CN113098030B - Frequency control optimization method for wind power plant participating in primary frequency modulation - Google Patents

Frequency control optimization method for wind power plant participating in primary frequency modulation Download PDF

Info

Publication number
CN113098030B
CN113098030B CN202110376547.1A CN202110376547A CN113098030B CN 113098030 B CN113098030 B CN 113098030B CN 202110376547 A CN202110376547 A CN 202110376547A CN 113098030 B CN113098030 B CN 113098030B
Authority
CN
China
Prior art keywords
frequency
wind farm
value
power
control
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202110376547.1A
Other languages
Chinese (zh)
Other versions
CN113098030A (en
Inventor
王巧
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN202110376547.1A priority Critical patent/CN113098030B/en
Publication of CN113098030A publication Critical patent/CN113098030A/en
Application granted granted Critical
Publication of CN113098030B publication Critical patent/CN113098030B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
    • H02J3/241—The oscillation concerning frequency
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • H02J3/48—Controlling the sharing of the in-phase component
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin
    • H02J2300/28—The renewable source being wind energy
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/76—Power conversion electric or electronic aspects

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Wind Motors (AREA)

Abstract

The invention discloses a frequency control optimization method for wind power plant participation primary frequency modulation, which comprises the following steps: s1, establishing a frequency control model; s2, acquiring the frequency of a power grid; s3, judging whether the current power grid frequency deviates from a set value; s4, obtaining an output control value of the inertia control model; obtaining an output control value of a frequency control model; s5, obtaining an output control value of the load shedding control model; s6, obtaining an output control value of the inertia control model; s7, obtaining the power output by the wind turbine generator to the power grid; s8, inputting the power output by the wind turbine generator to the power grid, the output power of the prime motor and the load power to a generator load model to obtain an adjusted frequency; s9, optimizing target parameters of the current frequency control model through reinforcement learning, and returning to the step S2 to realize continuous optimization. The method can not only rapidly respond to transient state change of system frequency and relieve pressure of the traditional frequency modulation unit, but also realize permanent participation of the wind power plant.

