CN114447951A

CN114447951A - Dynamic setting method for frequency deviation coefficient of AGC system

Info

Publication number: CN114447951A
Application number: CN202111277943.5A
Authority: CN
Inventors: 杨知方; 向明旭; 余娟; 黄俊凯; 许懿; 胡润滋; 杨渝璐; 刘伟
Original assignee: Chongqing University; State Grid Chongqing Electric Power Co Ltd
Current assignee: Chongqing University; State Grid Chongqing Electric Power Co Ltd
Priority date: 2021-10-30
Filing date: 2021-10-30
Publication date: 2022-05-06

Abstract

The invention discloses a dynamic setting method for an AGC system frequency deviation coefficient, which comprises the following steps: 1) acquiring power grid data; 2) calculating a natural frequency characteristic coefficient beta (t) of the power grid; 3) and according to the frequency deviation delta f, setting the frequency deviation B coefficient as a power grid natural frequency characteristic coefficient beta (t). The method can effectively predict the beta coefficient of the power grid, and the set B coefficient can better adapt to the operating condition of the power grid, thereby realizing the coordinated control of AGC and primary frequency modulation, effectively avoiding the occurrence of the phenomenon of frequency inverse modulation (namely the conflict between an AGC command and a primary frequency modulation command), and having important significance for maintaining the frequency stability of the power grid.

Description

Dynamic setting method for frequency deviation coefficient of AGC system

Technical Field

The invention relates to the field of power systems and automation thereof, in particular to a dynamic setting method for an AGC system frequency deviation coefficient.

Background

The frequency of the power system is one of three main indicators of the quality of electric energy, reflects the balance relation between the active power and the load of power generation, is an important control parameter for the operation of the power system, and is also closely related to the safety and efficiency of power equipment of a large number of users and power generation and supply equipment. For example, power system frequency deviations will greatly affect time-dependent peripheral equipment, and may also cause accelerated wear of parts such as turbine blades. Therefore, the quality of the frequency control effect directly influences the overall economic benefit of the society. Automatic Generation Control (AGC) is an important technical means for realizing power grid active power Control and ensuring system frequency quality. The AGC system can automatically adjust the generating power of the generator according to the real-time monitoring data of the power grid and the self adjusting characteristics of the generator set, so that the real-time balance of the generating power and the load is maintained, and the frequency quality of the power grid is guaranteed. In the AGC control process, the amount of system active adjustment is typically determined by calculating ACE. The AGC system includes a plurality of control modes, with different control modes having different ACE calculations. Among them, Tie-line frequency deviation Control (Tie-line frequency Bias Control) is the most widely used Control mode at present. The control objective of this mode is to control both the grid frequency and the tie line power around the planned values.

In the calculation process of ACE, the frequency deviation B coefficient is an important parameter, which determines the preference of AGC system to tie-line and frequency. When the B coefficient is larger than the β coefficient, the AGC control will be biased to adjust the frequency deviation, which will cause frequency overshoot, causing unnecessary adjustment costs. When the coefficient B is smaller than the coefficient beta, AGC control is biased to adjust a connecting line, frequency undershooting or even reverse adjustment is caused, the direction is opposite to the adjustment direction of primary frequency modulation of a power grid, and frequency recovery is not facilitated. When the coefficient B is equal to the coefficient beta, the AGC control system of each area is only responsible for processing the load disturbance generated in the area, and the frequency overshoot, undershoot and back regulation can be avoided. The beta coefficient reflects the relation between the grid frequency and the power under the steady state condition, and is the inherent characteristic of the power system. The beta coefficient is not fixed in value, but is influenced by system running states such as load fluctuation, spare capacity, unit start-stop modes and the like, and shows time-varying property and nonlinearity. This presents challenges to the proper tuning of the B coefficient.

In the past, the B coefficient in the industry is generally set according to expert experience, the B coefficient is often set to be 1% -2% of the annual maximum load, and the B coefficient is kept unchanged after setting and is only adjusted once a year. The method for fixing the B coefficient achieves better effect in the traditional power grid with insignificant random fluctuation characteristics. However, with the rapid development of new energy in China, the random fluctuation characteristic of the power grid is remarkably increased, the operation characteristic of the power grid becomes complicated and changeable, and the beta coefficient of the power grid is also greatly changed. In this case, the problem that the conventional method for fixing the B coefficient is difficult to track the change of the beta coefficient is highlighted. The invention also discovers that the inverse modulation phenomenon caused by the undersized B coefficient occurs in the operation process of the power grid by analyzing the actual data of a certain provincial power grid in China, so that the frequency of the power grid is deteriorated and the operation safety of the power grid is threatened.

In order to solve the above problems, the academic world proposes a segmented B coefficient setting method, which comprises: two-section type B coefficient setting based on the size of the system frequency deviation in a sectional manner is carried out, and a B coefficient set value is increased when the frequency deviation is larger so as to ensure that B is larger than beta; dividing the frequency deviation into 4 grades, and continuously increasing the setting value of the B coefficient along with the increase of the frequency deviation to ensure that B is more than beta; establishing an all-day time-interval B coefficient setting model according to a system operation mode, so that the model can better approximate a beta coefficient; the three-section type B coefficient setting method is established according to the setting condition of the primary frequency modulation dead zone of the thermal power generating unit and the hydroelectric generating unit. The segmented B coefficient setting method can relieve the phenomena of system underregulation and reverse regulation to a certain extent, but the setting value of each segment is still a fixed value set in advance, so that the B coefficient is still relatively fixed, and the operation conditions of a power grid such as start and stop of a unit, the reserve capacity of the unit and the like are difficult to take into account. In order to better track the change of the beta coefficient, a learner provides a beta coefficient point prediction method, which can predict the value of the beta coefficient according to the system power disturbance quantity and provide reference for setting the B coefficient. However, the research is carried out based on simulation data, historical data of the beta coefficient is still lacked in the power grid at present, and prediction of the actual value of the beta coefficient of the power grid is difficult to support.

