A state space-based multi-circuit direct current flow controller modeling and control parameter optimization method

Abstract

The application provides a state space-based multi-line DC power flow controller modeling and control parameter optimization method, the instability of the DC network ring network system caused by the addition of the DC power flow control is made up by adding the matrix E containing PI control, and the influence of the phase margin and the dynamic characteristic is considered, the game optimization of the control parameters in the dynamic time domain model is carried out based on the eigenvalue, and the application has high universality and strong portability, and has high guiding significance for improving the control performance and optimization ability of the line-to-line power flow controller. The application models the multi-line DC power flow controller, breaks through the barrier of the scale expansion of the DC power flow controller, has the advantages of simple structure, high universality, strong applicability and the like, and provides a model basis for the subsequent research of the DC power flow controller.

Description

Technical Field

[0001] This invention belongs to the field of power electronics technology, and in particular relates to a method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller. Background Technology

[0002] With the large-scale integration of distributed energy resources and energy storage into the power grid, traditional AC power grids face significant challenges in terms of power supply stability, economic efficiency, and the ability to accommodate high-penetration distributed power sources. As power electronics technology matures, DC transmission technology has become a crucial direction for power grid development. The development of high-voltage DC transmission technology has led to the evolution from point-to-point HVDC transmission systems to today's multi-terminal DC transmission networks. Similar to the development of traditional AC power grids, because ring networks significantly increase line utilization, enhance line redundancy, and improve power transmission reliability compared to radial networks, the topology of DC power grids will inevitably evolve from radial to ring structures.

[0003] However, for ring DC networks, although the input / output power flow at the outlet of each converter station can be completely determined through effective control of the converter station, the power flow of the ring network lines is not completely controlled. Without additional control devices, the distribution of line current is determined by the network port voltage and the distribution characteristics of line impedance. Therefore, the line current in the ring network may experience overload in some lines and insufficient utilization in the rest of the lines.

[0004] Therefore, in multi-terminal DC ring networks, the method of using an external DC current flow controller (DC CFC) is usually considered to achieve effective control of the line current. The control of the line current is achieved by changing the DC line resistance or the DC port voltage. Moreover, the line-to-line DC CFC does not require an external power supply and has advantages such as low cost and low loss, making it more suitable for the future development of DC power transmission.

[0005] However, the addition of a DC power flow controller may cause instability in DC ring network systems, posing a significant risk to the entire system. Commonly used steady-state models are far from sufficient to reveal the dynamic performance and stability of DC ring network systems. Dynamic stability refers to the ability of a synchronously operating power system to recover to its original operating state or transition to a new operating state after being subjected to minor, transient disturbances. Therefore, it is essential to establish a dynamic model of the DC ring network system, namely a multi-line DC power flow controller model, to analyze its stability. Simultaneously, improving the stability of the DC ring network system after the addition of a DC power flow controller is also a pressing issue that needs to be addressed. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method for modeling and optimizing control parameters of a multi-line DC power flow controller based on state space. This method overcomes the deficiencies of steady-state models in revealing the dynamic performance and stability of the system, and reduces the system instability caused by the addition of the DC power flow controller to the DC ring network system.

[0007] The present invention achieves the above-mentioned technical objectives through the following technical means.

[0008] A method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller includes the following steps:

[0009] Step 1: Construct a k-line DC CFC topology model and determine the operating mode of the common capacitor under different switching modes. Here, a k-line DC CFC consists of k DC-DC converters connected in parallel with a common capacitor C. cfc The energy exchanger consists of k DC-DC converters: SM-1, SM-2, ..., SM-k, where k is a positive integer;

[0010] Step 2: Based on the DC CFC topology model and operating mode in Step 1, establish a state-space based dynamic time-domain model of k-line DC CFC.

[0011] Step 3: Based on the k-line DC CFC dynamic time-domain model established in Step 2, determine the open-loop transfer function of the open-loop control system, add a matrix E containing PI control, and consider the influence of phase margin and dynamic characteristics. Design an eigenvalue-based game optimization method for control parameters to optimize the configuration of PI control parameters in matrix E, thereby compensating for the instability of the DC ring network system caused by the addition of DC power flow control.