Description

Frequency control optimization method for wind power plant participating in primary frequency modulation
Technical Field
The invention relates to the field of power grid frequency modulation, in particular to a frequency control optimization method for wind power plant participation primary frequency modulation.
Background
In recent years, wind energy is used as a clean novel energy source, and is an important research object for developing and utilizing renewable energy sources due to the characteristics of abundant reserves, low cost, and the like. However, because wind power generation has the characteristics of intermittence, randomness and volatility, a series of impacts are caused on the safe and stable operation of a power grid, particularly the frequency of the power grid is influenced, and the problems are particularly prominent in a system with high permeability of wind power.
Therefore, in order to maintain the frequency stability of the system while effectively utilizing wind energy resources, the wind farm should co-regulate the grid frequency as in conventional units. Although the wind turbine generator system does not have the inertial response and primary frequency modulation capability of the traditional wind turbine generator system, an active control system based on a power electronic converter provides a control basis for wind power participation frequency adjustment. Therefore, the wind turbine generator can realize the adjustment of the frequency of the power grid by adjusting the active output.
At present, a great deal of research is carried out on wind power participation frequency modulation at home and abroad, and the research can be roughly divided into two types, namely inertial control without standby capacity and primary frequency modulation control with standby capacity, including overspeed control and pitch angle control.
However, most frequency modulation methods basically do not consider the nonlinear characteristic of the frequency response of the system, only the control method of the active power system of the wind power plant is researched, and the active output of the wind power generator set is still not ideal when different frequency deviations are dealt with. Therefore, in order to fully utilize wind power resources, it is necessary to further optimize the active power output adjustment scheme in order to better respond to system frequency changes.
Disclosure of Invention
Aiming at the defects in the prior art, the frequency control optimization method for the wind power plant participating in primary frequency modulation solves the problem that the active output of a wind turbine generator is not ideal when the existing wind power plant is used for coping with different frequency deviations.
In order to achieve the aim of the invention, the invention adopts the following technical scheme:
the frequency control optimization method for the wind farm participating in primary frequency modulation comprises the following steps:
s1, establishing a frequency control model, and initializing target parameters of the frequency control model, wherein the target parameters comprise wind farm inertia, wind farm damping and auxiliary control constants of steady-state output change of the wind farm;
s2, acquiring the current power grid frequency;
s3, judging whether the current power grid frequency deviates from a set value, if so, entering a step S4, otherwise, returning to the step S2;
s4, inputting the deviation value of the power grid frequency into an inertia control model to obtain an output control value A1 of the inertia control model; inputting the deviation value of the power grid frequency into a current frequency control model to obtain an output control value
Figure BDA0003010085030000021
S5, inputting the value A1 into a load shedding control model to obtain an output control value A2 of the load shedding control model;
s6, feeding back the rotor rotating speed variation corresponding to the control value A2 to an inertia control model to obtain an output control value A3 of the inertia control model;
s7, combining the value A3 with the value
Figure BDA0003010085030000022
After addition, inputting the power model of the fan to obtain the power delta P output by the wind turbine generator to the power grid WF ；
S8, power delta P WF Inputting the output power and the load power of the prime motor into a generator load model to obtain an adjusted frequency;
s9, judging whether to continue frequency optimization, if so, optimizing target parameters of the current frequency control model through reinforcement learning, and returning to the step S2; otherwise, ending the frequency optimization.
Further, in step S1, a frequency control model is established, and the specific method for initializing the target parameters of the frequency control model is as follows:
taking the auxiliary control constant of wind farm inertia, wind farm damping and wind farm steady-state output change as a target parameter, and according to the formula:
Figure BDA0003010085030000031
Figure BDA0003010085030000032
Figure BDA0003010085030000033
Figure BDA0003010085030000034
establishing a frequency control model; wherein DeltaP FC The resulting output control values for the initial frequency control model,
Figure BDA0003010085030000035
for wind farm inertia M A Corresponding power, +.>
Figure BDA0003010085030000036
Damping D for wind farm A Corresponding power, +.>
Figure BDA0003010085030000037
Auxiliary control constant K for steady-state output change of wind farm P Corresponding power; Δf is a frequency variation value; Δf' is a value obtained by performing high-pass filtering on Δf; t represents time;
