New Energy Grid Low-Frequency Oscillation Suppression Method Based on FCM Wind Speed Clustering and TS Fuzzy Model

Abstract

A method for suppressing low-frequency oscillations in a new energy power grid based on FCM wind speed clustering and TS fuzzy model provided by the present invention includes the following steps: S1. Determine the regional new energy power system model, which includes Region A, Region B, and a wind farm; S2. Collect the wind speed data of the wind farm, and construct a wind speed data set V = {v1, v2, …, v n} and a membership matrix U = [u ij k×n , and solve the wind speed clustering center based on the fuzzy C-means clustering algorithm; where u ij represents the membership degree of the j-th daily average wind speed to the i-th wind speed clustering center; S3. Construct a TS fuzzy model based on the wind speed clustering center, and solve the feedback matrix coefficient K m of the doubly-fed induction generator controller based on the TS fuzzy model, and input the feedback matrix coefficient K m , the phase angle difference δ AB between Region A and Region B, and the angular velocity difference ω AB between Region A and the inertial center into the doubly-fed induction generator controller to control the doubly-fed induction generator unit to output according to the expected power.​

Description

Technical Field

[0001] The present invention relates to a method for suppressing power grid oscillations, and more particularly to a method for suppressing low-frequency oscillations in a new energy power grid based on FCM wind speed clustering and TS fuzzy model. Background Art

[0002] Developing renewable energy to make modern power systems low-carbon is of great strategic significance for solving energy crises and alleviating pollution and other problems. Compared with traditional energy sources, renewable energy is more economical and environmentally friendly, so the proportion of renewable energy in the global power system is increasing. Among all renewable energy sources (RES), wind energy dominates in modern power systems. Different wind turbine (WT) structures are used in commercial and academic fields, such as squirrel cage induction generators, permanent magnet synchronous generators, doubly-fed induction generators (DFIG), etc. Among these wind turbines, the doubly-fed asynchronous generator has received more attention due to its simple structure, high efficiency and good economy.

[0003] However, the structure of the DFIG is different from that of the synchronous generator (SG). This difference is mainly reflected in the weak coupling between the rotor and the stator of the DFIG, making the DFIG show weak damping characteristics. In addition, in order to maintain the stability of the DFIG grid-side voltage, the implementation of the phase-locked loop (PLL) exacerbates this weak damping characteristic and further deteriorates the system damping level. Therefore, integrating a large number of DFIGs into the power system to replace the SG may lead to varying degrees of weakening of the system inertia, operating characteristics and damping characteristics. In addition, considering that the wind resources required for doubly-fed converter power generation are located in remote areas, long-distance transmission lines must be used between the wind farm and the power grid, which will reduce the damping ratio of inter-area low-frequency oscillations (LFO) and increase the risk coefficient of inter-area instability of the wind energy penetrating the power system. With the upgrade of the global primary energy transmission demand and the rapid increase in the installed capacity of DFIGs worldwide, the problem of low-frequency oscillations has become more frequent.

[0004] Existing research on low-frequency oscillation (LFO) in power systems with wind energy penetration mainly focuses on oscillation mechanisms and suppression measures. In the research on LFO mechanisms, the mainstream direction is mainly based on mathematical modeling. By establishing mathematical models for the equipment in the new energy power system respectively, an overall mathematical model is formed and its control strategy is designed to suppress the LFO problems in the system. Currently, the commonly used methods include installing power system stabilizers (PSS) on synchronous generators (SG) in the power system and installing power oscillation dampers (POD) on doubly-fed induction generators (DFIG). Although these methods can suppress LFO in the new energy power grid to a certain extent, due to the expansion of the grid scale and the increase of non-linear new energy equipment, it is not only difficult to establish an accurate mathematical model when reaching a certain scale, but also the expansion of the grid structure will weaken the effect of the original suppression strategy. On the other hand, the investment in PSS and POD equipment will also increase the construction cost and operation and maintenance cost of the power system, reducing the investment benefit of new energy equipment.