[0012] Furthermore, the specific process of step 3 is as follows:

[0013] First, based on the k-line DC CFC dynamic time-domain model established in step 2, the open-loop transfer function of the open-loop control system is determined as follows:

[0014]

[0015] in, Indicates Δu c The inverse matrix, Δu c Δy represents the small perturbation of the control variable; gc_op represents the small perturbation of the output variable. 11 gc_op 1k gc_op k1 gc_op kkThese represent the data in the 1st row and 1st column, the 1st row and kth column, the kth row and 1st column, and the kth row and kth column of the k*k matrix, respectively.

[0016] When inputting Δu c Previously, a matrix E containing PI control was added;

[0017] Introducing a reference value for the input signal, then Δu c =Δe r ·E=(Δy ref -Δy)·E, where Δe r Δy represents the difference between the input signal reference value and the measured value. ref Indicates the reference value of the input signal;

[0018] The PI control in matrix E is divided into a two-layer control strategy, specifically a cooperative game strategy between branch current and common capacitor voltage and a non-cooperative game strategy between branch currents; then, the improved control variable method is used to optimize the PI control parameters.

[0019] Furthermore, the cooperative game strategy between the branch current and the common capacitor voltage is specifically as follows:

[0020] The control of SM-k is achieved by utilizing the combined effect of the branch current and the common capacitor voltage.

[0021] (Δi SM-kref -Δi SM-kpre )·D k +(Δu cref -Δu cpre )·(1-D k )=Δm k

[0022] Where, Δi SM-kref , Δi SM-kpre These represent the reference and measured values ​​of the current in the branch where SM-k is located, respectively; Δu cref , Δu cpre These represent the reference value and measured value of the common capacitor voltage, respectively; D k For coordination factor; Δm k This is the trigger signal for SM-k.

[0023] Furthermore, the specific implementation method of the non-cooperative game strategy among the branch currents is as follows:

[0024] S1: Based on the natural distribution of the DC ring network, calculate the branch current where SM-k is located and set it as the initial branch current reference value i. SM-kref ;

[0025] S2: With the DC CFC of line k in operation, and aiming at stabilizing the common capacitor voltage, the reference value of the branch current in SM-k is measured to obtain the current optimal reference value I′. SM-kref ;

[0026] S3: Comparison with i SM-kref and I′ SM-kref If the two are equal, the process ends, and the optimal branch current reference value is I′. SM-kref If the two are not equal, then use I′. SM-kref Replace i SM-kref And repeat S2.

[0027] Furthermore, the specific process of optimizing the PI control parameters using the improved control variable method is as follows:

[0028] S3.1: Calculate the optimal branch current reference value based on the non-cooperative game strategy among branch currents, and calculate the initial common capacitor voltage reference value based on the DC network power flow distribution and common capacitor capacity.

[0029] S3.2: Based on the cooperative game strategy between branch current and common capacitor voltage, select the coordination factor to further determine the trigger signal of each SM-k of DC CFC in line k;

[0030] S3.3: Initialize PI control parameters, let K p1 =……=K p(k-1) =K pk =0,

[0031] K i1 =K i1 =……=K ik =50;

[0032] S3.4: Calculate the eigenvalues ​​of the transfer function after adding matrix E;

[0033] S3.5: Determine if all eigenvalues ​​are on the left half-plane of s. If so, the DC ring network system is stable, proceed to S3.6; otherwise, let K... p1 ... K p(k-1) K pk Both increased by 0.5, K i1 ... K i(k-1) K ik All are increased by 50 and returned to S3.4; where s represents the Laplace transform;

[0034] S3.6: Calculate the phase margin of the branch current and capacitor voltage of DC CF after adding matrix E;

[0035] S3.7: Determine if the phase margin is greater than 70°. If so, output the optimal value for the PI control parameter; otherwise, set K... p1 ... K p(k-1) K pk Both increased by 0.5, K i1 ... K i(k-1) K ik Increase all by 50 and return to S3.4.