the initial value of wind farm inertia is set to 50, the initial value of wind farm damping is set to 50, and the initial value of auxiliary control constant for steady state output variation of wind farm is set to 10.
Further, the specific method for optimizing the target parameters of the frequency control model through reinforcement learning in step S9 includes the following sub-steps:
s9-1, according to the formula:
Figure BDA0003010085030000038
acquiring an evaluation value Q corresponding to the nth group of target parameters n The method comprises the steps of carrying out a first treatment on the surface of the Where N is the number of power system areas; i is the i-th region; Δt (delta t) i Time variation for the i-th region; Δf i Frequency change for the i-th region; initial evaluation value Q 0 Is minus infinity;
s9-2, according to the formula:
Q n+1 -Q n >h
judging whether the difference between the evaluation values corresponding to the target parameters in the two adjacent iteration processes is larger than a threshold h, if so, entering a step S9-3, otherwise, outputting the current target parameters;
s9-3, judging whether auxiliary control constants of the inertia of the current wind power plant, the damping of the wind power plant and steady-state output change of the wind power plant are respectively larger than a threshold value a, a threshold value b and a threshold value c, and if so, outputting current target parameters; otherwise, enter step S9-4;
s9-4, respectively carrying out gradient update on auxiliary control constants of the inertia of the current wind farm, the damping of the wind farm and the steady-state output change of the wind farm, and returning to the step S9-1.
Further, the threshold value h in step S9-2 is 10 -7 。
Further, the threshold value a in step S9-3 is 120.
Further, the threshold b in step S9-3 is set to 110.
Further, the threshold c in step S9-3 has a value of 30.
Further, the specific method for gradient update in step S9-4 is as follows:
according to the formula:
Figure BDA0003010085030000041
Figure BDA0003010085030000042
Figure BDA0003010085030000043
inertia to nth wind farm
Figure BDA0003010085030000044
Wind farm damping->
Figure BDA0003010085030000045
Auxiliary control constant for steady-state output change of wind farm
Figure BDA0003010085030000046
Gradient updating is respectively carried out to obtain the inertia of the (n+1) th wind power plant>
Figure BDA0003010085030000047
Wind farm damping->
Figure BDA0003010085030000048
Auxiliary control constant for steady-state output variation of wind farm>
Figure BDA0003010085030000049
Wherein α is the learning rate; />
Figure BDA00030100850300000410
And the frequency control model corresponding to the nth group of target parameters is represented to output control values.
The beneficial effects of the invention are as follows: according to the method, the frequency modulation control method of the wind power plant is formulated by combining the advantages of inertia control and primary frequency modulation control strategies, and the frequency control parameters in the frequency control model are optimized according to the nonlinear characteristic of frequency response and the reinforcement learning algorithm based on deterministic strategy gradient, so that the method can not only rapidly respond to transient variation of system frequency, relieve pressure of a traditional frequency modulation unit, but also realize permanent participation of the wind power plant.
Drawings
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a schematic diagram of a frequency control model;
FIG. 3 is a flow chart of a reinforcement learning optimization initial frequency control model parameter;
FIG. 4 is a step load ripple diagram in an embodiment;
FIG. 5 is a frequency deviation comparison chart of an embodiment;
FIG. 6 is a graph showing the comparison of the active outputs of a stroke motor group according to an embodiment;
FIG. 7 is a graph comparing region 1 frequency bias in a single region power system;
FIG. 8 is a graph comparing the active power output changes of zone 1 in a single zone power system;
fig. 9 is a comparison diagram of area 1ACE in a single area power system;
FIG. 10 is a graph showing frequency deviation of a single region under three schemes;
FIG. 11 is a comparison diagram of active outputs of wind turbines in a single region under three schemes;
FIG. 12 is a plot of region 1 frequency bias versus a dual region power system for three schemes;
FIG. 13 is a graph showing the active power output variation of region 1 in a dual region power system under three schemes;
fig. 14 is a comparison of regional 1ACE in a dual-regional power system for three scenarios.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
As shown in fig. 1, the frequency control optimization method for the wind farm participating in primary frequency modulation comprises the following steps:
s1, establishing a frequency control model, and initializing target parameters of the frequency control model, wherein the target parameters comprise wind farm inertia, wind farm damping and auxiliary control constants of steady-state output change of the wind farm;
s2, acquiring the current power grid frequency;
s3, judging whether the current power grid frequency deviates from a set value, if so, entering a step S4, otherwise, returning to the step S2;