[0005] Therefore, to solve the above technical problems, a new technical means is urgently needed. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a method for suppressing low-frequency oscillation in a new energy power grid based on FCM wind speed clustering and TS fuzzy model. By considering the influence of wind speed factors, performing wind speed clustering and TS fuzzy algorithm, the controller of the doubly-fed induction generator and the feedback matrix coefficients are determined, so that the doubly-fed induction generator outputs according to the expected power, thereby effectively suppressing the low-frequency oscillation in the new energy power grid, without the need to use other external devices, effectively reducing costs, and the whole process is relatively more concise and reliable than traditional technologies.

[0007] A method for suppressing low-frequency oscillation in a new energy power grid based on FCM wind speed clustering and TS fuzzy model provided by the present invention includes the following steps:

[0008] S1. Determine the model of the regional new energy power system, which includes Region A, Region B and a wind farm. Among them, Region A and Region B each have two synchronous generators. The electric energy generated by Region A and Region B is respectively input to Bus A and Bus B, and Bus A and Bus B are electrically connected. The wind farm has a doubly-fed induction generator set, and the electric energy generated by the doubly-fed induction generator set is input to Bus A;

[0009] S2. Collect the wind speed data of the wind farm, and construct a wind speed data set V = {v1, v2,..., v n} and a membership matrix U = [u ij k×n , and solve the wind speed clustering center based on the fuzzy C-means clustering algorithm; where u ij represents the membership degree of the jth daily average wind speed to the ith wind speed clustering center;​

[0010] S3. Construct a TS fuzzy model based on the wind speed clustering center, and solve the feedback matrix coefficient K of the doubly-fed induction generator controller based on the TS fuzzy model m , and the feedback matrix coefficient K m , the phase angle difference δ between region A and region B AB and the angular velocity difference ω between region A and the center of inertia AB are input into the doubly-fed induction generator controller to control the doubly-fed induction generator unit to output according to the desired power.

[0011] Furthermore, in step S2, solving the wind speed clustering center based on the modulus C-means algorithm specifically includes:

[0012] Construct the objective function of the modulus C-means clustering algorithm:

[0013]

[0014] k is the number of wind speed clustering centers, n is the number of samples of the daily average wind speed of the wind farm, m is the weight index affecting the membership matrix, d ij = ||V ci - v j || is the distance between the jth daily average wind speed and the ith clustering center, V ci is the ith clustering center, and v j is the jth daily average wind speed;

[0015] Solve formula (1) to obtain the wind speed clustering center V ci and the membership degree u ij :

[0016]

[0017]

[0018] Randomly generate a set of wind speed clustering centers and membership matrices, and perform iterative calculations on formulas (2) and (3) until convergence to obtain the final wind speed clustering center V C = [V C1 , V C2 , …, V Ck .

[0019] Furthermore, construct a TS fuzzy model based on the wind speed clustering center, and solve the feedback matrix coefficient K of the doubly-fed induction generator controller based on the TS fuzzy model i Specifically include:

[0020] Construct the state equation of the regional new energy power system:

[0021]

[0022] Among them, u = P w ;

[0023] Among them: A, B, and C are the state matrices of the system. x is the system state variable, u = P w is the system input variable, P w is the active power of the doubly-fed induction generator, and y is the system output variable;

[0024] Taking formula (4) as the premise variable of formula (5), the state equation of the system is fuzzified as:

[0025]

[0026] Among them: Among them, A1 to A4, B1 to B4, and C1 to C4 are the state matrices of the system when the antecedent variable v is respectively V c1 to V c4 in the case;

[0027] Performing singleton fuzzification and weighted averaging on formula (6), the TS fuzzy model of the new energy power system is obtained:

[0028]

[0029] Among them:

[0030] A m , B m , and C m are the state matrices of the system under the fuzzy set corresponding to the m-th wind speed clustering center, h m (v) is the weighted average coefficient of the system, and C m is a constant matrix, μ ml (v l ) is the membership degree of the l-th wind speed data to the m-th wind speed clustering center, and σ is the width of the membership function μ ml (·);

[0031] Constructing a doubly-fed induction generator controller using the same antecedent variable as formula (7):

[0032] K m is the feedback matrix of the controller under the fuzzy set corresponding to the m-th wind speed clustering center, and equating formula (7) to formula (9):

[0033]

[0034] Substituting formula (8) into formula (9) to obtain a closed-loop system, and the expression of the closed-loop system is:

