Two-stage optimization scheduling method for wind power high-permeability system based on frequency safety and peak regulation requirements

Abstract

The invention provides a two-stage optimization scheduling method for a wind power high-permeability system based on frequency safety and peak regulation requirements. The method comprises the following steps: obtaining a primary frequency modulation response model through primary frequency modulation response power of a thermal power generating unit, a wind power plant and an energy storage power station; constructing a first-order swing equation through the primary frequency modulation response model, and solving the first-order swing equation to obtain a system frequency safety index; respectively constructing constraint models of the thermal power generating unit, the wind power plant and the energy storage power station, and obtaining a frequency safety constraint relation through the constraint models and the system frequency safety indexes; constructing a first-stage scheduling model through the hybrid similarity of the new energy output and the load; constructing a second-stage scheduling model through the frequency security constraint relationship; and respectively solving the first-stage scheduling model and the second-stage scheduling model through a solver to obtain an optimal scheduling result. According to the method, mathematical models of two stages are constructed and solved to obtain an optimal operation state capable of meeting system safety and requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power system dispatching, and in particular to a two-stage optimization dispatching method for a wind power high penetration system based on frequency safety and peak load regulation requirements. Background Art

[0002] As the scale of new energy grid connection increases, the power system gradually shows a trend of increasing proportion of new energy and decreasing stock of traditional synchronous machines. The system inertia level and frequency anti-disturbance capability continue to decline, facing severe frequency safety challenges. Large-scale wind power access also makes power balance more difficult and increases the peak regulation pressure of the system.

[0003] The high penetration wind power system is difficult to deal with the frequency security problem by relying solely on the reserved frequency regulation reserve capacity of thermal power units. At the same time, large-scale wind power grid connection reduces the traditional synchronous inertia of the system and increases the power generation disturbance of the system. Therefore, it is necessary to design a two-stage optimization scheduling method for high penetration wind power system based on frequency security and peak regulation needs. Summary of the invention

[0004] The purpose of the present invention is to provide a two-stage optimization scheduling method for a wind power high penetration system based on frequency safety and peak load demand, so as to obtain the optimal operating state that can meet system safety and demand by constructing a two-stage mathematical model and solving it.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A two-stage optimization dispatching method for a wind power high penetration system based on frequency security and peak load demand comprises the following steps:

[0007] The primary frequency regulation response model is obtained through the primary frequency regulation response power of thermal power units, wind farms and energy storage power stations;

[0008] The first-order swing equation is constructed through a frequency modulation response model, and the system frequency safety index is obtained by solving the first-order swing equation; the system frequency safety index includes: initial frequency change rate, maximum frequency deviation value and quasi-steady-state frequency deviation;

[0009] Constraint models of thermal power units, wind farms and energy storage power stations are constructed respectively, and frequency safety constraint relationships are obtained through constraint models and system frequency safety indicators; frequency safety constraint relationships include: frequency change rate constraint relationship, frequency minimum point deviation value constraint relationship and quasi-steady-state frequency deviation constraint relationship;

[0010] The first-stage dispatch model is constructed through the mixed similarity of renewable energy output and load;

[0011] The second-stage scheduling model is constructed through the frequency safety constraint relationship; the second-stage scheduling model includes the second-stage objective function and constraint conditions;

[0012] The first-stage scheduling model and the second-stage scheduling model are solved by the solver to obtain the optimal scheduling result.

[0013] Optionally, the expression of the primary frequency modulation response model is: Among them, P pfr is the total power of the primary frequency response, is the primary frequency regulation response power of the energy storage power station, is the primary frequency regulation response power of thermal power units and wind farms. is the primary frequency modulation response power of thermal power units, wind farms and energy storage power stations, τ is the primary frequency modulation time state quantity, t db is the dead time, t nad is the lowest frequency point time, R is the frequency modulation rate, T d is the duration of a frequency response.

[0014] Alternatively, the first-order swing equation is: Where M is the total inertia of the system, f0 is the rated frequency of the system, Δf(τ) is the frequency deviation, D is the load damping coefficient, P load is the load level when the disturbance occurs, is the system disturbance power, P pfr is the total power of the primary frequency response, τ is the time state quantity of the primary frequency modulation;

[0015] The formula for calculating the initial frequency change rate is: Where RoCoF is the rate of change of system frequency, H is the total inertia of the system;

[0016] The maximum frequency deviation value is calculated as: Where Δf nadir is the frequency deviation value at the lowest frequency point, Δf db is the frequency deviation value of the FM dead zone, R is the primary FM rate;

[0017] The calculation formula for the quasi-steady-state frequency deviation is: Where Δf ∞ It is the frequency deviation value when the system frequency returns to the quasi-steady state after a frequency modulation is completed.