Disclosure of Invention

The invention aims to provide a dynamic setting method for an AGC system frequency deviation coefficient, which comprises the following steps:

1) and acquiring power grid data.

The power grid data comprise a day-ahead scheduling plan, a unit starting and stopping state on the day, a unit primary frequency modulation response characteristic of a unit in a starting state at the running time, and power grid load.

2) And calculating a natural frequency characteristic coefficient beta (t) of the power grid.

The step of calculating the natural frequency characteristic coefficient beta (t) of the power grid comprises

2.1) calculating the load damping characteristic D (t) of the power grid and the total primary frequency modulation capacity K of the generator set_gen(t)。

The grid load damping characteristic d (t) is as follows:

in the formula, K_damp(t) is a load damping coefficient. L (t) is the grid load, f_nRepresenting the nominal frequency of the system.

Total primary frequency modulation capability K of generator set_gen(t) is as follows:

in the formula, N represents the number of units with frequency modulation capability in the power grid at the moment t. k is a radical of_iAnd the frequency modulation characteristics of the ith frequency modulation unit.

Grid natural frequency characteristic coefficient beta (t), grid load damping characteristic D (t), and total primary frequency modulation capacity K of generator set_gen(t) satisfies the following relation:

β(t)＝D(t)+K_gen(t) (3)

in the formula, β (t) is a characteristic coefficient of the natural frequency of the power grid.

2.2) according to the load damping characteristic D (t) of the power grid and the total primary frequency modulation capacity K of the generator set_gen(t), calculating to obtain a power grid natural frequency characteristic coefficient beta (t), namely:

in the formula, K_damp(t) is a load damping coefficient. L (t) is the grid load, f_nRepresenting the nominal frequency of the system. And N represents the number of the units with frequency modulation capability in the power grid at the moment t. k is a radical of_iAnd the frequency modulation characteristic of the ith frequency modulation unit.

2.3) establishing a piecewise function of the frequency modulation characteristic of the ith generating set, namely:

in the formula, DB_iThe frequency modulation dead zone of the ith generating set. R_iAnd the frequency modulation coefficient of the ith generating set. U shape_i、L_iRespectively, an upper frequency modulation limit and a lower frequency modulation limit. Delta P_iThe power variation of the generator set. Δ f is the frequency deviation.

And 2.4) substituting the formula (5) into the formula (4) to obtain a natural frequency characteristic coefficient beta (t) of the power grid.

3) And according to the frequency deviation delta f, setting the frequency deviation B coefficient as a power grid natural frequency characteristic coefficient beta (t).

The frequency deviation B coefficient is used to calculate the area control deviation ACE.

The area control deviation ACE is as follows:

ACE＝(P_tie,a-P_tie,s)+B(f_a-f_s)＝ΔP_tie+BΔf (6)

in the formula, P_tie,a、P_tie,sThe actual value and the planned value of the power of the tie line are calculated. f. of_a、f_sAs actual value and meter of frequencyAnd dividing values. Delta P_tieAnd Δ f represent the tie line power deviation and the frequency deviation, respectively.

The area control deviation ACE is used to calculate an area adjustment requirement ARR.

The regional regulatory requirements ARR are as follows:

ARR＝-K_PACE-K_I∫ACEdt (7)

in the formula, K_PIs a proportional parameter in the PI control. K is_IIs an integral parameter in the PI control.

It is worth explaining that the natural frequency characteristic beta coefficient of the power grid is estimated, and dynamic setting of the frequency deviation B coefficient in the power grid automatic power generation control strategy is achieved accordingly, so that the control target that an automatic power generation control system in each control area is only responsible for local disturbance is met, the frequency back-tuning phenomenon caused by unreasonable setting of the B coefficient is effectively avoided, and the frequency quality of the power grid is ensured. In particular to the contents of estimation of a natural frequency characteristic coefficient of a power grid, a dynamic setting strategy of a frequency deviation B coefficient and the like.

It is worth explaining that, the invention firstly analyzes the actual operation data of a certain province network in China, reveals that the fixed B coefficient adopted by the power grid at present has the problem of frequency reverse adjustment caused by small in the power grid operation process; then estimating the primary frequency modulation capability of the power grid through actual operation data of the power grid according to the physical meaning of the natural frequency characteristic coefficient of the power grid, and further obtaining an estimated value of the beta coefficient; and finally, realizing dynamic setting of the B coefficient based on a beta coefficient estimation result. The results of example simulation on the two-region interconnected system show that the dynamic B coefficient setting method provided by the invention can be self-adaptive to the operation condition of a power grid, can reflect the change condition of the beta coefficient, and effectively avoids the condition that an AGC command is contradictory to a primary frequency modulation command.

The invention aims to pre-estimate the natural frequency characteristic beta coefficient of a power grid and accordingly realize the dynamic setting of the frequency deviation B coefficient in the power grid automatic generation control strategy, thereby meeting the control target that each control area automatic generation control system is only responsible for the local area disturbance, effectively avoiding the frequency back-tuning phenomenon caused by the unreasonable setting of the B coefficient and ensuring the frequency quality of the power grid. In particular to the contents of estimation of a natural frequency characteristic coefficient of a power grid, a dynamic setting strategy of a frequency deviation B coefficient and the like.