[0035] Analyze the stability of the closed-loop system (10) using Lyapunov stability theorem:

[0036] Let Z ml = A m - B m K l , where m < l such that Substitute it into formula (10) to get:

[0037]

[0038] According to Lyapunov stability theorem, if the equilibrium point of the continuous fuzzy control system described by system (11) is globally asymptotically stable, then there exists a common positive definite matrix P such that the following linear matrix inequality holds:

[0039]

[0040] where: Ξ 1m = A m Q - B m M m ;

[0041] Q = P -1 ; M m = K m Q;

[0042] Solve the inequality group (12) to obtain the matrices P and M that satisfy the inequality m , and obtain the feedback matrix coefficient K through K m = M m P m .

[0043] Furthermore, determine the coefficient A through the following method i

[0044]

[0045] where:,, U A represents the phase voltage of bus A; U B represents the phase voltage of bus B; X At and X Bt represent equivalent reactances; X L is the equivalent line reactance; X D is the dynamic reactance of the wind farm reactive power.

[0046] Furthermore, determine the coefficients B i and C i :

[0047]

[0048] M A and M B are the inertia time constants of Region A and Region B, and P A represents the total power generation of Region A; P B represents the total power consumption of Region B.

[0049] Advantages of the present invention: Through the present invention, by considering the influence of wind speed factors, performing wind speed clustering and the TS fuzzy algorithm, the double-fed induction generator controller and the feedback matrix coefficients are determined, enabling the double-fed induction generator to output according to the desired power, thereby effectively suppressing the low-frequency oscillation of the new energy power grid, without the need to use other external devices, effectively reducing costs, and the whole process is relatively more concise and reliable than the traditional technology. Description of the Drawings

[0050] The present invention will be further described below in conjunction with the drawings and embodiments:

[0051] Figure 1 is the topology diagram of the new energy power grid of the present invention.

[0052] Figure 2 is the membership function of the TS fuzzy model based on the FCMA clustering wind speed center of the present invention. Detailed Embodiment

[0053] The following further elaborates on the present invention in detail:

[0054] A method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model provided by the present invention includes the following steps:

[0055] S1. Determine the regional new energy power system model, which includes Region A, Region B, and a wind farm. Among them, Region A and Region B each have two synchronous generators. The electric energy generated by Region A and Region B is respectively input to Bus A and Bus B, and Bus A and Bus B are electrically connected. The wind farm has a double-fed induction generator set, and the electric energy generated by the double-fed induction generator set is input to Bus A;

[0056] S2. Collect the wind speed data of the wind farm, and construct a wind speed data set V = {v1, v2,..., v n} and a membership matrix U = [u ij k×n , and solve the wind speed clustering center based on the fuzzy C-means clustering algorithm; where u ij represents the membership degree of the jth daily average wind speed to the ith wind speed clustering center;

[0057] ​S3. Construct a TS fuzzy model based on the wind speed clustering centers, and solve the feedback matrix coefficient K of the doubly-fed induction generator controller based on the TS fuzzy model m , and use the feedback matrix coefficient K m , the phase angle difference δ between area A and area B AB , and the angular velocity difference ω between area A and the inertia center AB as input values into the doubly-fed induction generator controller to control the doubly-fed induction generator unit to output power according to the desired power; through the above method, by considering the influence of wind speed factors, performing wind speed clustering and the TS fuzzy algorithm, the doubly-fed induction generator controller and the feedback matrix coefficient are determined, enabling the doubly-fed induction generator to output power according to the desired power, thereby effectively suppressing the low-frequency oscillation of the new energy power grid, without the need to use other external devices, effectively reducing costs, and the entire process is relatively simpler and more reliable than traditional technologies

[0058] In this embodiment, in step S2, specifically including solving the wind speed clustering centers based on the fuzzy C-means algorithm

[0059] Construct the objective function of the fuzzy C-means clustering algorithm

[0060]

[0061] where k is the number of wind speed clustering centers, n is the number of samples of the daily average wind speed of the wind farm, m is the weight index affecting the membership matrix, and d ij = ||V ci - v j || is the distance between the j-th daily average wind speed and the i-th clustering center, V ci is the i-th clustering center, and v j is the j-th daily average wind speed

[0062] Solve formula (1) to obtain the wind speed clustering centers V ci and the membership degrees u ij :

[0063]

[0064]

[0065] Randomly generate a set of wind speed clustering centers and membership matrices, and perform iterative calculations on formulas (2) and (3) until convergence to obtain the final wind speed clustering centers V C = [V C1 , V C2 , …, V Ck .

[0066] In this embodiment, a TS fuzzy model is constructed based on the wind speed clustering center, and the feedback matrix coefficient K of the doubly-fed induction generator controller is solved based on the TS fuzzy model. i Specifically, it includes:

[0067] Construct the state equation of the regional new energy power system:

[0068]

[0069] Wherein, u = P w ;

[0070] Where: A, B, and C are the state matrices of the system. x is the system state variable, u = P w is the system input variable, P w is the active power of the doubly-fed induction generator, and y is the system output variable;

[0071] Taking formula (4) as the premise variable of formula (5), the state equation of the system is fuzzified as:

[0072]

[0073]

[0074] Where: Among them, A1 to A4, B1 to B4, and C1 to C4 are the state matrices of the system when the antecedent variable v is respectively V c1 to V c4 ;

[0075] Performing singleton fuzzification and weighted averaging on formula (6) to obtain the TS fuzzy model of the new energy power system:

[0076]

[0077] Wherein:

[0078] A m , B m and C m are the state matrices of the system under the fuzzy set corresponding to the mth wind speed clustering center, h m (v) is the weighted average coefficient of the system, and C m is a constant matrix, μ ml (v l ) is the membership degree of the lth wind speed data to the mth wind speed clustering center, and σ is the width of the membership function μ ml (·); the above fuzzification process and singleton fuzzification process both adopt existing technologies, and their processes and principles will not be elaborated here;

[0079] Construct a doubly-fed induction generator controller using the same antecedent variables as in formula (7):

[0080] K m is the feedback matrix of the controller under the fuzzy set corresponding to the m-th wind speed clustering center. Equate formula (7) to formula (9):

[0081]

[0082] Substitute formula (8) into formula (9) to obtain the closed-loop system. The expression of the closed-loop system is:

[0083] Analyze the stability of the closed-loop system (10) using Lyapunov stability theorem:

[0084] Let Z ml = A m - B m K l , m < l s.t. Substitute it into formula (10) to get:

[0085]

[0086] According to Lyapunov stability theorem, if the equilibrium point of the continuous fuzzy control system described by system (11) is globally asymptotically stable, then there exists a common positive definite matrix P such that the following linear matrix inequality holds:

[0087]

[0088] where: Ξ 1m = A m Q - B m M m ;

[0089] Q = P -1 ; M m = K m Q;

[0090] Solve the inequality group (12) to obtain the matrices P and M m , that satisfy the inequality. Through K m = M m P, obtain the feedback matrix coefficient K m . Among them, the process of solving the inequality can adopt the existing technology.

[0091] In this embodiment, determine the coefficient A through the following method i

[0092]

[0093] Among them: U A represents the phase voltage of busbar A; U B represents the phase voltage of busbar B; X At and X Bt represent equivalent reactance; X L is the equivalent line reactance; X D is the dynamic reactance of the reactive power of the wind farm.

[0094] In this embodiment, the coefficients B i and C i are determined by the following method:

[0095]

[0096] M A and M B are the inertia time constants of Region A and Region B, P A represents the total power generation of Region A; P B represents the total power consumption of Region B.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model, characterized by: The following steps are involved: S1. Determine a regional new energy power system model, the model includes region A, region B and a wind farm, wherein region A and region B respectively have two synchronous generators, the electric energy generated by region A and region B are respectively input to bus A and bus B, and bus A and bus B are electrically connected, and the wind farm has a double-fed induction generator set, and the electric energy generated by the double-fed induction generator set is input to bus A; S2. Collect wind speed data of wind farms and construct a wind speed data set V = {v1, v2, ..., v n } and the membership matrix U = [u ij ] k×n , and solve the wind speed cluster center based on the fuzzy C-means clustering algorithm; where u ij It represents the degree of membership between the j-th daily average wind speed and the i-th wind speed cluster center; S3. Construct a TS fuzzy model based on the wind speed cluster center and solve the feedback matrix coefficient K of the double-fed induction generator controller based on the TS fuzzy model m , and the feedback matrix coefficient K m , the phase angle difference δ between region A and region B AB and the angular velocity difference ω between the centers of inertia of region A and region B AB The input value is used in the doubly-fed induction generator controller to control the doubly-fed induction generator unit to output the desired power.

2. The method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model according to claim 1 is characterized in that: In step S2, solving the wind speed cluster center based on the fuzzy C-means clustering algorithm specifically includes: Construct the objective function of the fuzzy C-means clustering algorithm: k is the number of wind speed cluster centers, n is the number of samples of the daily average wind speed of the wind farm, m is the weight index affecting the membership matrix, d ij =||V ci -v j || is the distance between the jth daily average wind speed and the ith cluster center, V ci is the i-th cluster center, v j is the average wind speed on the jth day; Solve formula (1) to obtain the wind speed cluster center V ci and membership u ij : A set of wind speed cluster centers and membership matrices are randomly generated, and formulas (2) and (3) are iterated until convergence to obtain the final wind speed cluster center V C =[V C1 ,V C2 ,…,V Ck ](4), where the number of wind speed cluster centers is set to K = 4, the weight coefficient m is set to 2, and the maximum number of iterations is 10 6 , stop iteration threshold ε=10 -6 .

3. The method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model according to claim 2 is characterized in that: The TS fuzzy model is constructed based on the wind speed cluster center, and the feedback matrix coefficient K of the double-fed induction generator controller is solved based on the TS fuzzy model. i Specifically include: Construct the state equation of the regional new energy power system: in, u=P w ; Where: A, B and C are the state matrices of the system, x is the system state variable, u = P w is the system input variable, P w is the active power of the doubly-fed induction generator, y is the system output variable; Taking formula (4) as the premise variable of formula (5), the state equation of the system is fuzzified as follows: R 1 :If v is V c1 ,So R 2 :If v is V c2 ,So R 3 :If v is V c3 ,So R 4 :If v is V c4 ,So Among them: A1 to A4, B1 to B4 and C1 to C4 are the system variables v respectively. c1 To V c4 The state matrix of the system in this case; Formula (6) is processed by single-point fuzzification and weighted averaging to obtain the TS fuzzy model of the new energy power system: in: A m , B m and C m is the system state matrix under the fuzzy set corresponding to the mth wind speed cluster center, h m (v) is the weighted average coefficient of the system, and C m is a constant matrix, μ ml (v l ) is the membership of the lth wind speed data to the mth wind speed cluster center, σ is the membership function μ ml (·) width; The doubly-fed induction generator controller is constructed using the same premise variables as in formula (7): K m is the feedback matrix of the controller under the fuzzy set corresponding to the mth wind speed cluster center, and formula (7) is equivalent to formula (9): Substituting formula (8) into formula (9) yields a closed-loop system. The closed-loop system expression is: The stability of the closed-loop system (10) is analyzed using Lyapunov's stability theorem: Z ml =A m -B m K l ,m<lst Substituting into formula (10) we get: From the Lyapunov stability theorem, we know that if the equilibrium point of the continuous fuzzy control system described by system (11) is globally asymptotically stable, then there exists a common positive definite matrix P such that the following linear matrix inequality holds: Among them: X 1m =A m QB m M m ; Q=P -1 ;M m =K m Q; Solve the inequality group (12) to obtain the matrices P and M that satisfy the inequality m , through K m =M m P gets the feedback matrix coefficient K m .

4. The method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model according to claim 1 is characterized in that: The coefficient A is determined by the following method i Among them: A Indicates the phase voltage of bus A; U B Indicates the phase voltage of bus B; X At and X Bt Represents equivalent reactance; X L is the equivalent line reactance; X D is the dynamic reactance of the wind farm reactive power.

5. The method for suppressing low-frequency oscillation of a new energy power grid based on FCM wind speed clustering and TS fuzzy model according to claim 1 is characterized in that: The coefficient B is determined by the following method i and C i : M A and M B is the inertia time constant of area A and area B, P A represents the total power generation in area A; P B Indicates the total power consumption of area B.

Images