[0036] Furthermore, the expression for the matrix E is as follows:

[0037]

[0038] The PI control parameter in matrix E is: K p1 =……=K p(k-1) =K pk =1,K i1 =……=K i(k-1) =K ik =50; E 11 E 1k E k1 E kk These represent the data in the 1st row and 1st column, the 1st row and kth column, the kth row and 1st column, and the kth row and kth column of matrix E, respectively; PI1 represents E 11 PI parameters; PI k-1 E represents (k-1)(k-1) The PI parameter, E (k-1)(k-1) PI represents the data in the (k-1)th row and (k-1)th column of matrix E; k E represents kk The PI parameter; i SM-1ref Indicates the reference value of the branch current where SM-1 is located; i SM-(k-1)ref This represents the reference value of the branch current where SM-(k-1) is located; u cref Indicates the reference value of the common capacitor voltage; s represents the Laplace transform;

[0039]

[0040] Where, Δi SM-(k-1)ref , Δi SM-(k-1) These represent the reference and measured values ​​of the current in the branch containing SM-(k-1), respectively; Δi SM-1ref , Δi SM-1 These represent the reference and measured values ​​of the current in the branch where SM-1 is located, respectively; Δu cref , Δu cpre These represent the reference value and the measured value of the common capacitor voltage, respectively.

[0041] Furthermore, the specific process of step 2 is as follows:

[0042] Step 2.1: Based on the k-line DC CFC topology model and the operating mode of the common capacitor in Step 1, establish the k-line DC CFC mathematical model:

[0043] By charging and discharging the common capacitor in the DC CFC circuit of line k, we can obtain:

[0044]

[0045] Right now:

[0046]

[0047] Among them, u c Indicates the common capacitor voltage; t represents time; i cfc Indicates the value of the common capacitor current; m i Indicates the switch mode, where i represents the mode number, i = 1, ..., 4; SM-1 i SM-2 i SM-k These represent the branch currents of SM-1, SM-2, and SM-k, respectively; d c1 d c2 d ck S represents C1 S C2 S CK Duty cycle, S C1 S C2 S CK These represent the controllable switching devices IGBTs in SM-1, SM-2, and SM-k, respectively; d a S represents AK Duty cycle, S AK For the controllable switching device IGBT in SM-k;

[0048] According to Kirchhoff's current and voltage laws, the mathematical model of DC CFC for line k is as follows:

[0049]

[0050] Among them, R SM-k L SM-k These are the resistance and inductance of the branch containing SM-k, respectively; v N1 v N(k+1 ) represent the voltages at both ends of the branch where SM-k is located;

[0051] Step 2.2: Establish a state-space based steady-state time-domain model of the k-line DC CFC;

[0052] Step 2.3: Linearize the k-line DC CFC mathematical model established in Step 2.1 to obtain the small-signal model. Based on the steady-state time-domain model established in Step 2.2, rewrite the small-signal model into the form of state-space equations.

[0053] Furthermore, the specific process of step 2.2 is as follows:

[0054] The k-line DC CFC mathematical model in step 2.1 is rewritten in the form of the state-space equations shown below:

[0055]

[0056] in, Let x represent the differential component of x; x represent the state vector; u represent the control variable; y represent the output variable; and A, B, C, and D all represent state matrices.

[0057] Furthermore, the specific process of step 2.3 is as follows:

[0058] Step 2.3.1: Linearize the k-line DC CFC mathematical model established in Step 2.1 to obtain the small-signal model shown below:

[0059]

[0060] Where, Δi SM-1 , Δi SM-2 , Δi SM-k These represent the small current disturbances in the branches containing SM-1, SM-2, and SM-k, respectively; Δv N1 Δv N(k+1) These represent the small voltage disturbances at both ends of the branch containing SM-k; Δu c Δd represents the small disturbance in the voltage of the common capacitor. c1 , Δd c2 , Δd ck S represents C1 S C2 S CK The duty cycle has a small disturbance amount;

[0061] Step 2.3.2: Rewrite the small-signal model from Step 2.3.1 into the state-space equation form shown below:

[0062]

[0063] The input variables, output variables, and state variables are as follows:

[0064] Δx=[Δi SM-1 Δi SM-2 … ΔiSM-k Δv N1 Δv N2 … Δv Nk Δu c ] T

[0065] Δy=[Δi SM-1 Δi SM-2 … Δi SM-k Δu c ] T

[0066] Δu c =[Δd c1 Δd c2 … Δd ck ] T

[0067] Δu d =[Δi SM-1 Δi SM-2 … Δi SM-k Δv N(k+1) ] T

[0068] in, Δx represents the differential component of Δx; Δx represents the small perturbation of the state vector; Δu c , Δu d Both represent small disturbances in the control variable; Δy represents small disturbances in the output variable; A2, B c B d C2, D c D d Both represent state matrices; Δv N2 This represents the small voltage disturbance at the other end of the branch containing SM-1; Δv Nk It represents the small voltage disturbance at the other end of the branch containing SM-(k-1); T represents the matrix transpose.