The technical effect of the invention is undoubted, and in order to better adapt to the change situation of the beta coefficient and avoid the occurrence of the phenomenon of AGC frequency back regulation, the invention provides a natural frequency characteristic coefficient estimation method based on the actual operation data of a power grid, and provides a dynamic setting method of the AGC frequency deviation coefficient based on the natural frequency characteristic coefficient estimation method. The invention has the following effective effects:

1) according to the physical meaning of the natural frequency characteristic coefficient, a power grid natural frequency characteristic coefficient estimation method is provided;

2) a dynamic setting method of the B coefficient is further provided, and the method can be adaptive to the operation condition of the power grid. The method provided by the invention can effectively estimate the beta coefficient of the power grid, and the set B coefficient can better adapt to the operating condition of the power grid, thereby realizing the coordinated control of AGC and primary frequency modulation, effectively avoiding the occurrence of frequency back modulation (namely, the conflict between an AGC command and a primary frequency modulation command) and having important significance for maintaining the stability of the frequency of the power grid.

Drawings

FIG. 1 shows that the year 2020 is 1 month 14 days 13:09:15_～13:10:15, ACE history of a certain provincial power grid in China;

FIG. 2 shows that 9 and 28 months in 2019 show 19:23:50_～19:24:25, ACE history of a certain provincial power grid in china;

FIG. 3 shows the AGC basic control logic of a provincial power grid in China;

FIG. 4 is a schematic diagram of theoretical analysis of the contradiction of the commands occurring when the frequency decreases.

FIG. 5 is a schematic diagram illustrating a theoretical analysis of the occurrence of command conflicts with increasing frequency;

FIG. 6 is a primary frequency modulation response curve of a generator set;

FIG. 7 is a (a) load curve of a provincial power grid in China at 1 month and 14 days in 2020; (b) a natural frequency characteristic coefficient;

FIG. 8 is a frequency deviation coefficient dynamic tuning strategy;

FIG. 9 is a frequency response model of a two-region interconnected network;

fig. 10(a) shows the receiving grid frequency deviation coefficient; FIG. 10(b) shows the frequency deviation coefficient of the transmitting-end grid;

fig. 11(a) shows calculated values and components of the receiving-end power grid ACE in example 1; fig. 11(b) is the calculated value of the receiving-end grid ARR in the calculation example 1; fig. 11(c) shows AGC commands and primary frequency modulation commands of the thermal power generating unit of the receiving-end power grid in the embodiment 1; fig. 11(d) shows AGC commands and primary frequency modulation commands of a receiving-end grid hydroelectric generating set in the embodiment 1;

fig. 12(a) shows calculated values and components of the receiving-end power grid ACE in example 2; fig. 12(b) is calculated values of the receiving-end grid ARR in the embodiment 2; fig. 12(c) shows an AGC command and a primary frequency modulation command of a thermal power generating unit of a receiving end power grid in the calculation example 2; fig. 12(d) shows AGC commands and primary frequency modulation commands of a receiving-end grid hydroelectric generating set in the embodiment 2;

fig. 13(a) shows calculated values and components of the receiving-end power grid ACE in example 3; FIG. 13(b) is the calculated value of the receiving end grid ARR in the embodiment 3; fig. 13(c) shows AGC commands and primary frequency modulation commands of the thermal power generating unit of the receiving-end power grid in the embodiment 3; fig. 13(d) shows AGC commands and primary frequency modulation commands of a receiving-end grid hydroelectric generating set in the embodiment 3;

fig. 14(a) shows calculated values and components of the receiving-end grid ACE in example 4; FIG. 14(b) is the calculated value of the receiving-end grid ARR in the embodiment 4; fig. 14(c) shows an AGC command and a primary frequency modulation command of a thermal power generating unit of a receiving-end power grid in example 4; fig. 14(d) shows AGC commands and primary frequency modulation commands of the receiving-end grid hydroelectric generating set in the embodiment 4;

fig. 15(a) is receiving grid frequency deviation; fig. 15(b) shows the receive grid tie line exchange power deviation;

fig. 16(a) is a receiving-end grid load curve and a dispatching curve; fig. 16(b) is a load curve and a scheduling curve of a transmitting-end power grid;

fig. 17 shows the calculated value of the receiving grid ACE and its components (example 1);

fig. 18 shows the AGC and primary frequency modulation commands of the thermal power generating unit of the receiving end power grid (example 1).

Fig. 19 shows the receiving grid hydroelectric generating set AGC and primary frequency modulation command (example 1).

Fig. 20 shows the calculated values and their components of the receiving grid ACE (example 2).

Fig. 21 shows AGC and primary frequency modulation commands of a thermal power generating unit of a receiving-end power grid (example 2).

Fig. 22 shows the receiving grid hydroelectric generating set AGC and primary frequency modulation command (example 1).

Detailed Description

The present invention will be further described with reference to the following examples, but it should be understood that the scope of the subject matter described above is not limited to the following examples. Various substitutions and alterations can be made without departing from the technical idea of the invention and the scope of the invention is covered by the present invention according to the common technical knowledge and the conventional means in the field.

Example 1:

referring to fig. 1 to 22, a method for dynamically setting a frequency deviation coefficient of an AGC system includes the following steps:

1) and acquiring power grid data.

The grid load damping characteristic d (t) is as follows:

wherein N represents the value ofThe number of units with frequency modulation capability. k is a radical of formula_iAnd the frequency modulation characteristic of the ith frequency modulation unit.

β(t)＝D(t)+K_gen(t) (3)

in the formula, DB_iThe frequency modulation dead zone of the ith generating set is obtained. R_iAnd the frequency modulation coefficient of the ith generating set. U shape_i、L_iRespectively, an upper frequency modulation limit and a lower frequency modulation limit. Delta P_iThe power variation of the generator set. Δ f is the frequency deviation.