[0018] Optionally, the constraint model expression of the thermal power unit is: Among them, r g,max It is the limit value of the speed regulator of the thermal power unit participating in the primary frequency regulation rate; is the primary frequency regulation reserved power of the thermal power unit at time t, T dis the duration of a frequency response, u g,t is the operating state variable of the thermal power unit at time t, r g,t is the primary frequency regulation rate of the thermal power unit speed regulator at time t;

[0019] The constraint model expression of the wind farm is: Among them, P w,t is the actual output value of the wind farm at time t, is the maximum available output value of the online fan, l w,t is the wind turbine load reduction rate of the wind farm at time t, l w,max is the maximum adjustable load reduction rate of the wind farm, Reserve power for primary frequency regulation of the wind farm at time t;

[0020] The constraint model expression of the energy storage power station is: Among them, P e,t , and are the total output, discharge power and charging power of the energy storage power station participating in peak regulation at time t, and They are the reserved power for frequency regulation in the discharge state, the reserved power for frequency regulation in the charging state and the reserved power for primary frequency regulation of the energy storage power station at time t. and are the maximum charging power and maximum discharging power of the energy storage power station respectively. and u e,t They are respectively the charging state variable, discharging state variable and grid-connected state variable of the energy storage power station at time t.

[0021] Optionally, the frequency change rate constraint relationship is expressed as: Among them, M t is the total inertia of the system at time t, is the maximum disturbance power in period t, RoCoF max is the system frequency change rate limit, f0 is the system rated frequency;

[0022] The expression of the constraint relationship of the frequency minimum point deviation value is: |Δf nadir,t ≤Δf max ; where |Δf nadir,t | is the maximum frequency deviation value at time t, Δf max is the system frequency safety deviation limit;

[0023] The expression of the quasi-steady-state frequency deviation constraint is: in, is the quasi-steady-state frequency deviation limit, D is the load damping coefficient, P t pfris the primary frequency regulation reserved power of the wind farm at time t, P t load is the load level at which the disturbance occurs at time t.

[0024] Optionally, the calculation formula of the mixed similarity is: s mix =γ+β -1 ; Among them, s mix is the mixed similarity, γ is the tracking coefficient, and β is the volatility coefficient;

[0025] The calculation formula of the tracking coefficient is: Among them, P t L and are the grid load at time t and the average load in the dispatching period, and are the predicted output value of the wind farm and the average predicted output value within the dispatch period, T is the total number of time periods, and W is the number of wind farms;

[0026] The calculation formula for the coefficient of volatility is: in, and are the net load of the power grid at time t and the average net load during the dispatching period, respectively.

[0027] Optionally, the objective function expression of the first stage scheduling model is:

[0028] Among them, F1 is the mixed similarity coefficient of the objective function of the first-stage scheduling model, maxF1 is the maximum of F1, and st. is the constraint condition of the first-stage scheduling model. and They are the load forecast power, adjustable load response power and wind power forecast power time series within the dispatch period, and are the upper and lower boundaries of the adjustable load sequence, and They are respectively the load forecast power, adjustable load response power and wind power forecast power in period t.

[0029] Optionally, the expression of the second stage objective function is: minF2 = f G +f W +f E +f AR +f cw ; Among them, f G 、f W 、f E 、f AR and f cware the operating costs of thermal power units, wind farms, energy storage power stations and adjustable loads, respectively. cw Penalty costs for wind curtailment.

[0030] Optionally, the constraint conditions include system power balance constraint, power flow constraint and wind-fire energy storage operation constraint; the expression of the system power balance constraint is: Among them, P b,g,t , P b,w,t , P b,s,t and P b,e,t are the active outputs of the thermal power units, wind farms, energy storage power stations and pumped storage power stations connected to node b, is the load value of node b, is the active power flow of the transmission line l between node b and node b', is the set of transmission lines connected to node b;

[0031] The expression of the power flow constraint is: Among them, B bb' is the admittance between node b and node b', θ b,t and θ b',t are the phase angles of node b and node b' during period t, is the value of the transmission line between b and b', and are the upper and lower limits of active power of the transmission line between node b and node b' respectively;

[0032] Wind-thermal energy storage operation constraints include: thermal power unit operation constraints, wind farm operation constraints, battery energy storage capacity constraints and pumped storage operation constraints.

[0033] Optionally, the first-stage scheduling model and the second-stage scheduling model are solved by a solver to obtain an optimal scheduling result. The specific steps are: solving the first-stage scheduling model by an adaptive particle swarm algorithm to obtain an optimal adjustable load sequence; linearizing the second-stage scheduling model to obtain a mixed integer linear programming model; solving the optimal adjustable load sequence by the mixed integer linear programming model to obtain an optimal scheduling result; the expression of the mixed integer linear programming model is: Among them, y g,t ,y e,t and s,t are auxiliary variables, M is a constant, h g and h s are the inertia time constants of thermal power units and pumped storage units, and are the virtual inertia coefficients of the wind turbine and battery energy storage, S g , S w and S sare the capacity of thermal power units, wind farm installed capacity and pumped storage power station installed capacity, u g,t 、u e,t and u s,t are the operating state variables of thermal power units, energy storage power stations and pumped storage power stations at time t, R t is the system frequency modulation rate at time t.