[0069] Furthermore, in step 1, under different switching modes, the common capacitor has three operating states: bypass, charging, and discharging; the k-line DC CFC is equivalent to k controllable voltage sources connected in series on k DC branches; for SM-k, its average output voltage e within one control cycle SM-k as follows:

[0070]

[0071] Among them, i SM-k Indicates the current in the branch containing SM-k; m iThe represents the switching mode, where i represents the mode number, i = 1, ..., 4; where m1 represents bypass mode, m2 represents charging mode, m3 represents discharging mode, and m4 represents bypass mode; d a d ck S represents AK and S CK duty cycle; u c This indicates the common capacitor voltage.

[0072] The present invention has the following beneficial effects:

[0073] This invention compensates for the instability of DC network ring systems caused by the addition of DC power flow control by incorporating a matrix E containing PI control. Considering the effects of phase margin and dynamic characteristics, it performs game-theoretic optimization of control parameters in the dynamic time-domain model based on eigenvalues. This approach exhibits high versatility and strong portability, providing significant guidance for improving the control performance and optimization capabilities of inter-line power flow controllers. Furthermore, this invention models multi-line DC power flow controllers, breaking through the barriers to scaling up DC power flow controllers. It boasts advantages such as simple structure, high versatility, and strong applicability, providing a model foundation for subsequent research on DC power flow controllers. Attached Figure Description

[0074] Figure 1 This is a topology diagram of the DC power flow controller described in this invention;

[0075] Figure 2 This is a flowchart of the control parameter optimization method described in this invention;

[0076] Figure 3 Optimize the flowchart for control parameters. Detailed Implementation

[0077] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the scope of protection of the present invention is not limited thereto.

[0078] The flowchart of the state-space based multi-line DC power flow controller modeling and control parameter optimization method described in this invention is as follows: Figure 2 As shown, the specific steps include the following:

[0079] Step 1: Build a... Figure 1 The k-line DC-DC power flow controller (DC CFC) topology shown is composed of k full-bridge DC-DC converters (SM-1, SM-2, ..., SM-k) connected in parallel with a common capacitor C. cfc The energy exchanger is composed of four controllable switching devices (IGBTs) in each DC-DC converter. The four controllable switching devices (IGBTs) are divided into left and right arms, and the two controllable switching devices (IGBTs) on the same arm form complementary conduction.

[0080] Figure 1 In the diagram, SM-1 represents the first DC-DC converter, SM-2 represents the second DC-DC converter, and SM-k represents the kth DC-DC converter; S A1 S B1 S C1 S D1 This refers to the four controllable switching devices IGBTs in SM-1, S A2 S B2 S C2 S D2 This refers to the four controllable switching devices IGBTs in SM-2, S AK S BK S CK S DK N represents the four controllable switching devices IGBTs in SM-k; N1 represents the common port of SM-1, SM-2, ..., SM-k; N2 represents another port of SM-1; N3 represents another port of SM-2; N... k+1 This indicates another port of SM-k.

[0081] Under different switching modes, the common capacitor has three operating states: bypass, charging, and discharging; the k-line DC CFC can be equivalent to k controllable voltage sources connected in series on k DC branches; for SM-k, its average output voltage e within one control cycle SM-K (where k is a positive integer) is as follows:

[0082]

[0083] Among them, i SM-k Indicates the current in the branch containing SM-k; m i (i represents the mode number, i = 1, ..., 4) represents the switching mode, where m1 represents the bypass mode, m2 represents the charging mode, m3 represents the discharging mode, and m4 represents the bypass mode; d a d ck S represents AK and S CK The duty cycle, and d a Take 0.5; u c Indicates the common capacitor voltage; e SM-k From duty cycle d ck control.