3) And setting the frequency deviation B coefficient as a power grid natural frequency characteristic coefficient beta (t) according to the frequency deviation delta f, namely setting B to beta (t).

The area control deviation ACE is as follows:

ACE＝(P_tie,a-P_tie,s)+B(f_a-f_s)＝ΔP_tie+BΔf (6)

in the formula, P_tie,a、P_tie,sThe actual value and the planned value of the power of the tie line are calculated. f. of_a、f_sThe actual value and the planned value of the frequency. Delta P_tieAnd Δ f represent the tie line power deviation and the frequency deviation, respectively.

The regional regulatory requirements ARR are as follows:

ARR＝-K_PACE-K_I∫ACEdt (7)

in the formula, K_PIs a proportional parameter in the PI control. K_IIs an integral parameter in the PI control.

Example 2:

a dynamic setting method for an AGC system frequency deviation coefficient comprises the following steps:

1) analysis of actual operation data of Chinese provincial power grid

In the actual operation process of a certain actual power grid in China, the phenomenon that an AGC command contradicts a primary frequency modulation command exists, namely the AGC command and the primary frequency modulation command are opposite in adjustment direction, and the AGC frequency is reversely adjusted. The invention analyzes the actual operation data of the power grid when the phenomenon occurs, and finds out that the reason of causing the instruction contradiction phenomenon is that the fixed B coefficient set by the power grid is smaller. The following description is made in detail with reference to the actual case.

Case 1(2020, 1 month, 14 days 13:09:15 to 13:10:15)

In an interconnected power system, a power grid analyzed by the method belongs to a receiving-end power grid. As shown in fig. 1, when the power grid receives less power at 13:09:37 on 14/1/2020, the tie line exchange power deviation in ACE is positive. And the frequency deviation component in the ACE is negative, indicating that the grid frequency is lower than the planned value. In the AGC control system, when the ACE calculated value is positive, the power grid can lower the output of the generator, and conversely, the power grid can increase the output of the generator. Thus, to adjust the tie-line, the grid should now indicate the generator derate. To adjust the frequency, the grid should indicate to the generator to add power, in opposite directions. As can be seen from fig. 1, since the absolute value of the ACE frequency deviation component is smaller than the tie line exchange power deviation component, the final ACE calculation value is positive and is consistent with the tie line exchange power deviation component. And finally, the AGC sends an output instruction to the lower part of the frequency modulation unit of the power grid. As shown in table 1, several units in the power grid all receive AGC commands for down-regulating active power. And at the moment, the unit primary frequency modulation indicates the unit increased power due to the frequency reduction, and contradicts with an AGC instruction.

Table 12020 years 1 month 14 days 13:09:15 ~ 13:10:15 AGC anti-modulation instruction record

Case 2(2019, 9, 28, 19, 23, 50, 19, 24, 25)

As shown in fig. 2, in 2019, month 9, 28, day 19:23:50, the grid receives more power, and the tie line exchange power deviation is negative. At this point, the ACE frequency deviation component is positive, indicating that the grid frequency is above the projected value. Since the absolute value of the ACE frequency deviation component is less than the crosstie exchange power deviation component, the sum of the two, i.e., the ACE calculated value, is negative. And finally, the AGC control system sends an output increasing instruction to the frequency modulation station. As shown in table 2, some plant stations receive an AGC command for adjusting up the active power. And at the moment, the unit primary frequency modulation issues a force reducing instruction due to the frequency rise, and the force reducing instruction contradicts with the AGC instruction.

In table 22019, 9/28/19/23: 50-19/24/25 AGC reverse regulation instruction record

Date	Time	Plant station	AGC command	Regulating variable (MW)	Mode(s)
						2019-09-28	19:24:08	Plant 1	888.0-->918.0	30	AUTOR
2019-09-28	19:24:08	Plant 2	287.8-->302.8	15	AUTOR

In conclusion, when the frequency of the power grid is lower than a planned value and higher than the planned value, the phenomenon that the AGC command contradicts with the primary frequency modulation command, namely, the frequency is reversely adjusted, the frequency is not recovered, and the frequency quality of the power grid is threatened. The reason for this phenomenon is theoretically analyzed below.

The basic AGC control logic for this grid is shown in fig. 3. From the figure, it can be seen that the AGC command, i.e. the Area adjustment Requirement (ARR), is determined according to the Area control deviation ACE, and the calculation formula is:

ARR＝-K_PACE-K_I∫ACEdt (1)

in the formula, K_PProportional parameters in PI control; k_IFor integration in PI controlAnd (4) parameters.

According to the actual case analysis, it can be found that when the offset component of the exchange power of the tie line is opposite to the offset component of the frequency, the phenomenon that the AGC instruction contradicts the primary frequency modulation instruction occurs. As further explained below.

As shown in FIG. 4, assume a sending end grid load Δ L_sIncreasing, then the grid frequency of the sending end is delta f_sWith the receiving grid frequency Δ f_rWill decrease, being negative. At this time,. DELTA.P_tieRises to a positive value. At this time, if the B coefficient is smaller, the calculated ACE is a positive value, and the area adjustment requirement ARR calculated according to the formula (1) is a negative value, so the AGC will indicate the fm unit in the area to reduce the output. And the frequency deviation delta f of the receiving end power grid_rIf the output is negative, the primary frequency modulation control indicates the generator to increase the output, so that the adjustment direction of the AGC command is opposite, and the command contradiction phenomenon occurs.

As shown in FIG. 5, assume a sending end grid load Δ L_sDecreasing, then the grid frequency Δ f of the transmitting end_sWith the receiving grid frequency Δ f_rWill rise, being positive. At this time, the receiving end power grid receives less power, delta P_tieAnd increasing, negative. At this time, if the B coefficient is smaller, the calculated ACE is a negative value, and the area adjustment requirement ARR calculated according to equation (1) is a positive value, so the AGC will instruct the controlled units in the area to increase capacity. And the frequency deviation delta f of the receiving end power grid_rIf the output is positive, the primary frequency modulation control indicates the generator to reduce the output, so that the adjustment direction of the AGC command is opposite, and the command contradiction phenomenon occurs.

From the above analysis, it can be seen that the main reason causing the conflict between the AGC command and the primary frequency modulation command is that the AGC control tends to preferentially maintain the link transmission power constant, i.e. the frequency deviation coefficient B in the AGC control link is smaller. In contrast, the frequency deviation B coefficient can be set to the natural frequency characteristic coefficient β of the power grid, and the above-described contradiction can be prevented to some extent. For this, the natural frequency characteristic coefficient of the power grid needs to be estimated first, and will be described in the next section. The conclusion that the current B coefficient setting value of the power grid is smaller can be verified by comparing the current B coefficient setting value of the power grid with the estimated value of the natural frequency characteristic coefficient of the power grid.

2) Power grid natural frequency characteristic coefficient calculation method

The natural frequency characteristic coefficient essentially reflects the total primary frequency modulation capability of the power grid. At time t, the natural frequency characteristic coefficient β therefore depends on the damping characteristic D [ MW/Hz ] of the grid load at the current time]Total primary frequency modulation capability K with generator set_gen[MW/Hz]Then, the total frequency modulation capability of the power grid can be recorded as:

β(t)＝D(t)+K_gen(t) (2)

as can be seen from equation (2), in order to calculate the natural frequency characteristic coefficient β, the damping characteristic of the grid load and the total primary frequency modulation capability K of the generator set need to be calculated first_gen. The calculation formula of the load damping characteristic of the power grid is as follows:

wherein, K_dampThe load damping coefficient is usually 1-3; l [ MW ]]For grid load, f_n[Hz]Representing the nominal frequency of the system.

For total primary frequency modulation capability K of generator set_genAnd the value is the sum of the frequency modulation characteristics of all the generator sets of the power grid:

wherein N represents the number of units with frequency modulation capability in the power grid at the moment t, and k_i[MW/Hz]And the frequency modulation characteristic of the ith frequency modulation unit.

The natural frequency characteristic coefficient of the system obtained by substituting the formula (3) and the formula (4) into the formula (2) is

For the ith generating set, the frequency modulation characteristic k thereof_iIs dependent onFrequency modulation dead zone DB_i[Hz]Frequency modulation factor R_i[MW/Hz]Upper limit of frequency modulation U_i[MW]And lower limit of frequency modulation L_i[MW]. Make the power variation delta P of the generator set_iAnd the frequency deviation deltaf, the frequency modulation characteristic of the generator set is shown in fig. 6.

As can be seen from the above diagram, the frequency modulation characteristic of a single generator set should be a piecewise function, and then the frequency modulation characteristic of the ith generator set can be represented by the following piecewise function:

the natural frequency characteristic coefficient of the power grid can be obtained by substituting the formula (6) into the formula (5).

Based on the calculation method, the natural frequency characteristic coefficient of the provincial power grid in China is estimated, and is compared with the current B coefficient setting value of the provincial power grid, and the comparison result is shown in figure 7. The results calculated using the actual grid operating data of 1, 13/2020 are shown in fig. 7. The load damping coefficient is closely related to the frequency sensitivity of the load and is difficult to learn, so that the system natural frequency characteristic coefficient under the three conditions of no load, the daily highest load and the daily lowest load is considered to analyze the influence of the load damping on the natural frequency characteristic coefficient, and the load damping coefficient is 1.5.

As can be seen from fig. 7, the B coefficient set in the AGC control is slightly larger than the natural frequency characteristic coefficient between the intervals [ -0.04Hz,0.04Hz ], and there is no risk of back tuning. Outside the interval, the coefficient B is far smaller than the natural frequency characteristic coefficient of the power grid, so that the inverse regulation risk exists, namely, the contradiction phenomenon is easy to occur. The (-0.04 Hz,0.04 Hz) belongs to the normal fluctuation range of the power grid frequency, and the B coefficient is only slightly larger than the natural frequency characteristic coefficient in the interval, so that no obvious overshoot phenomenon is found. And outside the range of [ -0.04Hz,0.04Hz ], the coefficient B is far smaller than the characteristic coefficient of natural frequency, when the power grid frequency is obviously deviated, the AGC cannot effectively control the frequency and even weakens the power grid frequency regulation capability, and the case of the contradiction between the AGC and the primary frequency modulation response instruction, which occurs in 1 month and 14 days of 2020 in section (1), can be referred to.

In conclusion, the situation that the setting value of the B coefficient of the power grid researched by the invention is far smaller than the natural frequency characteristic coefficient beta exists, and finally, the conflict between the AGC command and the primary frequency modulation command is caused. Therefore, a new B coefficient setting strategy needs to be proposed to more effectively adapt to the change situation of the power grid β coefficient, so as to effectively suppress the phenomenon of the conflict between the AGC command and the primary frequency modulation command.

3) AGC system frequency deviation B coefficient dynamic setting method

According to the analysis, the main reason why the AGC and the primary frequency modulation command are contradictory is that the frequency deviation coefficient B in the AGC cannot effectively reflect the natural frequency characteristic coefficient β of the power grid. For this purpose, the frequency deviation coefficient B may be set to the natural frequency characteristic coefficient calculated by equation (5). The specific implementation steps mainly comprise: (1) determining the starting and stopping states of the unit on the day according to the day-ahead scheduling plan; (2) estimating load damping according to the load prediction result, and selecting daily average load for load damping calculation because the load damping has small influence on the natural frequency characteristic coefficient; (3) determining the primary frequency modulation response characteristic of the unit in a starting state at the operation moment according to the unit data reported by the plant station; (4) calculating a power grid natural frequency characteristic beta coefficient according to the formula (5); (5) and setting the B coefficient as a beta coefficient according to the frequency deviation. As can be seen from fig. 7, the grid natural frequency characteristic coefficient is a piecewise function related to the frequency deviation magnitude, and accordingly, the set B coefficient will also be a piecewise function related to the frequency deviation magnitude.

The frequency deviation coefficient B in the AGC is set as the natural frequency characteristic coefficient calculated by the formula (5), so that the variation condition of the beta coefficient can be adapted to, the coordination work of the AGC and the primary frequency modulation of the unit is realized, and the phenomenon of contradiction between an AGC command and a primary frequency modulation command is effectively inhibited. The control strategy logic is shown in fig. 8.

Example 3:

referring to fig. 9-22, a verification test of a dynamic tuning method for frequency deviation coefficients of an AGC system includes the following steps:

1) estimation of natural frequency characteristic coefficients

And collecting the actual operation data of the power grid, wherein the actual operation data of the set (set start and stop data, set primary frequency modulation response capacity, set output data, set standby capacity and the like) and the average load predicted on the same day are included. Subsequently, the power grid natural frequency characteristic β coefficient is calculated by equation (5).

2) Dynamic setting of frequency deviation coefficient

And (3) according to the frequency deviation monitored in the actual operation of the power grid, setting the AGC frequency deviation coefficient B as the natural frequency characteristic beta coefficient calculated in the step (1).

The effectiveness of the frequency deviation coefficient dynamic setting method provided by the invention is verified by adopting a frequency response model of the two-region interconnected power grid as shown in fig. 9. Wherein, the area 1 is a transmitting end power grid, the area 2 is a receiving end power grid, and the exchange power plan value of the tie line is 1500 MW. Referring to the actual situation of a power grid, the thermal power rated capacity of a receiving end power grid is 14000MW, the rotating speed unequal rate is 0.05 (the primary frequency modulation capacity is 5600MW/Hz), the primary frequency modulation dead zone is +/-0.033 Hz, and the primary frequency modulation regulating variable is +/-8%; the rated capacity of water and electricity is set to be 5000MW, the rotating speed unequal rate is 0.04 (the primary frequency modulation capacity is 2500MW/Hz), the primary frequency modulation dead zone is +/-0.05 Hz, and the primary frequency modulation regulating quantity is +/-20%; the inertia time constant of the power grid is 5s (the equivalent rotational inertia is 3800MW & s/Hz), and the load damping coefficient is 1.5. For a transmitting-end power grid, the rated capacity of thermal power is 11000MW, the rotating speed anisometry is 0.05 (the primary frequency modulation capacity is 4400MW/Hz), the primary frequency modulation dead zone is +/-0.033 Hz, and the primary frequency modulation regulating quantity is +/-8%; the rated capacity of water and electricity is set to 50000MW, the rotating speed unequal rate is 0.04 (the primary frequency modulation capacity is 2450 MW/Hz), the primary frequency modulation dead zone is +/-0.05 Hz, and the primary frequency modulation regulating quantity is +/-20%; the inertia time constant of the sending end area is 4.9s (equivalent moment of inertia is 12000MW & s/Hz), and the load damping coefficient is 1.

In simulation, the sampling period of AGC frequency and tie line exchange power is set to be 1s, the calculation period of ACE and ARR is set to be 4s, the command issuing delay is 2s, and the command is distributed according to the ratio of thermal power to water power capacity. The integral dead zone in the receiving-end power grid AGC is +/-6 MW, the upper limit and the lower limit of the integral are +/-25 MW, and the dead zone issued by an ARR instruction is +/-22 MW; the integral dead zone in the AGC of the transmitting end power grid is +/-18 MW, the upper limit and the lower limit of the integral are +/-75 MW, and the dead zone issued by an ARR instruction is +/-66 MW.

3) Effectiveness analysis

In this section, the effectiveness of the dynamic frequency deviation coefficient setting method provided by the present invention in suppressing the conflict between AGC and the primary frequency modulation instruction of the unit is firstly verified in a single time period, and four examples shown in table 3 are set. The frequency deviation coefficients used in the proposed method are shown in fig. 10. When the frequency deviation coefficient B in the receiving end power grid AGC is 1000MW/Hz, the AGC proportionality coefficient is 0.8 and the integral coefficient is 1.2; when the method provided by the invention is adopted, the AGC proportionality coefficient is 0.4, and the integral coefficient is 1. And the proportional coefficient and the integral coefficient of the AGC of the transmission end power grid are both constant to 0.5.

Table 3 example set-up for validity analysis

Electric network	EXAMPLE 1	EXAMPLE 2	EXAMPLE 3	EXAMPLE 4
					Receiving end electric network	B＝1000MW/Hz	The method mentioned	The method mentioned	B＝1000MW/Hz
Sending end electric network	B＝3000MW/Hz	The mentioned formulaMethod of	B＝3000MW/Hz	The method mentioned

The load of a receiving end power grid is set to be 16100MW, the load of a transmitting end power grid is set to be 49300MW, and the fault is set to be that a hydroelectric generating set with the rated capacity of 1000MW and the actual transmitting power of 800MW is cut off from the transmitting end power grid in the 20 th second. The total simulation time is 80 seconds, and the simulation results of examples 1-4 are shown in FIGS. 11-14.

As shown in fig. 11(a), since the frequency decreases and the tie exchange power decreases after the occurrence of the failure, the frequency component in the ACE is negative and the tie exchange power deviation is positive. The frequency deviation falls to about 59.9Hz at the deepest under this fault. At this time, as shown in fig. 10(a), the receiving-end grid frequency deviation coefficient B is 1000MW/Hz, which is much smaller than the actual natural frequency characteristic coefficient. Therefore, the ACE calculation value in fig. 11(a) is dominated by the tie line exchange power deviation component, which is a positive value. Thus, an ARR value of negative value indicates that the unit is derating, as shown in FIG. 11 (b). As shown in fig. 11(c) and (d), therefore, the time frequency is lower than the planned value, the primary frequency modulation commands of the thermal power generating unit and the hydroelectric generating unit both indicate the output increase of the generator, and the command contradiction phenomenon is generated contrary to the AGC command.

In order to suppress the occurrence of the contradiction, the receiving-end power grid and the transmitting-end power grid in the example 2 both adopt the method provided by the invention. As shown in fig. 12(a), the ACE frequency component increases significantly at this time, and the calculated value of ACE added to the line exchange power component tends to maintain the grid frequency stable to some extent. Since the fault occurs outside the region, the ARR value in the AGC assumes a positive value for only a short time, remaining at zero for the rest of the time. That is, when a fault occurs outside the receiving grid, the receiving grid provides only limited power support, as shown in fig. 12(c) and (d).

In the embodiment 3, only the receiving-end power grid adopts the method provided by the invention, and as can be seen from fig. 13(a) - (d), the instruction contradiction phenomenon is also inhibited. The reason is that a fault occurring in the transmitting-side grid affects the frequency of the receiving-side grid by affecting the crossline exchange power, and therefore, only the crossline exchange power can be affected no matter what control method is adopted by the transmitting-side grid. If the fault occurs outside the region, the receiving end power grid can ensure that AGC is not biased to control the power deviation of the connecting line by adopting the method, so that the command contradiction phenomenon is prevented, and the control mode of the region outside the receiving end power grid is hardly related. This also explains that in example 4, when the receiving-end grid adopts a fixed B coefficient and the transmitting-end grid adopts the proposed method, the receiving-end grid still has an obvious command contradiction phenomenon, as shown in fig. 14.

In terms of frequency quality, as shown in fig. 15, when the receiving-end power grid adopts the proposed method (examples 2 and 3), the dynamic process of frequency is effectively improved, and the steady-state frequency deviation is increased from-0.0641 Hz to-0.0601 Hz and-0.0617 Hz respectively. Since the receiving grid AGC only takes short time to account for the power shortage outside the area for frequency stabilization, the tie line switching power deviation will increase to some extent, but the tie line switching power deviation should be taken up by the sending grid causing the deviation.

In summary, only the fixed coefficient B in the receiving-end power grid is replaced by the method provided by the invention, so that the contradiction between the AGC instruction and the unit primary frequency modulation instruction can be effectively inhibited, and the power grid frequency quality is improved.

4) Feasibility verification

The last section verifies the effectiveness of the frequency deviation coefficient dynamic setting method in the aspect of inhibiting the contradiction phenomenon between AGC and primary frequency modulation instructions under the conditions of constant load and single fault. However, in the actual production process, various factors such as the random fluctuation of the load, the response scheduling instruction of the unit and the like all affect the effectiveness of the control strategy. In order to explore the feasibility of the method for practical production, the section further verifies the effectiveness of the method by simulating the practical operation process of the power grid.

In this example, a 1-hour load curve and a scheduling curve as shown in fig. 16 are considered, and the scheduling period is 5 minutes. The transmission-side grid transmits power to the receiving-side grid with a planned value of 1500MW, the receiving-side grid dispatching curve shown in fig. 16(a) takes the planned value of the tie line exchange power into account, and the transmission-side grid dispatching curve shown in fig. 16(b) deducts the tie line exchange power. The example parameter settings are as in the previous section.

The faults set in the simulation include: (1) in 12 th minute, a hydroelectric generating set with the rated capacity of 1000MW and the actual generating power of 800MW is cut off from the power grid at the sending end; (2) in 24 th minute, the load loss of the power grid at the transmitting end is 1000 MW; (3) in 36 th minute, the load of the receiving end power grid is lost by 300 MW; and (4) cutting off a thermal power generating unit with the rated capacity of 660MW and the actual generating power of 300MW from the receiving-end power grid in 48 minutes.

Under the above example settings, the control performance of the following two examples were tested: the AGC for both regions uses a fixed frequency deviation factor (eq. 1) and the method of the present invention (eq. 2).

Fig. 17 to 19 are simulation results of example 1. Fig. 17 shows that when the receiving-end grid fails with a fixed B coefficient, the ACE calculated value is biased to the component of the link exchange power deviation, which causes the phenomenon that the AGC command contradicts the primary frequency modulation command. And when the fault occurs in the receiving end power grid, the AGC can carry out normal adjustment by adopting a fixed frequency deviation coefficient. As shown in fig. 18 and 19, when a fault occurs in the transmission-side power grid, the AGC command and the primary frequency modulation command are in significantly opposite directions, and a command contradiction phenomenon occurs.

Fig. 20 to 22 show the simulation results of example 2. Fig. 20 shows that, after the receiving-end power grid adopts the method of the present invention, when the sending-end power grid fails, the ACE frequency deviation component is opposite to the exchange power deviation component of the tie line and has a similar magnitude, so that the ACE calculated value is maintained near zero, thereby avoiding the occurrence of command conflict. When the fault occurs in the receiving end power grid, the AGC can still be effectively controlled by adopting the method provided by the invention, and the difference between the AGC and the AGC adopting a fixed frequency deviation coefficient is not obvious. As shown in fig. 21 and 22, when a fault occurs in the transmission-side power grid, the AGC does not issue a power adjustment command basically, and only a primary frequency modulation provides a certain power support, so that a command contradiction phenomenon does not exist. And when the fault occurs in the receiving end power grid, the AGC can perform power grid regulation task in cooperation with the primary frequency modulation, so that the stability of the power grid is guaranteed.

The simulation experiment results show that the dynamic setting method of the frequency deviation B coefficient can make the B coefficient self-adaptive to the change of the natural frequency characteristic beta coefficient along with the operation condition of a power grid, thereby realizing the coordination work of AGC and the primary frequency modulation of a unit and effectively inhibiting the phenomenon of contradiction between an AGC command and the primary frequency modulation.

On the basis of analyzing the actual operation data of a certain provincial power grid in China, the invention finds that the fact that the setting value of the coefficient B of the AGC frequency deviation is smaller than the beta coefficient of the natural frequency characteristic of the power grid is the main reason for causing the conflict between the AGC command and the primary frequency modulation command. In contrast, the invention firstly provides a beta coefficient calculation method based on the physical meaning of the beta coefficient, and the beta coefficient of the power grid can be estimated through the actual operation data of the power grid. Therefore, a frequency deviation coefficient dynamic setting strategy is further provided, so that the B coefficient can be adaptive to the change of the natural frequency characteristic beta coefficient along with the operation condition of the power grid. Simulation experiment results show that the dynamic frequency deviation coefficient setting strategy provided by the invention can realize the coordination work of AGC and unit primary frequency modulation, effectively inhibit the phenomenon of contradiction between an AGC command and the primary frequency modulation, and verify the effectiveness of the method. In addition, the simulation result can further discover that the phenomenon of contradiction between an AGC instruction and primary frequency modulation can be effectively inhibited only by the local power grid by adopting the method provided by the invention.

Claims

1. A dynamic setting method for the frequency deviation coefficient of an AGC system is characterized by comprising the following steps:

1) and acquiring the power grid data.

2) Calculating a natural frequency characteristic coefficient beta (t) of the power grid;

2. The method for dynamically setting the frequency deviation coefficient of the AGC system as claimed in claim 1, wherein the power grid data includes a day-ahead scheduling plan, a unit start-stop state on the day, a unit primary frequency modulation response characteristic of a unit in a start-up state at the running time, and a power grid load.

3. The method for dynamically setting the frequency deviation coefficient of the AGC system as claimed in claim 1, wherein the step of calculating the natural frequency characteristic coefficient β (t) of the power grid comprises

1) Calculating the load damping characteristic D (t) of the power grid and the total primary frequency modulation capacity K of the generator set_gen(t)；

2) According to the load damping characteristic D (t) of the power grid and the total primary frequency modulation capacity K of the generator set_gen(t), calculating to obtain a power grid natural frequency characteristic coefficient beta (t), namely:

in the formula, K_damp(t) is the load damping coefficient; l (t) is the grid load, f_nRepresents a system rated frequency; n represents the number of the units with frequency modulation capability in the power grid at the moment t; k is a radical of_iThe frequency modulation characteristic of the ith frequency modulation unit;

3) establishing a piecewise function of the frequency modulation characteristic of the ith generating set, namely:

in the formula, DB_iThe frequency modulation dead zone is the frequency modulation dead zone of the ith generator set; r_iThe frequency modulation coefficient of the ith generator set; u shape_i、L_iRespectively an upper frequency modulation limit and a lower frequency modulation limit; delta P_iThe power variation of the generator set is obtained; Δ f is the frequency deviation;

4) substituting the formula (2) into the formula (1) to obtain a natural frequency characteristic coefficient beta (t) of the power grid.

4. The method for dynamically setting the frequency deviation coefficient of the AGC system as claimed in claim 3, wherein the load damping characteristic D (t) of the power grid is as follows:

in the formula, K_damp(t) is the load damping coefficient; l (t) is the grid load, f_nRepresenting the nominal frequency of the system.

5. The method of claim 3, wherein the total primary modulation capability K of the generator set is_gen(t) is as follows:

in the formula, N represents the number of units with frequency modulation capability in a power grid at the moment t; k is a radical of_iAnd the frequency modulation characteristic of the ith frequency modulation unit.

6. The method for dynamically setting the frequency deviation coefficient of the AGC system as claimed in claim 3, wherein the natural frequency characteristic coefficient β (t) of the power grid, the load damping characteristic D (t) of the power grid, and the total primary frequency modulation capability K of the generator set_gen(t) satisfies the following relation:

β(t)＝D(t)+K_gen(t) (5)

7. The method for dynamically tuning the frequency deviation coefficient of the AGC system according to claim 1, wherein the frequency deviation B coefficient is used for calculating a regional control deviation ACE;

the area control deviation ACE is as follows:

ACE＝(P_tie,a-P_tie,s)+B(f_a-f_s)＝ΔP_tie+BΔf (6)

in the formula, P_tie,a、P_tie,sThe actual value and the planned value of the power of the tie line are calculated; f. of_a、f_sThe actual value and the planned value of the frequency are obtained; delta P_tieAnd Δ f are respectively shown inThe tie line power deviation and frequency deviation are shown.

8. The method for dynamically tuning the frequency deviation coefficient of the AGC system as claimed in claim 7, wherein the area control deviation ACE is used for calculating an area adjustment requirement ARR;

the regional regulatory requirements ARR are as follows:

ARR＝-K_PACE-K_I∫ACEdt (7)

in the formula, K_PProportional parameters in PI control; k_IIs an integral parameter in the PI control.