s4, inputting the deviation value of the power grid frequency into an inertia control model to obtain an output control value A1 of the inertia control model; inputting the deviation value of the power grid frequency into a current frequency control model to obtain an output control value
Figure BDA0003010085030000061
S5, inputting the value A1 into a load shedding control model to obtain an output control value A2 of the load shedding control model;
s6, feeding back the rotor rotating speed variation corresponding to the control value A2 to an inertia control model to obtain an output control value A3 of the inertia control model;
s7, combining the value A3 with the value
Figure BDA0003010085030000062
After addition, inputting the power model of the fan to obtain the power delta P output by the wind turbine generator to the power grid WF ；
S8, power delta P WF Inputting the output power and the load power of the prime motor into a generator load model to obtain an adjusted frequency;
s9, judging whether to continue frequency optimization, if so, optimizing target parameters of the current frequency control model through reinforcement learning, and returning to the step S2; otherwise, ending the frequency optimization.
As shown in fig. 2, in step S1, a frequency control model is established, and the specific method for initializing the target parameters of the frequency control model is as follows:
taking the auxiliary control constant of wind farm inertia, wind farm damping and wind farm steady-state output change as a target parameter, and according to the formula:
Figure BDA0003010085030000063
Figure BDA0003010085030000064
Figure BDA0003010085030000071
Figure BDA0003010085030000072
establishing a frequency control model; wherein DeltaP FC The resulting output control values for the initial frequency control model,
Figure BDA0003010085030000073
for wind farm inertia M A Corresponding power, +.>
Figure BDA0003010085030000074
Damping D for wind farm A Corresponding power, +.>
Figure BDA0003010085030000075
Auxiliary control constant K for steady-state output change of wind farm P Corresponding power; Δf is a frequency variation value; Δf' is a value obtained by performing high-pass filtering on Δf; t represents time;
the initial value of wind farm inertia is set to 50, the initial value of wind farm damping is set to 50, and the initial value of auxiliary control constant for steady state output variation of wind farm is set to 10.
As shown in fig. 3, the specific method for optimizing the target parameters of the frequency control model by reinforcement learning in step S9 includes the following sub-steps:
s9-1, according to the formula:
Figure BDA0003010085030000076
acquiring an evaluation value Q corresponding to the nth group of target parameters n The method comprises the steps of carrying out a first treatment on the surface of the Where N is the number of power system areas; i is the i-th region; Δt (delta t) i Time variation for the i-th region; Δf i Frequency change for the i-th region; initial evaluation value Q 0 Is minus infinity;
s9-2, according to the formula:
Q n+1 -Q n >h
judging whether the difference between the evaluation values corresponding to the target parameters in the two adjacent iteration processes is larger than a threshold h, if so, entering a step S9-3, otherwise, outputting the current target parameters;
s9-3, judging whether auxiliary control constants of the inertia of the current wind power plant, the damping of the wind power plant and steady-state output change of the wind power plant are respectively larger than a threshold value a, a threshold value b and a threshold value c, and if so, outputting current target parameters; otherwise, enter step S9-4;
s9-4, respectively carrying out gradient update on auxiliary control constants of the inertia of the current wind farm, the damping of the wind farm and the steady-state output change of the wind farm, and returning to the step S9-1.
The threshold value h in step S9-2 is 10 -7 . In step S9-3, the value of the threshold value a is 120, the value of the threshold value b is 110, and the value of the threshold value c is 30. The specific method for gradient updating in the step S9-4 is as follows: according to the formula:
Figure BDA0003010085030000081
Figure BDA0003010085030000082
Figure BDA0003010085030000083
inertia to nth wind farm
Figure BDA0003010085030000084
Wind farm damping->
Figure BDA0003010085030000085
Auxiliary control constant for steady-state output change of wind farm
Figure BDA0003010085030000086
Gradient updating is respectively carried out to obtain the inertia of the (n+1) th wind power plant>
Figure BDA0003010085030000087
Wind farm damping->
Figure BDA0003010085030000088
Auxiliary control constant for steady-state output variation of wind farm>
Figure BDA0003010085030000089
Wherein α is the learning rate; />
Figure BDA00030100850300000810
And the frequency control model corresponding to the nth group of target parameters is represented to output control values.
In one embodiment of the invention, the LFC dynamic model is used as a test system, and the frequency control optimization method provided by the method is simulated and verified in MATLAB/Simulink, and the frequency modulation effect of the power system is tested by the interconnection power system of single area and double area. The wind speed of the wind power plant is changed and is set to randomly fluctuate within the range of 10.8-13.2 m/s; the set step load fluctuations are shown in fig. 4, i.e. from 10s, the load fluctuations vary between 10% and 20% 5% each time.