[0084] Step 2: Establish a state-space based k-line DC CFC dynamic time-domain model, as follows:

[0085] Step 2.1: Based on the DC CFC topology model and operating mode conditions in Step 1, establish the DC CFC mathematical model for the k-line:

[0086] By charging and discharging the common capacitor in the DC CFC circuit of line k, we can obtain:

[0087]

[0088] Right now:

[0089]

[0090] Where, d c1 S represents C1 duty cycle, i SM-1 d represents the current in the branch containing SM-1. c2 S represents C2 duty cycle, i SM-2 This indicates the current in the branch containing SM-2;

[0091] According to Kirchhoff's current and voltage laws, the mathematical model of DC CFC for line k is as follows:

[0092]

[0093] Among them, R SM-k L SM-k These represent the resistance and inductance of the branch containing SM-k, respectively; t represents time; v N1 v N(k+1) These are the voltages at both ends of the branch where SM-k is located;

[0094] Step 2.2: Establish a state-space based k-line DC CFC steady-state time-domain model:

[0095] The k-line DC CFC mathematical model in step 2.1 is rewritten in the form of the state-space equations shown below:

[0096]

[0097] in, Let x represent the differential component of x; x represents the state vector, including i SM-k v N(k+1) u c ; u represents the control variable, including the known port current and port voltage values; y represents the output variable, including i SM-k u c A, B, C, and D all represent state matrices;

[0098] Step 2.3: Establish a state-space based dynamic time-domain model of k-line DC CFC:

[0099] Step 2.3.1: Linearize the k-line DC CFC mathematical model established in Step 2.1 to obtain the small-signal model shown below:

[0100]

[0101] Where, Δi SM-k Indicates SM- k The small current disturbance in the branch; Δv N1 Δv N(k+1) These represent the small voltage disturbances at both ends of the branch containing SM-k; Δu c Δd represents the small disturbance in the voltage of the common capacitor. ck S represents CK Duty cycle small disturbance; Δd c1 S represents C1 Duty cycle small disturbance; Δd c2 S represents C2 The duty cycle has a small disturbance amount;

[0102] Step 2.3.2: Based on the steady-state time-domain model established in Step 2.2, the small-signal model in Step 2.3.1 is rewritten into the state-space equation form shown below:

[0103]

[0104] The input variables, output variables, and state variables are as follows:

[0105] Δx=[Δi SM-1 Δi SM-2 … Δi SM-k Δv N1 Δv N2 … Δv Nk Δu c ] T

[0106] Δy=[Δi SM-1 Δi SM-2 … Δi SM-k Δu c ] T

[0107] Δu c =[Δd c1 Δd c2 … Δd ck ] T

[0108] Δu d =[Δi SM-1 Δi SM-2 … ΔiSM-k Δv N(k+1) ] T

[0109] in, Δx represents the differential component of Δx; Δx represents the small perturbation of the state vector; Δu c , Δu d Both represent small disturbances in the control variable; Δy represents small disturbances in the output variable; A2, B c B d C2, D c D d Both represent state matrices; Δi SM-1 This represents the small current disturbance in the branch containing SM-1; Δi SM-2 This represents the small current disturbance in the branch containing SM-2; Δv N2 This represents the small voltage disturbance at the other end of the branch containing SM-1; Δv Nk It represents the small voltage disturbance at the other end of the branch containing SM-(k-1); T represents the matrix transpose.

[0110] Step 3: Considering the influence of phase margin and dynamic characteristics, a game-theoretic optimization method for control parameters based on eigenvalues ​​is designed to optimize the configuration of PI control parameters and compensate for the instability of the DC ring network system caused by the addition of DC power flow control. The specific process is as follows:

[0111] Based on the k-line DC CFC dynamic time-domain model in step 2, regarding Δu c The open-loop transfer function of the open-loop control system is:

[0112]

[0113] Among them, gc_op 11 This represents the data in the first row and first column of a k*k matrix; gc_op 1k This represents the data in the 1st row and kth column of a k*k matrix; gc_op k1 This represents the data in the k-th row and 1st column of a k*k matrix; gc_op kk This represents the data in the k-th row and k-th column of a k*k matrix; Indicates Δu c The inverse matrix;

[0114] By giving Δu c A phase increment is provided to improve the stability of the open-loop control system, i.e., at the input Δu c Previously, a matrix E containing PI control was added. The expression for matrix E is as follows:

[0115]