[0034] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: the two-stage optimization scheduling method for a wind power high penetration system based on frequency safety and peak-shaving demand provided by the present invention comprises: obtaining a primary frequency regulation response model through the primary frequency regulation response power of thermal power units, wind farms and energy storage power stations; constructing a first-order swing equation through the primary frequency regulation response model, and obtaining a system frequency safety index by solving the first-order swing equation; respectively constructing constraint models of thermal power units, wind farms and energy storage power stations, and obtaining a frequency safety constraint relationship through the constraint model and the system frequency safety index; constructing a first-stage scheduling model through the mixed similarity of new energy output and load; constructing a second-stage scheduling model through the frequency safety constraint relationship; solving the first-stage scheduling model and the second-stage scheduling model through a solver to obtain the optimal scheduling result. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0036] Figure 1 This is a flow chart of a two-stage optimization scheduling method for a wind power high penetration system according to an embodiment of the present invention;

[0037] Figure 2 A load prediction curve diagram of an embodiment of the present invention;

[0038] Figure 3 A wind farm prediction output diagram according to an embodiment of the present invention;

[0039] Figure 4 The minimum inertia requirement of the system in each period and the actual inertia result diagram of each scheme in the embodiment of the present invention;

[0040] Figure 5 A schematic diagram of a system frequency change rate according to an embodiment of the present invention;

[0041] Figure 6 A schematic diagram of the maximum frequency deviation value of an embodiment of the present invention;

[0042] Figure 7This is a unit combination diagram of Scheme 1 of an embodiment of the present invention;

[0043] Figure 8 This is a unit combination diagram of Scheme 2 of an embodiment of the present invention;

[0044] Fig. 9 This is a unit combination diagram of Scheme 3 of an embodiment of the present invention;

[0045] Fig.10 A load percentage value result diagram of each solution of an embodiment of the present invention;

[0046] Fig.11 Schematic diagram of the relationship between system disturbance power, inertia and primary frequency modulation rate requirement according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] like Figure 1 As shown, the embodiment of the present invention provides a two-stage optimization scheduling method for a wind power high penetration system based on frequency security and peak load demand, comprising the following steps:

[0050] Step 100: Obtain a primary frequency regulation response model through the primary frequency regulation response power of the thermal power unit, the wind farm and the energy storage power station;

[0051] Specifically, the primary frequency response resources of the system are composed of conventional thermal power units, fans and energy storage. In this embodiment, the response power of each resource during the primary frequency response process increases linearly. Among them, the fan in the load reduction state has a speed regulator climbing response characteristic similar to that of the conventional thermal power unit. When the energy storage participates in the primary frequency response, it can complete the release of all regulated power within a few milliseconds, and its response characteristic is similar to a step response. The conventional thermal power unit and the fan are used as frequency regulation resources, and the energy storage is used as a fast frequency regulation resource, so that they are respectively at t nad and t db The moment is fully responded, and the respective primary frequency modulation response power is obtained. The expressions of different primary frequency modulation response powers are:

[0052]

[0053]

[0054] Among them, t db ,t nad ,t ∞ They are the dead time, the lowest frequency point time and the time to reach the quasi-steady state, τ is the time state quantity of the primary frequency modulation; T d It is defined as the duration of a frequency response, and its calculation formula is: T d =t nad -t db ;u g 、r g are the operating state variables of thermal power unit g and the primary frequency regulation rate of the speed regulator, u g The value range is 0-1; are the response powers of thermal power units, wind farms and energy storage power stations participating in primary frequency regulation in the system, G, W and E are the numbers of thermal power units, wind farms and energy storage power stations respectively; They are the reserved powers of wind farm w and energy storage power station e participating in primary frequency regulation respectively.

[0055] More specifically, the participation of wind farms and energy storage power stations in frequency regulation can improve the regulation rate of the primary frequency response of the system. The primary frequency regulation response model of the system is obtained by the response power of thermal power units, wind farms and energy storage power stations participating in primary frequency regulation. The expression of the primary frequency regulation response model is:

[0056]

[0057] Among them, P pfr is the total power of the primary frequency response, is the primary frequency regulation response power of the energy storage power station, is the primary frequency regulation response power of thermal power units and wind farms. is the primary frequency modulation response power of thermal power units, wind farms and energy storage power stations, τ is the primary frequency modulation time state quantity, t db is the dead time, t nad is the time of the lowest frequency point, R is the primary frequency modulation rate, and “+” indicates that different frequency modulation resources jointly participate in the frequency response.

[0058] It should be noted that the frequency response in the system includes inertial response and primary frequency response, and the frequency response plays a key role in ensuring the frequency safety of the system.

[0059] Step 200: construct a first-order swing equation through a primary frequency modulation response model, and obtain a system frequency safety index by solving the first-order swing equation; the system frequency safety index includes: initial frequency change rate, maximum frequency deviation value and quasi-steady-state frequency deviation.

[0060] Specifically, the frequency modulation response model can be used to calculate the change of frequency deviation over time after the system is disturbed, and this change can be expressed in the form of a first-order swing equation. The expression of the first-order swing equation is:

[0061]

[0062] Where M is the total inertia of the system, f0 is the rated frequency of the system, Δf(τ) is the frequency deviation, D is the load damping coefficient, P load is the load level when the disturbance occurs, is the system disturbance power, P pfr is the total power of the primary frequency response, and τ is the time state quantity of the primary frequency modulation.

[0063] More specifically, the initial frequency change rate is only related to the system inertia level and the disturbance power scale when the disturbance occurs, and its calculation formula is:

[0064]

[0065] Where RoCoF is the rate of change of system frequency and H is the total inertia of the system.

[0066] The maximum frequency deviation value is calculated as:

[0067]

[0068] Where Δf nadir is the frequency deviation value at the lowest frequency point, Δf db is the frequency deviation value of the FM dead zone, and R is the primary FM rate.