And obtaining the dynamic response of the frequency deviation of the single-area power system according to the set parameter simulation. Fig. 5 and fig. 6 are diagrams of frequency deviation and active output of a wind turbine under two schemes of no participation in frequency modulation and primary frequency modulation. The simulation data can be analyzed, and the frequency response and the active power change of the wind turbine generator set under the two schemes are shown in table 1.
TABLE 1
Figure BDA0003010085030000091
As can be obtained from fig. 5, fig. 6 and table 1, when the load fluctuates, the control strategy of the wind farm participating in primary frequency modulation is feasible, and the frequency deviation of a single-area power system is obviously improved. Compared with the frequency modulation without participation, the average time of frequency recovery is reduced by 24.318s, the maximum frequency deviation is reduced by 80.042 percent, namely, the overshoot is smaller, and the stabilization time is shorter. Meanwhile, when the wind power plant does not participate in frequency modulation, the wind power plant output has no power change because of no contribution to frequency adjustment of the system. In the control strategy formulated by the research, the wind turbine generator runs on a load shedding curve, the output of the wind farm is changed according to a specific load change value, and the power is changed rapidly in a short time, so that the frequency is reduced or increased smoothly. This steep slope is due to the effective active power within the wind farm inertia. As can be seen from the simulation, the average value of the active power variation output per load fluctuation is 0.1242pu. Taking the first load fluctuation as an example, the gradient becomes gentle when the power deviation increases above 0.09pu, wherein the excess power of the power deviation greater than 0.09pu is generated by the generator deceleration and the reduction of the blade angle. After a time interval of about 10s, the output reaches steady state, the power increase at steady state being provided by the generator speed and blade angle decrease. Thus taking the first load fluctuation as an example, the wind farm provides an active power increment of about 0.09pu in steady state, rapidly moderating the dip in system frequency.
And for the double-area interconnection power system, when step load fluctuates, simulating to obtain the system response of whether wind power participates in primary frequency modulation. Fig. 7, 8, and 9 are frequency deviation, active power output variation, and region control error of region 1, respectively. According to the simulation data of the two areas, the frequency response, the active power change and ACE of the double-area interconnection power system under the two schemes are shown in table 2.
TABLE 2
Figure BDA0003010085030000101
As can be obtained from fig. 7, fig. 8, fig. 9 and table 2, in the dual-area power system, it is verified that the control strategy of the wind power participation primary frequency modulation provided by the method has a good effect, and is reflected in obviously improving the system frequency deviation and the tie power deviation, and the two aspects can be obtained from the values of the ACE in the simulation result. The maximum ACE was reduced by 64.389% compared to the absence of frequency modulation. For the system frequency response, the average time of frequency recovery when the wind farm participates in primary frequency modulation is reduced by 1.4551s, the maximum frequency deviation is reduced by 64.117 percent, the frequency adjustment speed is increased, and the frequency is more stable. And for the active change mean value of the wind turbine, similar to a single-area system, the output of the wind farm has no power change when the wind turbine does not participate in frequency modulation. When participating in frequency modulation, the average power variation of the wind farm injection system is 0.06243pu, namely when step load fluctuates, the wind farm helps the system to increase active output and quickly balance the increased part of load.
In order to better verify the effectiveness of the proposed frequency control optimization method, the method compares and analyzes the response condition of the frequency in the following control scheme on the basis of the participation of the wind turbine generator in primary frequency modulation:
(1) the frequency control model adopts initial parameters;
(2) the frequency control model is optimized by adopting a particle swarm algorithm;
(3) the frequency control model is optimized by adopting a reinforcement learning algorithm based on DPG.
PID parameters after the optimization of the single-area and double-area interconnected power system are shown in tables 3 and 4.
TABLE 3 Table 3
Proportional constant DA Integral constant KP Differential constant MA
101.5492 24.5020 112.3150
TABLE 4 Table 4
Area number Proportional constant DA Integral constant KP Differential constant MA
Zone
1 100.5739 13.7575 112.0129
Zone 2 101.4445 13.9692 111.2066
The three schemes are applied to a single-area power system, and the obtained simulation comparison chart is as follows: FIG. 10 depicts the frequency response characteristics under four schemes with step load fluctuations; FIG. 11 illustrates wind power output of a single area power system. From the simulation results, the frequency response and active power variation under the three schemes are shown in table 5.
TABLE 5
Figure BDA0003010085030000111