[0116] The PI control parameter in matrix E is: K p1 =……=K p(k-1) =Kpk=1,K i1 =……=K i(k-1) =K ik =50; E 11 This represents the data in the first row and first column of a k*k matrix; E 1k This represents the data in the 1st row and kth column of a k*k matrix; E k1 This represents the data in the k-th row and 1st column of a k*k matrix; E kk PI1 represents the data in the k-th row and k-th column of a k*k matrix; PI1 represents E 11 PI parameters; PI k-1 E represents (k-1)(k-1) The PI parameter, E (k-1)(k-1) PI represents the data in the (k-1)th row and (k-1)th column of a k*k matrix. k E represents kk The PI parameter; i SM-1ref Indicates the reference value of the branch current where SM-1 is located; i SM-(k-1)ref This represents the reference value of the branch current where SM-(k-1) is located; u cre f represents the reference value of the common capacitor voltage; s represents the Laplace transform.

[0117] Introducing a reference value for the input signal, then

[0118]

[0119] Where, Δe r Δy represents the difference between the input signal reference value and the measured value. ref Indicates the reference value of the input signal; Δi SM-(k-1)ref , Δi SM-(k-1) These represent the reference and measured values ​​of the current in the branch containing SM-(k-1), respectively; Δi SM-1ref , Δi SM-1 These represent the reference and measured values ​​of the current in the branch where SM-1 is located, respectively; Δu cref , Δu cpre These represent the reference value and the measured value of the common capacitor voltage, respectively.

[0120] Optimize the PI control parameters in matrix E:

[0121] First, the PI control in matrix E is divided into a two-layer control strategy, specifically a cooperative game strategy between branch current and common capacitor voltage and a non-cooperative game strategy between branch currents.

[0122] 1) The cooperative game strategy between branch current and common capacitor voltage is specifically expressed as follows:

[0123] By utilizing the combined effect of the branch current and common capacitor voltage of the SM-k submodule of the k-line DC CFC, the SM-k converter is controlled in a coordinated manner, i.e.:

[0124] (Δi SM-kref -Δi SM-kpre )·D k +(Δu cref -Δu cpre )·(1-D k )=Δm k

[0125] Where, Δi SM-kref Indicates the reference value of the current in the branch where SM-k is located; Δi SM-kpre D represents the measured current value of the branch where SM-k is located; k For coordination factor; Δm k This is the trigger signal for SM-k;

[0126] 2) The non-cooperative game strategy among branch currents, specifically implemented as follows:

[0127] S1: Based on the natural distribution of the DC ring network, calculate the branch current where SM-k is located and set it as the initial branch current reference value i. SM-kref ;

[0128] S2: With the DC CFC of line k in operation, and aiming at stabilizing the common capacitor voltage, the reference value of the branch current in SM-k is measured to obtain the current optimal reference value I′. SM-kref ;

[0129] S3: Comparison with i SM-kref and I′ SM-kref If the two are equal, the process ends, and the optimal branch current reference value is I′. SM-kref If the two are not equal, then use I′. SM-kref Replace i SM-kref , and repeat S2;

[0130] Then as Figure 3 The method shown here is to optimize the PI control parameters using the improved control variable method. The specific steps are as follows:

[0131] S3.1: Calculate the optimal branch current reference value based on the non-cooperative game strategy among branch currents, and calculate the initial common capacitor voltage reference value based on the DC network power flow distribution and common capacitor capacity.

[0132] S3.2: Based on the cooperative game strategy between branch current and common capacitor voltage, select the coordination factor to further determine the trigger signal of each SM-k of DC CFC in line k;

[0133] S3.3: Initialize PI control parameters, let K p1 =……=K p(k-1) =K pk =0,K i1 =K i1 =……=K ik =50;

[0134] S3.4: Calculate the eigenvalues ​​of the transfer function after adding matrix E;

[0135] S3.5: Determine if all eigenvalues ​​are on the left half-plane of s. If so, the DC ring network system is stable, proceed to S3.6; otherwise, let K... p1 ... K p(k-1) K pk Both increased by 0.5, K i1 ... K i(k-1) K ik Increase all by 50 and return to S3.4;

[0136] S3.6: Calculate the phase margin of the branch current and capacitor voltage of DC CF after adding matrix E;

[0137] S3.7: Determine if the phase margin is greater than 70°. If so, output the optimal value for the PI control parameter; otherwise, set K... p1 ... K p(k-1) K pk Both increased by 0.5, K i1 ... K i(k-1) K ik Increase all by 50 and return to S3.4.