[0069] After the primary frequency modulation is completed, the system frequency tends to a stable value. The quasi-steady-state frequency deviation is related to the primary frequency modulation power of the system and the load damping effect. The calculation formula for the quasi-steady-state frequency deviation is:

[0070]

[0071] Where Δf ∞ It is the frequency deviation value when the system frequency returns to the quasi-steady state after a frequency modulation is completed.

[0072] Step 300: construct constraint models for thermal power units, wind farms and energy storage power stations respectively, and obtain frequency safety constraint relationships through constraint models and system frequency safety indicators; frequency safety constraint relationships include: frequency change rate constraint relationships, frequency minimum point deviation value constraint relationships and quasi-steady-state frequency deviation constraint relationships;

[0073] Specifically, the constraint model is obtained by analyzing the inertia support and primary frequency regulation dynamic process of the system. The inertia support is provided by thermal power units, pumped storage units, wind farms and energy storage power stations. Thermal power units, wind farms and energy storage power stations participate in the primary frequency regulation of the system, and thermal power units also participate in the secondary frequency regulation of the system. The primary frequency regulation dynamic process is obtained through the maximum disturbance power of each frequency regulation period.

[0074] It should be noted that the power response rate of thermal power units is significantly different during normal peak load regulation and frequency regulation, and thermal power units in rated operation still have primary frequency response capability. The constraint model expression of thermal power units is:

[0075]

[0076] Among them, r g,max It is the limit value of the speed regulator of the thermal power unit participating in the primary frequency regulation rate; is the primary frequency regulation reserved power of the thermal power unit at time t, T d is the duration of a frequency response, u g,t is the operating state variable of the thermal power unit at time t, r g,t It is the primary frequency regulation rate of the speed regulator of the thermal power unit at time t.

[0077] The wind turbines in the wind farm participate in the system frequency regulation in the form of dynamic load reduction. The output power and primary frequency regulation capability of the wind turbines are related to the dynamic load reduction rate. The constraint model expression of the wind farm is:

[0078]

[0079] Among them, P w,t is the actual output value of wind farm w at time t, is the maximum available output value of the online fan; l w,t is the wind turbine load reduction rate of wind farm w at time t, l w,max is the maximum adjustable load reduction rate of the wind farm w, The reserved power is for the primary frequency regulation of wind farm w at time t.

[0080] Energy storage power stations have the characteristics of rapid response and stable charging and discharging. They can participate in the long-term peak regulation and short-term frequency regulation of the system. The impact of energy storage power stations in the primary frequency regulation of the system is related to their charging and discharging state and charging and discharging power. The constraint model expression of energy storage power stations is:

[0081]

[0082] Among them, P e,t , and are the total output, discharge power and charging power of the energy storage power station e participating in peak regulation at time t, and They are the reserved power for frequency regulation in the discharge state, the reserved power for frequency regulation in the charging state and the reserved power for primary frequency regulation of the energy storage station e at time t. and are the maximum charging power and maximum discharging power of the energy storage station e, respectively. and u e,t They are respectively the charging state variable, discharging state variable and grid-connected state variable of the energy storage power station e at time t.

[0083] Furthermore, fire, wind and storage have complementary effects in the system frequency regulation process. The expression of primary frequency regulation power and primary frequency regulation rate in the frequency regulation process is:

[0084]

[0085] Among them, P t pfr is the total power of the system at time t, R t is the system frequency modulation rate at time t.

[0086] Specifically, the frequency change rate is related to the total inertia of the system, which is calculated as:

[0087]

[0088] Among them, h g and h s are the inertia time constants of thermal power units and pumped storage units, and are the virtual inertia coefficients of wind turbines and energy storage power stations in wind farms, respectively. g , S w and S s are the capacity of thermal power unit g, the installed capacity of wind farm w and the installed capacity of pumped storage power station s, respectively. is the rated power of the energy storage station e, u s,t is the online state variable of the unit of pumped storage power station s, u s,t The value range is 0-1.

[0089] The expression of the frequency change rate constraint is:

[0090]

[0091] Among them, M t is the total inertia of the system at time t, is the maximum disturbance power in period t, RoCoF max is the system frequency change rate limit;

[0092] The expression of the constraint relationship of the frequency minimum point deviation value is:

[0093] |Δf nadir,t ≤Δf max ;

[0094] Where |Δf nadir,t | is the maximum frequency deviation value at time t, Δf max is the system frequency safety deviation limit;

[0095] In order to ensure that the quasi-steady-state frequency deviation does not exceed the limit and prevent over-regulation, the primary frequency response power should be greater than the minimum frequency safety requirement and not exceed the disturbance power value. The expression of the quasi-steady-state frequency deviation constraint is:

[0096]

[0097] in, is the quasi-steady-state frequency deviation limit, D is the load damping coefficient, P t pfr is the primary frequency regulation reserved power of the wind farm at time t, P t load is the load level at which the disturbance occurs at time t.

[0098] It should be noted that after the primary frequency regulation of the system, the system active power is completely balanced by mobilizing the spinning reserve of the thermal power units and adjusting the reference value. The expression of the spinning reserve constraint relationship of the thermal power units that meets the secondary frequency regulation requirements of the system is:

[0099]

[0100] in, and They are the upper and lower standby of thermal power unit g at time t respectively.