As can be obtained from fig. 10, 11 and table 5, when the step load fluctuates, the recovery time average value of the frequency and the maximum frequency deviation are larger, and the active power output is also not ideal for the scheme that the wind turbine participates in primary frequency modulation but does not optimize the frequency modulation controller; after the particle swarm optimization is carried out, the maximum frequency deviation is reduced by 8.475%, the frequency recovery time average value is increased by 2.4111s, and the method has a certain improvement effect on the frequency response characteristic, but the improvement effect is not obvious, and the recovery speed of the system frequency is slowed down; after the DPG-based reinforcement learning algorithm is utilized to optimize, the frequency recovery time average value is obviously reduced, the maximum frequency deviation is reduced by 41.005%, and the output change average value of the wind power plant reaches 0.1418pu, so that the active power output by the wind turbine generator is more ideal and is closer to the amplitude of load fluctuation after the algorithm is optimized, the frequency deviation of a single-area power system is obviously reduced, and the system frequency is more quickly stabilized.
The three schemes are applied to the double-area interconnection power system, and the obtained simulation comparison chart is as follows: FIG. 12 depicts the frequency response characteristics of region 1 at step load fluctuations; FIG. 13 illustrates wind power output for region 1; fig. 14 depicts the ACE of region 1. As can be seen from the simulation results, the frequency response, active power variation and ACE response of the dual-area interconnected power system under the three schemes are shown in table 6.
TABLE 6
Figure BDA0003010085030000121
As can be seen from fig. 12, 13, 14 and table 6, similar to the case of a single-area power system, when a step load disturbance occurs in the system, the active power output is not ideal due to the fact that the wind turbine generator participates in primary frequency modulation but does not optimize the frequency modulation controller, so that the average value of the system frequency and the ACE recovery time is longest, and the maximum frequency deviation and the ACE are also largest; after the particle swarm optimization, the response of ACE of the power system is mainly compared, the average value of ACE recovery time is reduced by 1.8339s, the maximum value of ACE is reduced by 9.734%, and the method has a certain improvement effect on regional control deviation, but the overshoot and the variation amplitude of stable time are smaller, and the optimization effect is poor; after the DPG-based reinforcement learning algorithm is used for optimization, the ACE recovery time average value is obviously reduced, the maximum value of ACE is reduced by 27.812%, and the output change average value of the wind power plant is 0.8527pu, so that the active power output by the wind turbine generator is optimal and is closest to the amplitude of load fluctuation after the algorithm is optimized, the adjustment process is stable, and the maximum value of ACE of the system is obviously reduced.
According to the analysis, the optimization method provided by the invention has great help to the improvement of the system response, is simultaneously applied to a double-area interconnection power system, and also verifies the effectiveness of the DPG-based reinforcement learning algorithm for simultaneously setting a plurality of PID controllers.
In summary, the invention provides a frequency control optimization scheme of the wind turbine participating in primary frequency modulation based on the traditional LFC model. On the basis of the establishment of an active power control strategy of the wind turbine, the primary frequency modulation controller is designed (a frequency control model is formed), and parameters of the PID controller are optimized by utilizing a DPG-based strengthening algorithm, so that the primary frequency modulation controller can respond to the change of the system frequency better. Finally, the effectiveness of the active power control strategy and the optimization scheme of the wind power plant is verified through simulation. The main conclusion of the invention is as follows:
1) The active power of the wind turbine generator is controlled by combining the advantages of inertia control and load shedding control, so that the system can rapidly respond to transient changes of system frequency, and permanent participation of a wind farm is realized.
2) Compared with the wind turbine active power control method that wind power stations do not participate in frequency modulation, the wind turbine active power control method provided by the method has the advantages that the recovery time of the frequency is reduced by about 24 seconds on average, the amplitude of frequency fluctuation is reduced by 70% on average, the system frequency stability is improved, and the wind turbine active power control method can effectively realize the wind turbine participation in primary frequency modulation.
3) By using the frequency control optimization method based on the reinforcement learning algorithm, compared with the response of a system which is not optimized and is optimized by using the particle swarm algorithm, simulation shows that the optimized regional control deviation is reduced by about 30 percent compared with the non-optimized average, is reduced by about 10 percent compared with the optimized average of the particle swarm algorithm, and the active power output of the wind turbine is optimal, so that the pressure of the traditional frequency modulation unit is relieved.