[0138] The embodiments described above are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the protection scope of the present invention.

[0139] The embodiments described above are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller, characterized in that, Includes the following steps: Step 1: Build The DC-CFC topology model of the line is used to determine the operating modes of the common capacitor under different switching modes. DC CFC for lines is An energy exchanger consisting of a DC-DC converter connected in parallel with a common capacitor. It is a positive integer; Step 2: Based on the DC CFC topology model and operating modes in Step 1, establish a state-space based model. Line DC CFC dynamic time-domain model; Step 3: Based on the steps established in step 2 The DC CFC dynamic time-domain model of the line is used to determine the open-loop transfer function of the open-loop control system. A matrix E containing PI control is added. Considering the effects of phase margin and dynamic characteristics, a game-theoretic optimization method for control parameters based on eigenvalues ​​is designed to optimize the configuration of PI control parameters in matrix E, thereby compensating for the instability of the DC ring network system caused by the addition of DC power flow control. The specific process of step 3 is as follows: First, based on the steps established in step 2... Based on the DC CFC dynamic time-domain model of the line, the open-loop transfer function of the open-loop control system is determined as follows: ； in, express The inverse matrix, This represents a small disturbance in the control variable; This represents a small perturbation in the output variable; , , , They represent The data in the first row and first column of the matrix, the data in the first row and first column... Column data, the first Data in row 1 and column 2 Line number Column data; In the input Previously, a matrix E containing PI control was added; Introducing a reference value for the input signal, then ,in, This represents the difference between the input signal reference value and the measured value. Indicates the reference value of the input signal; The PI control in matrix E is divided into a two-layer control strategy, specifically a cooperative game strategy between branch current and common capacitor voltage and a non-cooperative game strategy between branch currents; then, the improved control variable method is used to optimize the PI control parameters.

2. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 1, characterized in that, The specific cooperative game strategy between the branch current and the common capacitor voltage is as follows: The current in the branch containing SM-k and the voltage of the common capacitor are used to coordinate and control SM-k, that is: ； in, The DC-DC converters are: SM-1, SM-2, ..., SM- ; , They represent SM- Reference and measured values ​​of the current in the branch circuit; , These represent the reference value and the measured value of the common capacitor voltage, respectively. As a coordinating factor; For SM- The trigger signal.

3. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 2, characterized in that, The specific implementation method of the non-cooperative game strategy among the branch currents is as follows: S1: Calculate SM- based on the natural distribution of the DC ring network. Set the current of the branch to the initial reference value for the branch current. ; S2: With the DC CFC of the line put into operation, and the common capacitor voltage stabilized, SM- was measured. The optimal branch current reference value is obtained by taking the current reference value of the branch. ; S3: Comparison and If the two are equal, the process ends, and the optimal branch current reference value is obtained. If the two are not equal, then use replace And repeat S2.

4. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 3, characterized in that, The specific process of optimizing the PI control parameters using the improved control variable method is as follows: S3.1: Calculate the optimal branch current reference value based on the non-cooperative game strategy among branch currents, and calculate the initial common capacitor voltage reference value based on the DC network power flow distribution and common capacitor capacity. S3.2: Based on the cooperative game strategy between branch current and common capacitor voltage, select the coordination factor to further determine the trigger signal of each SM-k of DC CFC in line k; S3.3: Initialize the PI control parameters, let , ; S3.4: Calculate the eigenvalues ​​of the transfer function after adding matrix E; S3.5: Determine if all eigenvalues ​​are in If so in the left half-plane, it indicates that the DC loop network system is stable, and proceed to S3.6; otherwise, let... ... , All increased by 0.

5. ... , All are increased by 50, and returned to S3.4; among them, Represents the Laplace transform; S3.6: Calculate the phase margin of the branch current and capacitor voltage of DC CF after adding matrix E; S3.7: Determine if the phase margin is greater than 70º. If so, output the optimal value for the PI control parameter; otherwise, set... ... , All increased by 0.

5. ... , Increase all by 50 and return to S3.