[0101] Step 400: constructing a first-stage dispatching model through the hybrid similarity between the output of new energy and the load;

[0102] Specifically, the higher the matching degree between the new energy output curve and the load curve and the smoother the net load, the safer and more economical the power system operation will be. The calculation formula of hybrid similarity is:

[0103] s mix =γ+β -1 ;

[0104] Among them, s mix is the mixed similarity, γ is the tracking coefficient, and β is the volatility coefficient; the calculation formula of the tracking coefficient is:

[0105]

[0106] Among them, P t L and are the grid load at time t and the average load in the dispatching period, and are the predicted output value of the wind farm and the average predicted output value within the dispatch period, T is the total number of time periods, and W is the number of wind farms; the calculation formula of the fluctuation coefficient is:

[0107]

[0108] in, and are the net load of the power grid at time t and the average net load during the dispatching period, respectively.

[0109] It should be noted that the tracking coefficient γ ranges from [-1, 1], which indicates the degree of fit between the output of new energy and the load. A positive value indicates that the change trends of the two are the same. The larger the γ, the higher the fit between the new energy and the load. The fluctuation coefficient β ranges from [0, +∞], which indicates the smoothness of the net load curve of the system. The larger the β, the higher the volatility of the net load curve.

[0110] Specifically, the objective function of the first-stage optimization model is constructed by the maximum value of the new energy-load hybrid similarity to optimize the adjustable load power. The expression of the objective function is:

[0111]

[0112] Among them, F1 is the mixed similarity coefficient of the objective function of the first-stage scheduling model, maxF1 is the maximum of F1, and st. indicates that the following formula is the constraint condition of the first-stage scheduling model. and They are the load forecast power, adjustable load response power and wind power forecast power time series within the dispatch period, and are the upper and lower boundaries of the adjustable load sequence, and They are respectively the load forecast power, adjustable load response power and wind power forecast power in period t. The load forecast curve is as follows: Figure 2 shown.

[0113] Step 500: constructing a second-stage scheduling model through a frequency safety constraint relationship; the second-stage scheduling model includes a second-stage objective function and constraint conditions;

[0114] Specifically, the expression of the second stage objective function is:

[0115] minF2=f G +fW +f E +f AR +f cw ;

[0116] Among them, f G 、f W 、f E 、f AR and f cw are the operating costs of thermal power units, wind farms, energy storage power stations and adjustable loads, respectively. cw Penalty costs for wind curtailment.

[0117] It should be noted that the operating cost of thermal power is G The cost of coal-fired power generation is f G1 , start-stop cost G2 , standby cost f G3 and the primary frequency regulation reserve cost f G4 The respective expressions are as follows:

[0118]

[0119] Among them, a g , b g and c g are the cost coefficients of thermal power unit g, s g ,d g and e g are the startup, shutdown and standby cost coefficients of thermal power unit g, P g,t , and are the output variable, startup action variable and shutdown action variable of thermal power unit g in period t, is the frequency regulation power reserve cost coefficient of thermal power unit g.

[0120] The operating cost of a wind farm includes operation and maintenance cost, primary frequency regulation power reserve cost and wind abandonment cost. The expression of the operating cost of a wind farm is:

[0121]

[0122] Among them, w , and are the operation and maintenance cost coefficient, frequency regulation power reserve coefficient and wind abandonment penalty coefficient of wind farm w, is the abandoned wind power of wind farm w during period t.

[0123] The operating cost of the energy storage power station includes the peak load regulation and frequency regulation costs. The expression of the operating cost of the energy storage power station is:

[0124]

[0125] Among them, e and They are the operation and maintenance cost coefficient and frequency regulation standby cost coefficient of the energy storage power station e respectively.

[0126] The expression of adjustable load operating cost is:

[0127]

[0128] Among them, c AR is the unit power adjustable load operating cost, where P t AR The time series in the first stage scheduling model The value of is a fixed value in the second stage scheduling model.

[0129] Specifically, the constraints include system power balance constraints, power flow constraints, and wind-fire energy storage operation constraints; the expression of the load power that can be adjusted after optimization is:

[0130] P t load =P t L -P t AR ;

[0131] It can be seen that the expression of the system power balance constraint is:

[0132]

[0133] Among them, P b,g,t , P b,w,t , P b,s,t and P b,e,t are the active outputs of the thermal power units, wind farms, energy storage power stations and pumped storage power stations connected to node b, is the load value of node b, is the active power flow of the transmission line l between node b and node b', is the set of transmission lines connected to node b. The expression of power flow constraint is:

[0134] Among them, B bb' is the admittance between node b and node b', θ b,t and θ b',t are the phase angles of node b and node b' during period t, is the value of the transmission line between b and b', and are the upper and lower limits of the active power of the transmission line between node b and node b' respectively; the expression of the phase angle constraint is:

[0135]

[0136] in, and are the lower and upper limits of the phase angle of node b respectively.

[0137] More specifically, the wind and thermal storage operation constraints include: thermal power unit operation constraints, wind farm operation constraints, battery energy storage capacity constraints and pumped storage operation constraints. The operation constraints of thermal power units include unit output constraints, ramp constraints, standby constraints and start-stop constraints. The expression of thermal power unit operation constraints is as follows:

[0138] u g,t P g,min ≤P g,t ≤u g,t P g,max

[0139]

[0140] Among them, P g,min and P g,max are the upper and lower limits of the output of thermal power unit g, and are respectively the up and down ramp regulation rate constraints of thermal power unit g, and are the actual start and stop time of the thermal power unit, and are the minimum start and stop time of unit g respectively.

[0141] The expression of wind farm operation constraints is:

[0142]

[0143] in, is the predicted power of wind farm w at time t.

[0144] The expression of battery energy storage capacity constraint is:

[0145]

[0146] Among them, E e,t is the power of energy storage station e during period t, and are the charging and discharging efficiency coefficients of the energy storage station e, E e,max and E e,min They are the upper and lower limits of the power of the energy storage power station e.

[0147] The expression of pumped storage operation constraint is:

[0148]

[0149] Among them, P s,t is the output of the pumped storage power station s at time t, P s max and P s min are the maximum and minimum output of the pumped storage power station s, is the existing water volume in the upper reservoir of the pumped storage power station s at time t, η s is the hydropower conversion coefficient of pumped storage station s, and are the maximum and minimum storage capacities of the upper reservoir of pumped storage power station s, respectively.

[0150] Step 600: Solve the first stage scheduling model and the second stage scheduling model respectively through a solver to obtain an optimal scheduling result.

[0151] Specifically, the first-stage scheduling model is first solved by using an adaptive particle swarm algorithm with strong global optimization capability to obtain the optimal adjustable load sequence, which is used as the input of the second-stage scheduling model; secondly, the second-stage scheduling model is linearized to obtain a mixed integer linear programming model; finally, the optimal adjustable load sequence is solved by the mixed integer linear programming model to obtain the optimal scheduling result.

[0152] More specifically, the specific steps of linearizing the second-stage scheduling model to obtain the mixed integer linear programming model include:

[0153] By Δf nadir The expression and constraints of get the complete expanded form of the maximum frequency deviation:

[0154]

[0155] The equality boundary of the nonlinear inequality constraint is obtained through the second-stage scheduling model; the expression of the equality boundary equation is:

[0156]

[0157] in, is a particular solution of the equation, is H at time t when the equation is established t and R t Special value.

[0158] The full expansion form of the maximum frequency deviation is converted into a general inequality constraint; the expression of the general inequality constraint is:

[0159]

[0160] Among them, H t *Rt It is a typical nonlinear expression of the product of two variables.

[0161] Introduce auxiliary variable y g,t ,y e,t ,y s,t and a sufficiently large constant M, linearize the nonlinear expression of the product of two variables to obtain a mixed integer linear programming model; the expression of the mixed integer linear programming model is:

[0162]

[0163] Among them, h g and h s are the inertia time constants of thermal power units and pumped storage units, and are the virtual inertia coefficients of the wind turbine and battery energy storage, S g , S w and S s are the capacity of thermal power units, wind farm installed capacity and pumped storage power station installed capacity, u g,t 、u e,t and u s,t are the operating state variables of thermal power units, energy storage power stations and pumped storage power stations at time t, R t is the system frequency modulation rate at time t.

[0164] It should be noted that the Yalmip toolbox is used in combination with the commercial optimization software GUROBI to solve each expression, and the operating status and output of each thermal power plant, wind farm output, energy storage power station and pumped storage power station under the typical day with the best economic efficiency that meets all constraints, as well as the actual situation of participating in peak load regulation and frequency regulation, are obtained. Figure 3 shown.

[0165] The embodiment of the present invention provides three solutions. The maximum disturbance power in each time period of the three solutions is 5% of the load value, and the first-stage scheduling model based on the new energy-load hybrid similarity is not considered.

[0166] Scheme 1: The traditional unit combination model that does not consider the primary frequency regulation of the system and only considers the peak regulation cost, and the reserve power is set to 2% of the system load; Scheme 2: Considering the dynamic frequency constraint but not considering the frequency regulation of the wind farm; Scheme 3: Considering the dynamic frequency constraint and considering the wind and fire storage constraint. The minimum inertia demand of the system in each period and the actual inertia results of each scheme are as follows Figure 4 The system frequency change rate and maximum frequency deviation value under the three schemes are shown as follows: Figure 5 and Figure 6 The unit combinations of each scheme are as follows Figure 7-Figure 9 The economic dispatch results of each scheme are shown in Table 1:

[0167] Table 1 Economic dispatch results

[0168]

[0169]

[0170] It should be noted that although Scheme 1 can meet the power supply demand of the system, due to the lack of consideration of dynamic frequency constraints, the system inertia is insufficient during the peak load period of 11:00-12:00, resulting in system frequency change rate values ​​of -0.2514Hz / s and -0.2601Hz / s, respectively, exceeding the system maximum frequency change rate limit of 0.25Hz / s, resulting in wind abandonment or load shedding events. During the period of 11:00-12:00, the start and stop states of the generator sets of Scheme 1 and Scheme 2 are the same, but the overall inertia level of Scheme 2 is slightly higher than the minimum inertia requirement of the system. The overall inertia level of Scheme 3 is significantly improved, and the inertia level in each period is much higher than the minimum inertia requirement of the system.

[0171] Both Scheme 2 and Scheme 3 take into account dynamic frequency constraints, and the maximum frequency deviation value in each time period is less than the frequency deviation limit of the constraint condition of 0.8Hz. The overall system frequency deviation value of Scheme 2 is relatively high, and the maximum frequency deviation reaches 0.7485Hz at the peak load time of 12:00. Compared with Scheme 2, Scheme 3 takes into account the participation of fans in inertia support and primary frequency regulation. The maximum frequency deviation value is relatively stable overall and the mean value is small. It is only greater than Scheme 2 in the 15:00-17:00 period. The maximum frequency deviation values ​​in the remaining time periods are all less than Scheme 2, and the maximum frequency deviation value in each time period is only 0.4726Hz.

[0172] Specifically, the system frequency regulation capability is defined as the percentage of the disturbance power that the system bears to the load power, and the dispatching models of Scheme 2 and Scheme 3 are solved to calculate the system frequency regulation capability and operating cost. The calculation results are as follows: Fig.10 shown.

[0173] It should be noted that the dispatching cost of Scheme 3 is better than that of Scheme 2 as a whole, and the system inertia support capacity is greatly improved after the wind turbine participates in the system frequency regulation. The maximum frequency regulation capacity of the system reaches 8.511%, which is higher than 5.086% of Scheme 2, and has better frequency stability.

[0174] Furthermore, the relationship between the system disturbance power, inertia and primary frequency modulation rate requirement is as follows: Fig.11As shown. The frequency regulation capacity coverage operation area of ​​Scheme 2 and Scheme 3 corresponds to the red and yellow line areas in the figure respectively, and the extreme operation points are the red dots (13.3, 8800, 105.1) and the yellow square dots (12.5, 16000, 186.1) in the figure respectively. In order to meet the maximum frequency deviation constraint of the system, the primary frequency regulation rate requirements of Scheme 2 and Scheme 3 are 13.3MW / s and 12.5MW / s respectively. The maximum primary frequency regulation rate that the thermal power unit can provide is 16.62MW / s, which is greater than the primary frequency regulation power requirements of the two schemes, and the battery energy storage and wind turbine also provide primary frequency regulation power, and the primary frequency regulation power reserve is sufficient. After the two schemes continue to increase the disturbance power on the basis of the extreme disturbance power of 105.1MW and 186.1MW corresponding to the extreme operation point, the dispatching model cannot be solved. At this time, there is still a surplus of primary frequency regulation capacity in the system, indicating that the main factor restricting the upper limit of the system frequency regulation capacity at this time is not the primary frequency regulation capacity but the insufficient system inertia.

[0175] It is understandable that the frequency regulation cost of Scheme 2 increased by 6.517% compared with Scheme 1, and the frequency regulation capacity of Scheme 3 increased by 67.3% compared with Scheme 2, which improved the safety and economy of the system frequency.

[0176] The beneficial effects of the present invention are as follows:

[0177] 1) The first-stage dispatching model was established by combining wind, fire, storage and storage to regulate system frequency;

[0178] 2) The adjustable load is optimized by taking the new energy-load hybrid similarity as the target, thus improving the reliability of system scheduling;

[0179] 3) In the process of constructing the second-stage scheduling model, various constraints were fully considered to improve the stability and accuracy of system scheduling.

[0180] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0181] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand, characterized in that: The steps include: The primary frequency regulation response model is obtained through the primary frequency regulation response power of thermal power units, wind farms and energy storage power stations; Constructing a first-order swing equation through the primary frequency modulation response model, and obtaining a system frequency safety index by solving the first-order swing equation; The system frequency safety indicators include: initial frequency change rate, maximum frequency deviation value and quasi-steady-state frequency deviation; Constraint models of thermal power units, wind farms and energy storage power stations are constructed respectively, and frequency safety constraint relationships are obtained through the constraint models and the system frequency safety indicators; the frequency safety constraint relationships include: frequency change rate constraint relationships, frequency minimum point deviation value constraint relationships and quasi-steady-state frequency deviation constraint relationships; The first-stage dispatch model is constructed through the mixed similarity of renewable energy output and load; The second-stage scheduling model is constructed by the frequency safety constraint relationship; the second-stage scheduling model includes the second-stage objective function and constraint conditions; The first stage scheduling model and the second stage scheduling model are solved by the solver respectively to obtain the optimal scheduling result.

2. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The expression of the primary frequency modulation response model is: Among them, P pfr is the total power of the primary frequency response, is the primary frequency regulation response power of the energy storage power station, is the primary frequency regulation response power of thermal power units and wind farms. is the primary frequency modulation response power of thermal power units, wind farms and energy storage power stations, τ is the primary frequency modulation time state quantity, t db is the dead time, t nad is the lowest frequency point time, R is the frequency modulation rate, T d is the duration of a frequency response.

3. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The first-order swing equation is: Where M is the total inertia of the system, f0 is the rated frequency of the system, Δf(τ) is the frequency deviation, D is the load damping coefficient, P load is the load level when the disturbance occurs, is the system disturbance power, P pfr is the total power of the primary frequency response, τ is the time state quantity of the primary frequency modulation; The calculation formula of the initial frequency change rate is: Where RoCoF is the rate of change of system frequency, H is the total inertia of the system; The calculation formula of the maximum frequency deviation value is: Where Δf nadir is the frequency deviation value at the lowest frequency point, Δf db is the frequency deviation value of the FM dead zone, R is the primary FM rate; The calculation formula of the quasi-steady-state frequency deviation is: Where Δf ∞ It is the frequency deviation value when the system frequency returns to the quasi-steady state after a frequency modulation is completed.

4. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The constraint model expression of the thermal power unit is: Among them, r g,max It is the limit value of the speed regulator of the thermal power unit participating in the primary frequency regulation rate; is the primary frequency regulation reserved power of the thermal power unit at time t, T d is the duration of a frequency response, u g,t is the operating state variable of the thermal power unit at time t, r g,t is the primary frequency regulation rate of the thermal power unit speed regulator at time t; The constraint model expression of the wind farm is: Among them, P w,t is the actual output value of the wind farm at time t, is the maximum available output value of the online fan, l w,t is the wind turbine load reduction rate of the wind farm at time t, l w,max is the maximum adjustable load reduction rate of the wind farm, Reserve power for primary frequency regulation of the wind farm at time t; The constraint model expression of the energy storage power station is: Among them, P e,t , and are the total output, discharge power and charging power of the energy storage power station participating in peak regulation at time t, and They are the reserved power for frequency regulation in the discharge state, the reserved power for frequency regulation in the charging state and the reserved power for primary frequency regulation of the energy storage power station at time t. and are the maximum charging power and maximum discharging power of the energy storage power station respectively. and u e,t They are respectively the charging state variable, discharging state variable and grid-connected state variable of the energy storage power station at time t.

5. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The expression of the frequency change rate constraint relationship is: Among them, M t is the total inertia of the system at time t, is the maximum disturbance power in period t, RoCoF max is the system frequency change rate limit, f0 is the system rated frequency; The expression of the constraint relationship of the frequency minimum point deviation value is: |Δf nadir,t |≤Δf max ; where |Δf nadir,t | is the maximum frequency deviation value at time t, Δf max is the system frequency safety deviation limit; The expression of the quasi-steady-state frequency deviation constraint relationship is: in, is the quasi-steady-state frequency deviation limit, D is the load damping coefficient, P t pfr is the primary frequency regulation reserved power of the wind farm at time t, P t load is the load level at which the disturbance occurs at time t.

6. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The calculation formula of the mixed similarity is: mix =γ+β -1 ; Among them, s mix is the mixed similarity, γ is the tracking coefficient, and β is the volatility coefficient; The calculation formula of the tracking coefficient is: Among them, P t L and are the grid load at time t and the average load in the dispatching period, and are the predicted output value of the wind farm and the average predicted output value within the dispatch period, T is the total number of time periods, and W is the number of wind farms; The calculation formula of the fluctuation coefficient is: in, and are the net load of the power grid at time t and the average net load during the dispatching period, respectively.

7. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The objective function expression of the first stage scheduling model is: Among them, F1 is the mixed similarity coefficient of the objective function of the first-stage scheduling model, maxF1 is the maximum of F1, and st. is the constraint condition of the first-stage scheduling model. and They are the load forecast power, adjustable load response power and wind power forecast power time series within the dispatch period, and are the upper and lower boundaries of the adjustable load sequence, and They are respectively the load forecast power, adjustable load response power and wind power forecast power in period t.

8. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The expression of the objective function of the second stage is: minF2 = f G +f W +f E +f AR +f cw ; Among them, f G 、f W 、f E 、f AR and f cw are the operating costs of thermal power units, wind farms, energy storage power stations and adjustable loads, respectively. cw Penalty costs for wind curtailment.

9. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The constraints include system power balance constraints, power flow constraints and wind-fire energy storage operation constraints; the expression of the system power balance constraint is: Among them, P b,g,t , P b,w,t , P b,s,t and P b,e,t are the active outputs of the thermal power units, wind farms, energy storage power stations and pumped storage power stations connected to node b, is the load value of node b, is the active power flow of the transmission line l between node b and node b', is the set of transmission lines connected to node b; The expression of the power flow constraint is: Among them, B bb' is the admittance between node b and node b', θ b,t and θ b',t are the phase angles of node b and node b' during period t, is the value of the transmission line between b and b', and are the upper and lower limits of active power of the transmission line between node b and node b' respectively; The wind-fire storage operation constraints include: thermal power unit operation constraints, wind farm operation constraints, battery energy storage power constraints and pumped storage operation constraints.

10. The two-stage optimization scheduling method for wind power high penetration system based on frequency security and peak load demand according to claim 1 is characterized in that: The first stage scheduling model and the second stage scheduling model are solved by the solver respectively to obtain the optimal scheduling result. The specific steps are: solving the first stage scheduling model by the adaptive particle swarm algorithm to obtain the optimal adjustable load sequence; linearizing the second stage scheduling model to obtain a mixed integer linear programming model; solving the optimal adjustable load sequence by the mixed integer linear programming model to obtain the optimal scheduling result; the expression of the mixed integer linear programming model is: Among them, y g,t ,y e,t and s,t are auxiliary variables, M is a constant, h g and h s are the inertia time constants of thermal power units and pumped storage units, and are the virtual inertia coefficients of the wind turbine and battery energy storage, S g , S w and S s are the capacity of thermal power units, wind farm installed capacity and pumped storage power station installed capacity, u g,t 、u e,t and u s,t are the operating state variables of thermal power units, energy storage power stations and pumped storage power stations at time t, R t is the system frequency modulation rate at time t.

Images