Claims (8)

1. The frequency control optimization method for the primary frequency modulation of the wind farm is characterized by comprising the following steps of:
s1, establishing a frequency control model, and initializing target parameters of the frequency control model, wherein the targets comprise wind farm inertia, wind farm damping and auxiliary control constants of steady-state output change of the wind farm;
s2, acquiring the current power grid frequency;
s3, judging whether the current power grid frequency deviates from a set value, if so, entering a step S4, otherwise, returning to the step S2;
s4, inputting the deviation value of the power grid frequency into an inertia control model to obtain an output control value A1 of the inertia control model; inputting the deviation value of the power grid frequency into a current frequency control model to obtain an output control value
Figure FDA0003010085020000012
S5, inputting the value A1 into a load shedding control model to obtain an output control value A2 of the load shedding control model;
s6, feeding back the rotor rotating speed variation corresponding to the control value A2 to an inertia control model to obtain an output control value A3 of the inertia control model;
s7, combining the value A3 with the value
Figure FDA0003010085020000011
After addition, inputting the power model of the fan to obtain the power delta P output by the wind turbine generator to the power grid WF ；
S8, power delta P WF Inputting the output power and the load power of the prime motor into a generator load model to obtain an adjusted frequency;
s9, judging whether to continue frequency optimization, if so, optimizing target parameters of the current frequency control model through reinforcement learning, and returning to the step S2; otherwise, ending the frequency optimization.
2. The method for optimizing frequency control of a wind farm participating in primary frequency modulation according to claim 1, wherein in step S1, a frequency control model is established, and the specific method for initializing target parameters of the frequency control model is as follows:
taking the auxiliary control constant of wind farm inertia, wind farm damping and wind farm steady-state output change as a target parameter, and according to the formula:
Figure FDA0003010085020000021
Figure FDA0003010085020000022
Figure FDA0003010085020000023
Figure FDA0003010085020000024
establishing a frequency control model; wherein DeltaP FC The resulting output control values for the initial frequency control model,
Figure FDA0003010085020000025
for wind farm inertia M A Corresponding power, +.>
Figure FDA0003010085020000026
Damping D for wind farm A Corresponding power, +.>
Figure FDA0003010085020000027
Auxiliary control constant K for steady-state output change of wind farm P Corresponding power; Δf is a frequency variation value; Δf' is a value obtained by performing high-pass filtering on Δf; t represents time;
the initial value of wind farm inertia is set to 50, the initial value of wind farm damping is set to 50, and the initial value of auxiliary control constant for steady state output variation of wind farm is set to 10.
3. The method for optimizing frequency control of a wind farm participating in primary frequency modulation according to claim 1, wherein the specific method for optimizing the target parameters of the frequency control model by reinforcement learning in step S9 comprises the following sub-steps:
s9-1, according to the formula:
Figure FDA0003010085020000028
acquiring an evaluation value Q corresponding to the nth group of target parameters n The method comprises the steps of carrying out a first treatment on the surface of the Where N is the number of power system areas; i is the i-th region; Δt (delta t) i Time variation for the i-th region; Δf i Frequency change for the i-th region; initial evaluation value Q 0 Is minus infinity;
s9-2, according to the formula:
Q n+1 -Q n >h
judging whether the difference between the evaluation values corresponding to the target parameters in the two adjacent iteration processes is larger than a threshold h, if so, entering a step S9-3, otherwise, outputting the current target parameters;
s9-3, judging whether auxiliary control constants of the inertia of the current wind power plant, the damping of the wind power plant and steady-state output change of the wind power plant are respectively larger than a threshold value a, a threshold value b and a threshold value c, and if so, outputting current target parameters; otherwise, enter step S9-4;
s9-4, respectively carrying out gradient update on auxiliary control constants of the inertia of the current wind farm, the damping of the wind farm and the steady-state output change of the wind farm, and returning to the step S9-1.
4. The method for optimizing frequency control of wind farm participation in primary frequency modulation according to claim 3, wherein the value of the threshold h in step S9-2 is 10 -7 。
5. A method for optimizing frequency control of a wind farm engaged in primary frequency modulation according to claim 3, wherein the threshold value a in step S9-3 is 120.
6. A method for optimizing frequency control of a wind farm engaged in primary frequency modulation according to claim 3, wherein the threshold b in step S9-3 is valued at 110.
7. The method for optimizing frequency control of a wind farm participating in primary frequency modulation according to claim 3, wherein the value of the threshold c in step S9-3 is 30.
8. The method for optimizing frequency control of wind farm participation in primary frequency modulation according to claim 3, wherein the specific method for gradient update in step S9-4 is as follows:
according to the formula:
Figure FDA0003010085020000031
Figure FDA0003010085020000032
Figure FDA0003010085020000033
inertia to nth wind farm
Figure FDA0003010085020000034
Wind farm damping->
Figure FDA0003010085020000035
Auxiliary control constant for steady-state output variation of wind farm>
Figure FDA0003010085020000036
Gradient updating is respectively carried out to obtain the inertia of the (n+1) th wind power plant>
Figure FDA0003010085020000037
Wind farm damping->
Figure FDA0003010085020000038
Auxiliary control constant for steady-state output variation of wind farm>
Figure FDA0003010085020000041
Wherein α is the learning rate; />
Figure FDA0003010085020000042
And the frequency control model corresponding to the nth group of target parameters is represented to output control values.
CN202110376547.1A 2021-04-07 2021-04-07 Frequency control optimization method for wind power plant participating in primary frequency modulation Active CN113098030B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110376547.1A CN113098030B (en) 2021-04-07 2021-04-07 Frequency control optimization method for wind power plant participating in primary frequency modulation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110376547.1A CN113098030B (en) 2021-04-07 2021-04-07 Frequency control optimization method for wind power plant participating in primary frequency modulation

