response model through Bernstein polynomials to obtain an algebraic expression of multi - area frequency dynamics;

[0085] S3. According to the algebraic expression of the multi - area frequency dynamics, construct multi - area frequency security constraints;

[0086] S4. Embed the multi - area frequency security constraints into the unit commitment problem to construct a multi - area frequency - secure unit commitment model;

[0087] S5. By using the adaptive two-layer decomposition algorithm, decompose the mixed-integer programming problem of the multi-region frequency security-constrained unit commitment model into a master problem and a sub-problem; by solving the master problem and the sub-problem, obtain a multi-region frequency security unit scheduling scheme.

[0088] In this embodiment, in step S1, based on the dynamic primary frequency regulation response of synchronous generators, the governor dead zone, and the optimal allocation of frequency regulation reserve, construct a multi-region frequency response model;

[0089] Specifically, the expression of the multi-region frequency response model is:

[0090] When the frequency deviation is less than the governor dead zone:

[0091]

[0092] When the frequency deviation is greater than the governor dead zone:

[0093]

[0094] In the formula, H i is the total inertia of region i; Δf i (τ) is the frequency deviation at time τ in region i; is the load damping ratio of region i; is the load power of region i; is the total primary frequency regulation response power at time τ in region i; is the disturbance power of region i; Ψ i is the set of regions connected to region i; X ij is the reactance of the tie line between region i and region j; η is an integral variable; Δf i DB is the action dead zone value of region i; is the critical point time of the governor dead zone; is the set of synchronous generators in region i; is the primary frequency regulation power of generator g at time τ; T g is the response constant of generator g; K g is the droop coefficient of generator g; U g is a 0-1 variable indicating whether unit g is online, and if so, it is 1.

[0095] In this embodiment, in step S2, approximate the integral-differential algebraic equation of the multi-region frequency response model by using Bernstein polynomials to obtain an algebraic expression of multi-region frequency dynamics;

[0096] Specifically, divide the time range L of the frequency dynamic response into multiple segments, and the length of each segment is ls Normalize the range of each l s to [0, 1], and then use Bernstein polynomial splines to dynamically approximate the frequency of each segment. The frequency Δf i,s (τ) of segment s in region i is approximated by the Bernstein polynomial as follows:

[0097]

[0098] where B m,k (τ) is the m-th degree Bernstein polynomial; is the Bernstein spline coefficient corresponding to Δf i,s (τ); and B m (τ) are column vectors composed of and B m,k (τ), respectively.

[0099] Similarly, the Bernstein polynomial approximations of other continuous-time functions in equations (3)-(7) are respectively:

[0100]

[0101] Then integrate both sides of equation (3) from 0 to τ to obtain:

[0102]

[0103] After normalizing the time range, to keep the physical meaning of the equation unchanged, the differential term in equation (14) needs to be divided by l s , while the integral term needs to be multiplied by l s . Equation (14) can be further written as:

[0104]

[0105] For part a in equation (15), apply the integral property of the Bernstein polynomial and equation (8); for part b in equation (15), apply the integral property of the Bernstein polynomial and equation (12); and for part c, since there are two integral signs, apply the integral property of the Bernstein polynomial to both integrals. Therefore, equation (15) is equivalent to:

[0106]

[0107] where J T is the coefficient matrix of the integral property of the Bernstein polynomial; is the column vector of the Bernstein spline coefficients corresponding to ; ΔP i L,B is the column vector of the Bernstein spline coefficients corresponding to The Bernstein spline coefficient column vector; m is the order of the Bernstein polynomial; k = 0, 1…, m; both i and j are regions; For segment s Initial value It is necessary to satisfy the initial conditions shown in Equation (17).

[0108] Similarly, Equations (5)-(7) are equivalent to:

[0109]

[0110]

[0111] In the formula, Corresponding to The Bernstein spline coefficient column vector; Is the Initial value of segment s; Δf i DB,B Is the dead zone value Δf i DB The column vector composed of.

[0112] In this embodiment, in step S3, according to the algebraic expression of the multi-region frequency dynamics, a multi-region frequency security constraint is constructed;

[0113] Specifically, the multi-region frequency security constraint considers three indicators including: the rate of change of frequency at the initial moment, the deviation of the lowest frequency point, and the deviation of the quasi-steady state frequency. It is assumed that the limit values of the three indicators for all regions are the same.

[0114] Rate of change of frequency at the initial moment constraint:

[0115] At the initial moment τ = 0 + Is located in segment s = 1. Applying the differential property of the Bernstein polynomial, the RoCoF constraint at the initial moment for any region i can be expressed as:

[0116]

[0117] In the formula, RoCoF max Is the limit value of the rate of change of frequency at the initial moment.

[0118] Deviation of the lowest frequency point constraint:

[0119] The frequency deviation at all times should be less than its maximum limit value. Applying the solution space transformation property of the Bernstein polynomial, the deviation of the lowest frequency point constraint for any region i can be expressed as:

[0120]

[0121] In the formula, V is the augmented matrix for the transformation of the inequality equation, which is a constant matrix; Δfmax is the lowest frequency deviation limit value.

[0122] Quasi-steady state frequency deviation constraint:

[0123] Due to the oscillation attenuation, the QSS frequency deviation index does not show significant regional differences. Using the final value theorem, the QSS frequency deviation constraint expression can be deduced as:

[0124]

[0125] In the formula, Δf ss,max is the quasi-steady state frequency deviation limit value.

[0126] Primary frequency regulation reserve constraint:

[0127] The primary frequency regulation power of each unit cannot exceed its primary frequency regulation reserve power. Therefore, this constraint can also determine the optimal reserve of each unit. Applying the solution space transformation property of Bernstein polynomials, the primary frequency regulation reserve constraint of each unit can be expressed as:

[0128]

[0129] In the formula, R g is the frequency regulation reserve power of unit g.

[0130] In this embodiment, the complete multi-area frequency security constraint is given by equations (16)-(25), including the multi-area frequency response model based on Bernstein polynomial approximation (equations (16)-(21)); the frequency change rate constraint (equation (22)); the lowest frequency point constraint (equation (23)); the quasi-steady state frequency deviation constraint (equation (24)) and the primary frequency regulation reserve constraint (equation (25)).

[0131] In this embodiment, in step S4, the multi-area frequency security constraint is embedded into the unit commitment problem to construct a multi-area frequency security constraint unit commitment model;

[0132] Specifically, the expression of the multi-area frequency security constraint unit commitment model is:

[0133]

[0134]

[0135] In the formula, Ω G and Ω T and Ω W and Ω B and Ω L are the sets of thermal power units, scheduling periods, wind power units, buses, and branches respectively; U gt is a 0-1 variable indicating whether unit g is online at time t, where 1 means online and 0 means offline; Pgt is the active power output of unit g during period t; R gt is the primary frequency regulation reserve power of unit g during period t; ΔP wt is the curtailed wind power of wind turbine w during period t; is the load shedding power of bus b during period t; is the fuel cost of unit g during period t; p gtμ is the active power output of unit g in the μ-th segment during period t; c gμ are respectively the start-up and shut-down cost, primary frequency regulation reserve cost, on-line fixed cost and fuel marginal cost of the μ-th segment of unit g; c W and c B are respectively the unit curtailed wind cost and load shedding cost; N μ is the number of segments of the fuel cost; and are respectively the maximum and minimum technical output of unit g; H i,t is the total inertia of region i during period t; f 0 is the reference frequency, i.e., 50 Hz. h g is the inertia of unit g; N a is the number of regions; and are respectively the ramp-up limit and ramp-down limit of unit g; and are respectively the minimum running time and minimum shutdown time of unit g; and P wt are respectively the predicted output and actual output of wind turbine w during period t; and are respectively the predicted load and actual load of bus b during period t; ζ lg 、ζ lw 、ζ lb are respectively the power transfer factors of branch l with respect to unit g, wind turbine w and bus b; is the power flow limit of branch l.

[0136] In the above formula, formula (26) is the objective function of the multi - area frequency security - constrained unit commitment model, including the start - up / shut - down cost, fuel cost, primary frequency regulation reserve cost, wind curtailment and load shedding penalty cost. The multi - area frequency security constraints (16) - (25) ignore the index t of the UC time period, and formula (27) indicates that the frequency security constraints need to be satisfied at any time t. Constraint (28) linearizes the fuel cost of the unit piece - by - piece. Formula (29) calculates the total inertia of area i. Formulas (30) - (32) are the upper and lower limits of the unit output constraints (reserving primary frequency regulation reserve power), ramping - up constraints and ramping - down constraints respectively. Formulas (33) - (36) are the minimum on - time / off - time constraints of the unit. Formulas (37) - (39) are the wind curtailment constraint, load shedding constraint and system power balance constraint respectively. Formula (40) is the DC power flow constraint of the branch.

[0137] In this embodiment, in step S5, through the adaptive two - layer decomposition algorithm, the mixed - integer programming problem of the multi - area frequency security - constrained unit commitment model is decomposed into a master problem and a sub - problem; by solving the master problem and the sub - problem, a multi - area frequency - secure unit scheduling scheme is obtained.

[0138] Specifically, the adaptive two - layer decomposition algorithm decomposes the large - scale mixed - integer programming problem into a master problem and a sub - problem for iterative calculation through the Benders decomposition strategy and the adaptive low - frequency point - time estimation model, where the master problem solves the unit start - up / shut - down and frequency regulation reserve allocation, and the sub - problem verifies the multi - area frequency security constraints.

[0139] Specifically, there are a large number of 0 - 1 variables and continuous variables in the multi - area frequency security constraints (16) - (25). Therefore, the proposed multi - area frequency security - constrained unit commitment model is a complex large - scale mixed - integer programming problem, which is difficult to solve directly with a commercial solver and requires the design of a reasonable acceleration algorithm.

[0140] Among them, the frequency change rate index and the frequency lowest - point deviation index are affected by the inter - area frequency response difference, while the quasi - steady - state frequency deviation index is not related to it. Therefore, the multi - area frequency response model (16) - (21), the frequency change rate constraint (22) and the frequency lowest - point deviation constraint (23) based on the Bernstein polynomial approximation do not need to consider the complete frequency dynamic segments s. For example, when L = 30S, assume l s = 2, where s = 1,.., 15. Assume that the frequency lowest - point time t nadir appears at s = 4 (i.e., between 6 - 8S). Therefore, only the constraints (16) - (23) at s = 1, 2, 3, 4 need to be considered, rather than the complete segments s = 1, 2,…, 15. This will greatly reduce the scale of the model.

[0141] Ignore the regional frequency response differences and use the frequency dynamics based on the center of inertia to estimate the segment s at the moment of the lowest point nadir . When the frequency reaches the lowest point, the primary frequency regulation response power of the entire system is equal to the disturbance power, and the dynamics of the primary frequency regulation response power are simulated using a linear ramp:

[0142]

[0143] In the formula, ΔP M (τ) is the total increment of the mechanical power of the system at time τ; τ nadir is the time when the frequency reaches the lowest point; ΔP L is the total power disturbance of the system.

[0144] The swing equation based on the center of inertia is:

[0145]

[0146] In the formula, H sys is the total inertia of the system; Δf(τ) is the frequency deviation at time τ.

[0147] Substitute Equation (41) into Equation (42) to get:

[0148]

[0149] Integrating Equation (43) gives the time-domain expression of Δf(τ) for an approximate parabola:

[0150]

[0151] τ in Equation (44) is an unknown variable, and τ nadir is an unknown constant. Therefore, to obtain the lowest frequency point, τ nadir needs to be obtained first.

[0152] Based on Equation (44), the total primary frequency regulation response power of the system in the frequency domain is:

[0153]

[0154] In the formula, G sys (s) is the sum of the dynamic responses of all units:

[0155]

[0156] In the formula, T g is the response constant of generator g; K g is the droop coefficient of generator g.

[0157] Substitute Equation (46) into Equation (45) to get:

[0158]

[0159] Taking the inverse Laplace transform of Equation (47), P PF,sys (τ) can be obtained. Also, since P PF ,sys (τ) is equal to ΔP L when the frequency reaches the lowest point, thus:

[0160] P PF,sys (τ nadir ) = ΔP L (48)

[0161] Solving Equation (48) can obtain the value of τ nadir , and thus the section s nadir where τ nadir is located can be obtained.

[0162] In this embodiment, Benders decomposition is used to decompose the multi - area frequency - security - constrained unit commitment model (26)-(40) into a UC master problem and a frequency verification sub - problem, and the master - sub problems are solved iteratively.

[0163] UC master problem:

[0164] The master problem is to solve the UC to obtain the operation mode of the entire system without considering the multi - area frequency - security constraints. During iteration, the cut - plane fed back by the sub - problem needs to be supplemented in the master - problem constraints, and its model is as follows:

[0165]

[0166] After solving the master problem, the unit start - stop result U gt and the primary frequency regulation reserve power R gt are obtained. They determine the feasibility of the sub - problem, are variables passed to the sub - problem, and are regarded as known parameters during the solution process of the sub - problem.

[0167] The present invention calls a frequency sub - problem mode. Among them, are respectively the column vectors composed of the optimal solutions.

[0168] Frequency verification sub - problem:

[0169] The sub - problem checks whether the system operation mode obtained in the master problem satisfies the multi - area frequency - security constraints at each time period t. Non - negative slack variables St are introduced to ensure that the sub - problem has a solution. Its model is:

[0170]

[0171] Wherein, W is a matrix composed of coefficients on the left side of the inequality signs in constraints (22) and (23); F max is a column vector composed of constants on the right side of the inequality signs in constraints (22) and (23).

[0172] If the target value is greater than 0, it indicates that the solution obtained from the master problem will cause the frequency index to exceed the limit during the t period, and the corresponding infeasible cutting plane needs to be returned to the master problem:

[0173]

[0174] Wherein, is the optimal value of the sub-problem; and are respectively the optimal solutions of the dual variables of constraints and respectively.

[0175] In a possible embodiment, an example of the unit commitment scheduling with improved frequency dynamic constraints for a two-area 12-node system is provided as follows:

[0176] As Figure 2 shown, there are 12 buses, 13 branches and 1 tie line in this system. The time range L of the frequency response is set to 30S, the length l s of each segment is set to 2S, the anticipated disturbance amount is set to 10% of the load demand. The load damping rate is set to 1% / Hz, and the frequency index limits RoCoF max , Δf max and Δf ss,max are respectively set to 0.5Hz / s, 0.5Hz, 0.3Hz, and the governor dead zone Δf i DB is set to 0.02Hz. The primary frequency regulation reserve cost is set to 15$ / MWh, and the curtailment / load shedding costs c W and c B are respectively set to 50$ / MWh and 100$ / MWh. The total system load peak is 540MW, and the load in Area 1 is twice that in Area 2. The installed capacity of the wind turbines is 350MW, all located in Area 2.

[0177] The multi-area frequency response model is approximated using Bernstein polynomials to obtain an algebraic expression for the multi-area frequency dynamics, and a multi-area frequency security constraint is constructed, considering the frequency change rate, the lowest frequency point, and the quasi-steady state frequency deviation constraint;

[0178] Embed multi - area frequency security constraints into the unit commitment problem to obtain a multi - area frequency security - constrained unit commitment model;

[0179] Decompose this large - scale mixed - integer programming problem into a unit commitment master problem and a frequency verification sub - problem through an adaptive two - layer decomposition algorithm. Using the Gurobi commercial solver to solve the model can obtain a unit scheduling plan that ensures multi - area frequency security. The economic cost results of the system optimal scheduling plan as shown in Table 1 are obtained.

[0180]

[0181] Table 1 Economic costs of the system optimal scheduling plan

[0182] Taking the results of Region 2 as an example, Figure 3 This is the comparison result between the multi - area frequency dynamics based on Bernstein polynomial approximation of the present invention and the Simulink numerical simulation. Compare the multi - area frequency - constrained unit commitment model of the present invention with the traditional frequency - constrained unit commitment model based on the center of inertia, Figure 4 is the deviation result of the lowest frequency point of the two, Figure 5 is the result of the initial frequency change rate of the two.

[0183] In summary, based on the dynamic primary frequency regulation response of a synchronous generator, the dead zone of a governor, and the optimal allocation of frequency regulation reserve, the present invention constructs a multi - area frequency response model; approximates the integral - differential algebraic equations of the multi - area frequency response model through Bernstein polynomials to obtain an algebraic expression of multi - area frequency dynamics; constructs multi - area frequency security constraints according to the algebraic expression of multi - area frequency dynamics; embeds the multi - area frequency security constraints into the unit commitment problem to construct a multi - area frequency security - constrained unit commitment model; decomposes the mixed - integer programming problem of the multi - area frequency security - constrained unit commitment model into a master problem and a sub - problem through an adaptive two - layer decomposition algorithm; and obtains a multi - area frequency - secure unit scheduling scheme by solving the master problem and the sub - problem. The present invention simultaneously considers the differentiated dynamic frequency response of units, the action dead zone, and the optimal allocation of frequency regulation reserve in the multi - area frequency response model, and uses Bernstein polynomials to model the integral - differential algebraic equations in the time domain. On this basis, multi - area frequency security constraints based on Bernstein polynomial approximation are obtained. In addition, an adaptive two - layer decomposition algorithm is proposed to decompose the original problem into a master problem and a sub - problem, and at the same time, the problem scale is adaptively reduced by estimating the moment of the lowest frequency point in the sub - problem, greatly reducing the dimension and time of problem optimization. The multi - area frequency response model dynamics based on Bernstein polynomial approximation of the present invention is tight and can more accurately simulate multi - area frequency dynamics compared with the prior art. Compared with the traditional single - area model based on the center of inertia, the frequency index values of the proposed method can more accurately reflect the actual situation of regional frequency response differences.

[0184] It should be noted that the method of the embodiments of the present disclosure can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In such a distributed scenario, one of the multiple devices can only execute one or more steps of the method of the embodiments of the present disclosure, and these multiple devices will interact with each other to complete the described method.

[0185] It should be noted that some embodiments of the present disclosure have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the above - described embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the specific order or continuous order shown to achieve the desired result. In certain embodiments, multi - tasking and parallel processing are also possible or may be advantageous.

[0186] Embodiment 2

[0187] See Figure 6, Embodiment 2 of the present invention further provides a multi - area frequency - secure unit scheduling device based on Bernstein polynomials, including:

[0188] A multi - area frequency response model construction module 001, configured to construct a multi - area frequency response model based on the dynamic primary frequency regulation response of synchronous generators, the governor dead zone, and the optimal allocation of frequency regulation reserve;

[0189] A multi - area frequency - dynamic algebraic expression acquisition module 002, configured to approximate the integral - differential algebraic equation of the multi - area frequency response model through Bernstein polynomials to obtain an algebraic expression of multi - area frequency dynamics;

[0190] A multi - area frequency - security constraint construction module 003, configured to construct multi - area frequency - security constraints according to the algebraic expression of the multi - area frequency dynamics;

[0191] A multi - area frequency - security - constraint unit commitment model construction module 004, configured to embed the multi - area frequency - security constraints into the unit commitment problem to construct a multi - area frequency - security - constraint unit commitment model;

[0192] A multi - area frequency - security - constraint unit commitment model solving module 005, configured to decompose the mixed - integer programming problem of the multi - area frequency - security - constraint unit commitment model into a master problem and a sub - problem through an adaptive two - layer decomposition algorithm; and obtain a multi - area frequency - secure unit scheduling scheme by solving the master problem and the sub - problem.

[0193] In this embodiment, in the multi - area frequency response model construction module 001, the expression of the multi - area frequency response model is:

[0194] When the frequency deviation is less than the governor dead zone:

[0195]

[0196] When the frequency deviation is greater than the governor dead zone:

[0197]

[0198] In the formula, H i is the total inertia of area i; Δf i (τ) is the frequency deviation at time τ in area i; is the load damping rate of area i; is the load power of area i; is the total primary frequency regulation response power at time τ in area i; is the disturbance power of area i; Ψ i is the set of areas connected to area i; X ijis the reactance of the tie line between region i and region j; η is the integration variable; Δf i DB is the action dead zone value of region i; is the critical point time of the governor dead zone; is the set of synchronous generators in region i; is the primary frequency regulation power of generator g at time τ; T g is the response constant of generator g; K g is the droop coefficient of generator g; U g is a 0-1 variable indicating whether unit g is online, and is 1 if so.

[0199] In this embodiment, in the multi-region frequency dynamic algebraic expression acquisition module 002, the algebraic expression of the multi-region frequency dynamic is:

[0200]

[0201] In the formula, l s divides the time range L of the frequency dynamic response into multiple segments, and each segment has a length; both i and j are regions; is the column vector composed of; is the Bernstein spline coefficient corresponding to the time-domain frequency; B m,k (τ) is the m-th order Bernstein polynomial; is the initial value of; J T is the coefficient matrix of the integral property of the Bernstein polynomial; is the column vector of Bernstein spline coefficients corresponding to the total frequency regulation power in the time domain; ΔP i L,B is the column vector of Bernstein spline coefficients corresponding to the disturbance power; is the column vector of Bernstein spline coefficients corresponding to the frequency regulation power of unit g in the time domain; is the initial value of; Δf i DB,B is the dead zone value Δf i DB the vector composed of; m is the order of the Bernstein polynomial; k = 0, 1…, m.

[0202] In this embodiment, in the multi-region frequency safety constraint construction module 003, the multi-region frequency safety constraints include: the initial moment frequency change rate constraint, the frequency lowest point deviation constraint, the quasi-steady state frequency deviation constraint, and the primary frequency regulation reserve constraint;

[0203] The expression of the initial moment frequency change rate constraint is:

[0204]

[0205] In the formula, RoCoF max is the limit value of the frequency change rate at the initial moment;

[0206] The expression of the lowest frequency point deviation constraint is:

[0207]

[0208] In the formula, V is the augmented matrix of the inequality equation transformation, which is a constant matrix; Δf max is the lowest frequency point deviation limit value;

[0209] The expression of the quasi-steady state frequency deviation constraint is:

[0210]

[0211] In the formula, Δf ss,max is the quasi-steady state frequency deviation limit value;

[0212] The expression of the primary frequency regulation reserve constraint is:

[0213]

[0214] In the formula, R g is the primary frequency regulation reserve power of unit g.

[0215] In this embodiment, in the multi-area frequency security constraint unit commitment model construction module 004, the objective function expression of the multi-area frequency security constraint unit commitment model is:

[0216]

[0217] In the formula, Ω G , Ω T , Ω W , Ω B are the sets of thermal power units, scheduling periods, wind turbines, and buses respectively; U gt is a 0-1 variable indicating whether unit g is online in period t, 1 means online, and 0 means offline; R gt is the primary frequency regulation reserve power of unit g in period t; are the start-stop cost and primary frequency regulation reserve cost of unit g respectively; is the fuel cost of unit g in period t; c W and c B are the unit wind curtailment cost and load shedding cost respectively; ΔP wt is the wind curtailment power of wind turbine w in period t; is the load shedding power of bus b in period t.

[0218] In this embodiment, in the multi-region frequency security-constrained unit commitment model solving module 005, during the process of decomposing the mixed-integer programming problem of the multi-region frequency security-constrained unit commitment model into the master problem and the sub-problem through the adaptive two-layer decomposition algorithm, the mixed-integer programming problem is decomposed into the master problem and the sub-problem for iterative calculation through the Benders decomposition strategy and the adaptive low-frequency moment estimation model; through the solution of the master problem, the unit start-stop and frequency regulation reserve allocation strategies are obtained; through the solution of the sub-problem, the multi-region frequency security constraints are verified.

[0219] It should be noted that the information interaction, execution process, etc. between the above system modules, since they are based on the same concept as the method embodiment in Embodiment 1 of the present application, bring the same technical effects as the method embodiment of the present application. The specific content can be seen in the description in the method embodiment shown above in the present application, and will not be elaborated here.

[0220] Embodiment 3

[0221] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which program codes of a multi-region frequency security unit scheduling method based on Bernstein polynomials are stored, and the program codes include instructions for executing the multi-region frequency security unit scheduling method based on Bernstein polynomials in Embodiment 1 or any possible implementation manner thereof.

[0222] The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center integrating one or more available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid-state drive (SSD)).

[0223] Embodiment 4

[0224] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0225] The processor and the memory complete communication with each other through a bus; the memory stores program instructions executable by the processor, and the processor can execute the multi-region frequency security unit scheduling method based on Bernstein polynomials in Embodiment 1 or any possible implementation manner thereof by calling the program instructions.

[0226] Specifically, the processor can be implemented by hardware or software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor that is implemented by reading software code stored in a memory. The memory can be integrated in the processor or can exist independently outside the processor.

[0227] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable systems. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.).

[0228] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general computing system. They can be concentrated on a single computing system or distributed on a network composed of multiple computing systems. Optionally, they can be implemented by program codes executable by the computing system. Thus, they can be stored in a storage system and executed by the computing system. And in some cases, the steps shown or described can be executed in a different order than here, or they can be separately made into individual integrated circuit modules, or multiple modules or steps of them can be made into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0229] Although the present invention has been described in detail above with general descriptions and specific embodiments, based on the present invention, some modifications or improvements can be made, which are obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of the present invention claimed.

Claims

1. A multi-regional frequency safety unit dispatching method based on Bernstein polynomials, characterized in that: include: A multi-region frequency response model is constructed based on the dynamic primary frequency regulation response of synchronous generators, the optimal allocation of speed regulator dead zone and frequency regulation reserve; Approximating the integro-differential algebraic equation of the multi-region frequency response model by Bernstein polynomials to obtain an algebraic expression of the multi-region frequency dynamics; constructing a multi-region frequency security constraint based on the algebraic expression of the multi-region frequency dynamics; The multi-regional frequency safety constraint is embedded into the unit commitment problem to construct a multi-regional frequency safety constraint unit commitment model; The mixed integer programming problem of the multi-regional frequency safety constraint unit combination model is decomposed into a main problem and sub-problems through an adaptive two-layer decomposition algorithm; and a multi-regional frequency safety unit scheduling plan is obtained by solving the main problem and the sub-problems.

2. The multi-region frequency safety unit scheduling method based on Bernstein polynomials according to claim 1 is characterized in that: The expression of the multi-region frequency response model is: When the frequency deviation is less than the dead zone of the speed regulator: When the frequency deviation is greater than the dead zone of the speed regulator: In the formula, H i is the total inertia of region i; Δf i (τ) is the frequency deviation at time τ in region i; is the load damping rate of region i; is the load power of area i; is the total primary FM response power in region i at time τ; is the disturbance power of region i; i is the set of regions connected to region i; X ij is the reactance of the tie line between area i and area j; η is the integration variable; is the action dead zone value of area i; It is the critical point time of the speed regulator dead zone; is the set of synchronous generators in region i; is the primary frequency modulation power of generator g at time τ; T g is the response constant of the generator g; K g is the droop coefficient of the generator g; U g It is a 0-1 variable, indicating whether unit g is online, if so, it is 1.

3. The multi-regional frequency safety unit scheduling method based on Bernstein polynomials according to claim 2 is characterized in that: The algebraic expression of the multi-region frequency dynamics is: In the formula, l s To divide the time range L of the frequency dynamic response into multiple segments, the length of each segment; i and j are both regions; for The column vector composed of is the Bernstein spline coefficient corresponding to the time domain frequency; B m,k (τ) is the m-th degree Bernstein polynomial; for The initial value of J T is the coefficient matrix of the integral property of Bernstein polynomial; is the column vector of Bernstein spline coefficients corresponding to the total FM power in the time domain; is the column vector of Bernstein spline coefficients corresponding to the disturbance power; is the column vector of Bernstein spline coefficients corresponding to the frequency modulation power of the time domain unit g; yes The initial value of Dead zone value A vector composed of; m is the order of Bernstein polynomial; k=0,1…,m.

4. The multi-regional frequency safety unit scheduling method based on Bernstein polynomials according to claim 3 is characterized in that: The multi-region frequency safety constraints include: initial frequency change rate constraints, frequency minimum point deviation constraints, quasi-steady-state frequency deviation constraints and primary frequency regulation standby constraints; The expression of the frequency change rate constraint at the initial moment is: Where, RoCoF max is the limit of the frequency change rate at the initial moment; The expression of the frequency minimum point deviation constraint is: Where V is the enhanced matrix of the inequality equation transformation, which is a constant matrix; Δf max is the frequency minimum point deviation limit; The expression of the quasi-steady-state frequency deviation constraint is: Where Δf ss,max It is the quasi-steady-state frequency deviation limit; The expression of the primary frequency regulation reserve constraint is: In the formula, R g is the frequency regulation standby power of unit g.

5. The multi-regional frequency safety unit scheduling method based on Bernstein polynomials according to claim 4 is characterized in that: The objective function expression of the multi-region frequency safety constraint unit commitment model is: In the formula, Ω G ,Ω T ,Ω W ,Ω B are the collection of thermal power units, dispatching periods, wind power units, and buses; U gt is a 0-1 variable, indicating whether the unit g is online during the period t, 1 for online and 0 for offline; R gt is the primary frequency regulation reserve power of unit g in period t; They are the start-up and shutdown cost and primary frequency regulation standby cost of unit g respectively; is the fuel cost of unit g during period t; c W and c B are the unit wind curtailment cost and load shedding cost respectively; ΔP wt is the abandoned wind power of wind turbine w in period t; is the load shedding power of bus b during period t.

6. The multi-region frequency safety unit scheduling method based on Bernstein polynomials according to claim 5 is characterized in that: In the process of decomposing the mixed integer programming problem of the multi-region frequency safety constraint unit commitment model into the main problem and the sub-problems by using the adaptive two-layer decomposition algorithm, the mixed integer programming problem is decomposed into the main problem and the sub-problems by using the Benders decomposition strategy and the adaptive frequency low point moment estimation model to perform iterative calculations; By solving the main problem, the unit start-stop and frequency regulation standby allocation strategy is obtained; By solving the sub-problems, the multi-region frequency safety constraints are verified.

7. A multi-regional frequency safety unit dispatching device based on Bernstein polynomials, adopting a multi-regional frequency safety unit dispatching method based on Bernstein polynomials as claimed in any one of claims 1 to 6, characterized in that: include: A multi-region frequency response model building module is used to build a multi-region frequency response model based on the dynamic primary frequency regulation response of synchronous generators, the governor dead zone and the optimal allocation of frequency regulation reserves; A module for obtaining an algebraic expression of multi-region frequency dynamics, used for approximating the integral-differential algebraic equation of the multi-region frequency response model by Bernstein polynomials to obtain an algebraic expression of the multi-region frequency dynamics; A multi-region frequency safety constraint construction module, used to construct a multi-region frequency safety constraint according to an algebraic expression of the multi-region frequency dynamics; A multi-regional frequency safety constraint unit combination model construction module is used to embed the multi-regional frequency safety constraint into the unit combination problem and construct a multi-regional frequency safety constraint unit combination model; The multi-regional frequency safety constraint unit combination model solving module is used to decompose the mixed integer programming problem of the multi-regional frequency safety constraint unit combination model into a main problem and sub-problems through an adaptive two-layer decomposition algorithm; and obtain a multi-regional frequency safety unit scheduling plan by solving the main problem and the sub-problems.

8. The multi-region frequency safety unit dispatching device based on Bernstein polynomials according to claim 7 is characterized in that: In the multi-region frequency response model construction module, the expression of the multi-region frequency response model is: When the frequency deviation is less than the dead zone of the speed regulator: When the frequency deviation is greater than the dead zone of the speed regulator: In the formula, H i is the total inertia of region i; Δf i (τ) is the frequency deviation at time τ in region i; is the load damping rate of region i; is the load power of area i; is the total primary FM response power in region i at time τ; is the disturbance power of region i; i is the set of regions connected to region i; X ij is the reactance of the tie line between area i and area j; η is the integration variable; is the action dead zone value of area i; It is the critical point time of the speed regulator dead zone; is the set of synchronous generators in region i; is the primary frequency modulation power of generator g at time τ; T g is the response constant of the generator g; K g is the droop coefficient of the generator g; U g It is a 0-1 variable, indicating whether unit g is online, if so, it is 1.

9. The multi-region frequency safety unit dispatching device based on Bernstein polynomials according to claim 8, characterized in that: In the algebraic expression acquisition module of the multi-region frequency dynamics, the algebraic expression of the multi-region frequency dynamics is: In the formula, l s To divide the time range L of the frequency dynamic response into multiple segments, the length of each segment; i and j are both regions; for The column vector composed of is the Bernstein spline coefficient corresponding to the time domain frequency; B m,k (τ) is the m-th degree Bernstein polynomial; for The initial value of J T is the coefficient matrix of the integral property of Bernstein polynomial; is the column vector of Bernstein spline coefficients corresponding to the total FM power in the time domain; is the column vector of Bernstein spline coefficients corresponding to the disturbance power; is the column vector of Bernstein spline coefficients corresponding to the frequency modulation power of the time domain unit g; yes The initial value of Dead zone value A vector composed of; m is the order of Bernstein polynomial; k=0,1…,m.

10. The multi-region frequency safety unit dispatching device based on Bernstein polynomials according to claim 9, characterized in that: In the multi-region frequency safety constraint construction module, the multi-region frequency safety constraint includes: initial moment frequency change rate constraint, frequency minimum point deviation constraint, quasi-steady-state frequency deviation constraint and primary frequency modulation standby constraint; The expression of the frequency change rate constraint at the initial moment is: Where, RoCoF max is the limit of the frequency change rate at the initial moment; The expression of the frequency minimum point deviation constraint is: Where V is the enhanced matrix of the inequality equation transformation, which is a constant matrix; Δf max is the frequency minimum point deviation limit; The expression of the quasi-steady-state frequency deviation constraint is: Where Δf ss,max It is the quasi-steady-state frequency deviation limit; The expression of the primary frequency regulation reserve constraint is: In the formula, R g is the frequency regulation standby power of unit g.

Images