4.

5. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 1, characterized in that, The expression for matrix E is as follows: ； in, The DC-DC converters are: SM-1, SM-2, ..., SM- The PI control parameters in matrix E are: , ; , , , These represent the data in the first row and first column of matrix E, respectively. Column data, the first Data in row 1 and column 2 Line number Column data; express The PI parameter; express PI parameter, Represents the first in matrix E Line number Column data; express PI parameters; This indicates the reference value of the current in the branch where SM-1 is located; This represents the reference value of the branch current where SM-(k-1) is located; Indicates the reference value for the common capacitor voltage; Represents the Laplace transform; ； in, , These represent the reference value and the measured value of the current in the branch where SM-(k-1) is located, respectively; , These represent the reference value and the measured value of the current in the branch where SM-1 is located, respectively. , These represent the reference value and the measured value of the common capacitor voltage, respectively.

6. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 1, characterized in that, The specific process of step 2 is as follows: Step 2.1: According to Step 1 The DC-CFC topology model of the line and the operating modes of the common capacitor are established. Line DC CFC mathematical model: right The charging and discharging of the common capacitor in the DC CFC circuit yields the following: ； Right now: ； in, The DC-DC converters are: SM-1, SM-2, ..., SM- ; Indicates the common capacitor voltage; Indicates time; Indicates the value of the common capacitor current; Indicates the common capacitance of DC CFC; Indicates the switch mode. Indicates the pattern number. ; , , They represent SM-1, SM-2, and SM- respectively. Current in the branch; , , They represent , , duty cycle, , , They represent SM-1, SM-2, and SM- respectively. IGBTs are controllable switching devices in computers. express duty cycle, For SM- IGBTs are controllable switching devices in computers. According to Kirchhoff's current and voltage laws, the mathematical model of DC CFC for line k is as follows: ； in, , SM- The resistance and inductance of the branch circuit; , SM- The voltage at both ends of the branch; Step 2.2: Establish a state-space based system DC-CFC steady-state time-domain model of the line; Step 2.3: For the data established in Step 2.1... The line DC CFC mathematical model is linearized to obtain a small-signal model. Based on the steady-state time-domain model established in step 2.2, the small-signal model is rewritten into the form of state-space equations.

7. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 6, characterized in that, The specific process of step 2.2 is as follows: In step 2.1 The mathematical model of DC-CFC circuit can be rewritten in the form of the state-space equations shown below: ； in, express The trace components; Represents the state vector; Indicates control variables; Indicates the output variable; , , , Both represent state matrices.

8. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 7, characterized in that, The specific process of step 2.3 is as follows: Step 2.3.1: For the data established in Step 2.1 Linearizing the DC-CFC mathematical model of the line yields the small-signal model shown below: ； in, , , These represent the branch where SM-1 is located, the branch where SM-2 is located, and SM- The current disturbance in the branch is small; , They represent SM- Small voltage disturbances at both ends of the branch; This indicates a small disturbance in the voltage of the common capacitor; , , They represent , , The duty cycle has a small disturbance amount; Step 2.3.2: Rewrite the small-signal model from Step 2.3.1 into the state-space equation form shown below: ； The input variables, output variables, and state variables are as follows: ； ； ； ； in, express The trace components; This represents a small perturbation in the state vector; , Both represent small disturbances in the control variables; This represents a small perturbation in the output variable; , , , , , Both represent state matrices; This indicates the small voltage disturbance at the other end of the branch where SM-1 is located; Indicates SM-( -1) Small voltage disturbance at the other end of the branch; This indicates the matrix transpose.

9. The method for modeling and optimizing control parameters of a state-space based multi-line DC power flow controller according to claim 1, characterized in that, In step 1, under different switching modes, the common capacitor has three working states: bypass, charging and discharging. The DC CFC of the line is equivalent to A controllable voltage source is connected in series and installed On the DC branch; for SM- The average output voltage within one control cycle as follows: ； in, The DC-DC converters are: SM-1, SM-2, ..., SM- ; Indicates SM- Current in the branch; Indicates the switch mode. Indicates the pattern number. ;in, Indicates bypass mode. Indicates the charging mode. Indicates the discharge mode. Indicates bypass mode; , They represent and Duty cycle; This indicates the common capacitor voltage.

Images