Publications (2)

Publication Number Publication Date
CN113098030A CN113098030A (en) 2021-07-09
CN113098030B true CN113098030B (en) 2023-06-16

Family

ID=76675094

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110376547.1A Active CN113098030B (en) 2021-04-07 2021-04-07 Frequency control optimization method for wind power plant participating in primary frequency modulation

Country Status (1)

Country Link
CN (1) CN113098030B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102064548A (en) * 2010-12-17 2011-05-18 中国电力科学研究院 Transient frequency deviation considered method for optimizing rotation reserve of once frequency modulation
CN108832658A (en) * 2018-06-22 2018-11-16 三峡大学 A kind of wind power penetration limit calculation method considering frequency constraint and wind-powered electricity generation frequency modulation
CN109888796A (en) * 2019-03-06 2019-06-14 中国电力科学研究院有限公司 A kind of battery energy storage power station control method and system for primary frequency regulation of power network
CN110048440A (en) * 2019-05-29 2019-07-23 国网陕西省电力公司电力科学研究院 A kind of wind power generating set participates in the control method and model of primary frequency regulation of power network
CN111864813A (en) * 2020-06-23 2020-10-30 国网辽宁省电力有限公司电力科学研究院 Wind/thermal power combined frequency control method based on virtual weight coefficient

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160160839A1 (en) * 2014-12-09 2016-06-09 State Grid Corporation Of China Method for controlling inertia response of variable-speed wind turbine generator

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102064548A (en) * 2010-12-17 2011-05-18 中国电力科学研究院 Transient frequency deviation considered method for optimizing rotation reserve of once frequency modulation
CN108832658A (en) * 2018-06-22 2018-11-16 三峡大学 A kind of wind power penetration limit calculation method considering frequency constraint and wind-powered electricity generation frequency modulation
CN109888796A (en) * 2019-03-06 2019-06-14 中国电力科学研究院有限公司 A kind of battery energy storage power station control method and system for primary frequency regulation of power network
CN110048440A (en) * 2019-05-29 2019-07-23 国网陕西省电力公司电力科学研究院 A kind of wind power generating set participates in the control method and model of primary frequency regulation of power network
CN111864813A (en) * 2020-06-23 2020-10-30 国网辽宁省电力有限公司电力科学研究院 Wind/thermal power combined frequency control method based on virtual weight coefficient

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
双馈风力发电机的一次调频控制;胡浪;李时东;袁清晨;徐大为;;湖北民族学院学报(自然科学版)(第03期);28-33 *
风电调频的控制参数选择及其影响分析;黑阳;付媛;王毅;张祥宇;;电测与仪表(第05期);16-21 *

Also Published As

Publication number Publication date
CN113098030A (en) 2021-07-09

Similar Documents

Publication Publication Date Title
Zholtayev et al. Adaptive super-twisting sliding mode control for maximum power point tracking of PMSG-based wind energy conversion systems
CN109936152B (en) Grid frequency modulation control method after high-permeability wind power grid connection and wind driven generator
CN105180135A (en) Unit plant coordination control method and system
CN114844064B (en) Self-adaptive variable parameter frequency modulation method and device for double-fed variable-speed pumped storage unit
Burkart et al. Nonlinear control of wind turbines: An approach based on switched linear systems and feedback linearization
Belghazi et al. Pitch angle control for variable speed wind turbines using genetic algorithm controller
Joo et al. Power regulation of variable speed wind turbines using pitch control based on disturbance observer
Dhar et al. Study on pitch angle control of a variable speed wind turbine using different control strategies
CN115459303A (en) Self-adaptive control method for participating in primary frequency modulation of power grid by battery energy storage
CN113471989A (en) Intelligent micro-grid secondary frequency control method based on wolf optimization method
Zeghdi et al. Improved backstepping control of a DFIG based wind energy conversion system using ant lion optimizer algorithm
CN115276039A (en) Rotor kinetic energy nonlinear control method suitable for frequency adjustment of wind power grid-connected system
Malmir et al. Controlling megawatt class WECS by ANFIS network trained with modified genetic algorithm
Bakir et al. Experimental evaluation of water cycle technique for control parameters optimization of double-fed induction generator-based wind turbine
CN113098030B (en) Frequency control optimization method for wind power plant participating in primary frequency modulation
Hemeyine et al. Robust takagi sugeno fuzzy models control for a variable speed wind turbine based a DFI-generator
CN117028141A (en) Variable pitch coordinated optimization control method for wind turbine generator
CN116345490A (en) Wind turbine generator system improved stepping inertia control frequency modulation strategy based on optimized power curve
CN116154804A (en) Fan-water turbine primary frequency modulation control method considering standby flexibility
CN115182844A (en) Bounded UDE torque control method for variable-speed wind generating set
Wang et al. LPV active power control and robust analysis for wind turbines
Narasimalu et al. Pitch angle control for horizontal axis wind turbine: A comparative study
CN113595153A (en) Output power optimization method and device of new energy cluster
Ilham et al. Monitoring of wind energy conversion system by on dimensional adaptive tuning fuzzy logic controller
Hou et al. Study of modeling and intelligent control on AGC system with wind power

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant