Parameter optimization method for converter station of wind power plant based on samson swarm algorithm

Abstract

The invention belongs to the field of offshore wind power generation system control strategies, and discloses a wind power plant converter station parameter optimization method based on a samson swarm algorithm, and the method comprises the steps: S1, constructing a structure of an offshore wind power plant converter station; s2, analyzing control systems of the sending-end converter station and the receiving-end converter station, and determining parameters needing to be set in the control systems; and S3, optimizing parameters which need to be set by adopting a salmons algorithm. The offshore wind power plant transmits electric energy to a power utilization side through a high-voltage direct-current power transmission line through a sending-end converter station and a receiving-end converter station; the sending end converter station and the receiving end converter station adopt MMC converters, and a control system comprises a large number of parameters needing to be set, and the parameters can be obtained through a samcat swarm optimization algorithm.

Description

Technical Field

[0001] The invention belongs to the field of offshore wind power generation, and in particular relates to a method for optimizing parameters of a converter station in a wind farm based on a sand cat swarm algorithm. Background Art

[0002] Offshore wind farms have large capacity and long transmission distances, and transmitting the output power of offshore wind farms through high-voltage direct current transmission lines has the advantages of high efficiency, etc. That is, the AC power output by the offshore wind farm is converted into DC power through the sending-end converter station, and then the energy is transmitted through the high-voltage direct current transmission line, and finally the DC power is converted into AC power through the receiving-end converter station.

[0003] Both the sending-end converter station and the receiving-end converter station can be constructed with MMC converters, and their control strategies include at least 49 parameters that need to be set. Once the parameters are set improperly, the performance of the control systems of the sending-end converter station and the receiving-end converter station will be greatly reduced.

[0004] The Sand Cat Swarm Optimization Algorithm is used to optimize the parameters that need to be set in the control systems of the sending-end converter station and the receiving-end converter station.

[0005] Therefore, there is still room for improvement in the prior art. Summary of the invention

[0006] The purpose of the present invention is to provide a method for optimizing the parameters of a wind farm converter station based on a sand cat swarm algorithm; it includes control strategies for a sending-end converter station and a receiving-end converter station, summarizes the parameters that need to be set in the control strategy, and then proposes to optimize these parameters using a sand cat swarm optimization algorithm.

[0007] To achieve the above-mentioned purpose, the technical solution of the present invention is: a method for optimizing the parameters of a wind farm converter station based on a sand cat group algorithm, comprising the following steps: S1: constructing the structure of an offshore wind farm converter station, the structure of the offshore wind farm comprising a sub-wind farm, a 35kV collector, a 220kV bus, a transformer, a sending-end converter station and a receiving-end converter station, the sub-wind farm converges through a 35kV collector, and then is connected to the 220kV bus through a step-up transformer, the 220kV bus is connected to the sending-end converter station to convert AC power into DC power, and is transmitted over long distances through a DC transmission line. When reaching the load side, the DC transmission line is connected to the receiving-end converter station to convert DC power into AC power to supply power to the load;

[0008] S2: Analyze the control systems of the sending-end converter station and the receiving-end converter station, and clarify the parameters that need to be set in the control system;

[0009] S3: Use the Sand Cat algorithm to optimize the parameters that need to be set.

[0010] The sending-end converter station and the receiving-end converter station use MMC composed of half-bridge submodules as converters.

[0011] Furthermore, the circulating current i in the MMC converter in the sending-end converter station Wg_cirj It is expressed as formula (1); Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. Wg_cirj It can also be expressed as formula (2).

[0012]

[0013] In the formula, I Wg_dc is the DC component in the converter bridge arm commutation, I Wg_k is the amplitude of the kth harmonic component in the bridge arm commutation; the harmonic components in the circulating current are basically even-numbered harmonic components; i Wg_pj is the upper arm current of the MMC converter in the sending-end converter station, i Wg_nj is the lower arm current of the MMC converter in the sending-end converter station; the subscript Wg represents the power grid on the wind farm side.

[0014] The quasi-proportional resonant controller can have a large gain at a specified harmonic frequency, and its transfer function is expressed as:

[0015]

[0016] In the formula, ω r is the resonant frequency, ω c is the cut-off frequency, K r is the resonance coefficient of the quasi-proportional resonant controller.

[0017] Taking the suppression of the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current in the control strategy of the sending-end converter station as an example, the circulating current suppression controller needs to include a second-order low-pass filter, a PI controller 1, and three QR controllers QR controller 1, QR controller 2, and QR controller 3 to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current respectively.

[0018] Furthermore, the circulating current i in the MMC converter in the receiving end converter station G_cirj It is expressed as formula (4); Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. The circulating current i of the MMC converter in the receiving end converter station is G_cirj It can also be expressed as formula (5):

[0019]

[0020] In the formula, I G_dcis the DC component in the converter bridge arm commutation, I G_k is the amplitude of the kth harmonic component in the bridge arm commutation; the harmonic components in the circulating current are basically even-numbered harmonic components; i G_pj is the upper arm current of the MMC converter in the sending-end converter station, i G_nj is the lower arm current of the MMC converter in the sending-end converter station.

[0021] Taking the control strategy of the receiving-end converter station to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current as an example, the circulating current suppression controller needs to include a second-order low-pass filter, a PI controller 2, and three QR controllers QR controller 4, QR controller 5, and QR controller 6 to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current respectively.

[0022] Furthermore, the control strategy of the converter of the sending-end converter station is divided into a d-axis voltage control link and a q-axis voltage control link; wherein, the d-axis voltage control link includes an outer-loop PI controller 3, an inner-loop controller 5, and an output limiter of the inner-loop and outer-loop controllers; the q-axis voltage control link includes an outer-loop PI controller 4, an inner-loop controller 6, and an output limiter of the inner-loop and outer-loop controllers. In addition, the input information of the outer-loop controller of the d-axis is the d-axis voltage, and the output information is the d-axis current reference value, which is also the input information of the PI controller 5. The input information of the outer-loop controller of the q-axis is the q-axis voltage, and the output information is the q-axis current reference value, which is also the input information of the PI controller 5.

[0023] The d-axis voltage and q-axis voltage are obtained by coordinate transformation of the three-phase voltages a, b, and c; the d-axis current and q-axis current are obtained by coordinate transformation of the three-phase currents a, b, and c; the coordinate transformation expression can be expressed as:

[0024]

[0025] In the formula, u Wg-A-LP 、u Wg-B-LP 、u Wg-C-LP Represents the low-frequency component of the three-phase grid voltage on the wind farm side; i Wg-A-LP 、i Wg-B-LP 、i Wg-C-LP Represents the low-frequency component of the three-phase grid current on the wind farm side; u Wg-α The α-axis component, u represents the low-frequency component of the three-phase grid voltage on the wind farm side, in the αβ axis. Wg-β represents the β-axis component of the low-frequency component of the three-phase grid voltage on the wind farm side in the αβ axis; i Wg-α The α-axis component of the low-frequency component of the three-phase grid current on the wind farm side is represented by i Wg-βrepresents the β-axis component of the low-frequency component of the three-phase grid current on the wind farm side in the αβ axis; u Wg-d The d-axis component of the low-frequency component of the three-phase grid voltage on the wind farm side is represented by u Wg-q The q-axis component of the low-frequency component of the three-phase grid voltage on the wind farm side, i Wg-d The d-axis component, i represents the low-frequency component of the three-phase grid current on the wind farm side, in the dq axis. Wg-q Represents the q-axis component of the low-frequency component of the three-phase grid current on the wind farm side in the dq axis; θ Wg Represents the voltage vector angle of the power grid on the wind farm side;

[0026] u Wg-A-LP 、u Wg-B-LP 、u Wg-C-LP 、i Wg-A-LP 、i Wg-B-LP 、i Wg-C-LP for u Wg-A 、u Wg-B 、u Wg-C 、i Wg-A-LP 、i Wg-B 、i Wg-C Obtained through low-pass filter 1 and low-pass filter 2.

[0027] Furthermore, the modulation strategy of the MMC converter in the sending-end converter station adopts a nearest level approximation modulation algorithm, which takes into account the capacitor voltage balance of the MMC submodules.

[0028] Furthermore, the control system of the MMC in the receiving-end converter station is divided into a DC bus voltage control loop and a reactive power control loop; wherein the DC bus voltage loop is divided into a voltage outer loop and a current inner loop, the voltage outer loop adopts a PI controller 6, and the current inner loop adopts a PI controller 8; the reactive power loop is divided into a reactive power outer loop and a current inner loop, the reactive power outer loop adopts a PI controller 7, and the current inner loop adopts a PI controller 9.

[0029] The input of the voltage outer loop controller is the collected DC bus voltage u dc and DC bus voltage reference value u dc_ref The output is the reference value i of the three-phase current component on the d-axis of the AC side power grid of the receiving converter station. G_d_ref The inner loop controller is i G_d_ref The difference between the actual value of the three-phase current component on the d-axis of the AC side grid of the receiving converter station. The output is the output three-phase voltage component u of the MMC converter of the receiving converter station on the d-axis. G_MMC_d .

[0030] The input of the reactive power outer loop controller is the reactive power Q G and reactive power reference value QG_ref The difference is the component reference value i of the three-phase current on the q axis of the AC side power grid of the receiving converter station. G_q_ref ; The input of the inner loop controller is i G_q_ref The component i of the three-phase current on the q axis of the AC side power grid of the receiving converter station G_q The output is the component u of the three-phase voltage on the dq axis of the AC side grid of the receiving converter station. G_ MMC _q .

[0031] u G_d and u G_q It can be obtained by formula (10)-(11), i G_d and i G_q Through formula (12)-(13):

[0032]

[0033]

[0034] In the formula, u G-A-LP 、u G-B-LP 、u G-C-LP 、i WG-A-LP 、i G-B-LP 、i G-C-LP for u G-A 、u Wg-B 、u Wg-C 、i G-A-LP 、i G-B 、i G-C It is obtained through low-pass filter 4 and low-pass filter 5.

[0035] Second aspect: The control strategy of the receiving-end converter station and the sending-end converter station includes 11 PI controllers, 6 QR controllers, 5 first-order low-pass filters, and 2 second-order low-pass filters. Each PI controller contains two parameters, each QR controller contains 3 parameters, each first-order low-pass filter contains one parameter, and each second-order low-pass filter contains 2 parameters. That is, there are 49 parameters that need to be set in the control system of the receiving-end converter station and the sending-end converter station. These 49 parameters can be optimized by the sand cat swarm optimization algorithm.

[0036] The step S3 specifically comprises the following steps:

[0037] Step 1: Initialize the location of the sand cat population;

[0038] Step 2: In the exploration phase, each cat in the group updates its next position based on the optimal solution, its current position, and its sensitivity range. The new iterative position is between the prey position and the current position.

[0039] Step 3: During the predation phase, the sand cat is set to approach the hunting position at a random angle and eventually obtain the optimal position.

[0040] Furthermore, population initialization is expressed as:

[0041] X i = l bi +a i ×(u bi -l bi ) (14)

[0042] Where, X i is the position of the i-th sand cat individual, a i is a random number between 0 and 1, u bi represents the upper boundary of the position of the i-th sand cat individual, l bi Represents the lower boundary of the location of the i-th sand cat individual.

[0043] The position of a sand cat represents a combination of 49 parameters, so X i is a 1×49 array. bi , l bi It is also a 1×49 array.

[0044] Furthermore, the search phase can be expressed as:

[0045]

[0046] In the formula, S Mi is the auditory characteristic of the i-th sand cat individual and can be set to a constant, such as 2. Gi is the sensitivity range of the i-th sand cat, t represents the current iteration number, and T is the total iteration number.

[0047] To avoid the r of every sand cat Gi The sand cat swarm optimization algorithm tends to be consistent, causing it to fall into a local optimal solution. The sensitivity update formula of each sand cat is shown in formula (16).

[0048] r i (t) = a i × Gi (t) (16)

[0049] In the formula, a i is a random number between 0 and 1, and the i-th sand cat updates its iterative position according to formula (17).

[0050] Pos i (t+1)=r i (t)×(Posbc i (t)-rand(0,1)×Posc i(t)) (17)

[0051] In the formula, Posbc i (t) is the current optimal solution position, Posc i (t) is the position of the i-th sand cat in the t-th iteration, Pos i (t+1) is the iterative position of the i-th sand cat.

[0052] Furthermore, the sign of the transition of the sand cat group from the search phase to the predation phase is R i , when |R i |>1, the sand cat enters the exploration phase; when |R i |≤1, the sand cat enters the predation stage; the R i It is expressed as:

[0053] R i (t) = 2 × r Gi (t)×rand(0,1)-r Gi (t) (18)

[0054] Furthermore, the predation stage formula can be expressed as:

[0055] Pos rndi (t)=|rand(0,1)×Posbc i (t)-Posc i (t)| (19)

[0056] Pos i (t+1)=Posbc i (t)-R i (t)×Pos rndi (t)×cos(rand(0,1)×2π) (20)

[0057] Furthermore, the Pos obtained by the sand cat group optimization algorithm is i (T) is the final optimization result, Pos i (T) is a 1×49 array, i.e., the values ​​of 49 parameters of the control systems of the receiving-end converter station and the sending-end converter station.

[0058] Beneficial effect: The present invention obtains the parameters that need to be set in the control strategy of the receiving-end converter station and the sending-end converter station of the offshore wind farm through the sand cat swarm optimization algorithm instead of obtaining them through experience, which can reduce the parameter setting debugging time, improve the accuracy of parameter setting, and thus improve the performance of the control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 The invention relates to an offshore wind farm structure.

[0060] Figure 2 It is a topological diagram of the MMC composed of the half-bridge submodules of the present invention.

[0061] Figure 3 It is a flow chart of the circulating current suppression link of the sending-end converter of the present invention.

[0062] Figure 4 It is a flow chart of the circulating current suppression link of the receiving-end converter of the present invention.

[0063] Figure 5 It is a control block diagram of the sending-end converter of the present invention.

[0064] Figure 6 It is the dq axis of the voltage and current of the power grid on the wind farm side of the present invention.

[0065] Figure 7 It is a modulation block diagram of the MMC converter of the sending-end converter station of the present invention.

[0066] Figure 8 It is a control strategy block diagram of the MMC converter in the receiving-end converter station of the present invention.

[0067] Fig. 9 It is the voltage component of the dq axis on the wind turbine generator set side of the present invention.

[0068] Fig.10 It is a modulation block diagram of the MMC converter of the receiving-end converter station of the present invention.

[0069] Fig.11 This is a block diagram of the grid voltage vector angle calculation of the present invention.

[0070] Fig.12 It is a step flowchart of the present invention.

[0071] Fig.13 It is a step diagram of the parameters of the control system optimized by the sand cat group optimization algorithm in the present invention. DETAILED DESCRIPTION

[0072] In order to make the objectives, technical solutions and advantages of the present invention more clear and explicit, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0073] like Fig.12As shown, the present invention discloses a method for optimizing parameters of a wind farm converter station based on a sand cat group algorithm, comprising the following steps: S1: constructing the structure of an offshore wind farm converter station, wherein the structure of the offshore wind farm comprises a sub-wind farm, a 35kV collector, a 220kV bus, a transformer, a sending-end converter station and a receiving-end converter station, wherein the sub-wind farm is converged through a 35kV collector, and then connected to the 220kV bus through a step-up transformer, and the 220kV bus is connected to the sending-end converter station to convert AC power into DC power, and then long-distance power transmission is performed through a DC transmission line. When reaching the load side, the DC transmission line is connected to the receiving-end converter station to convert DC power into AC power to supply power to the load;

[0074] S2: Analyze the control systems of the sending-end converter station and the receiving-end converter station, and clarify the parameters that need to be set in the control system;

[0075] S3: Use the Sand Cat algorithm to optimize the parameters that need to be set.

[0076] First aspect: The structure of offshore wind farms Figure 1 As shown, it includes a sub-wind farm, a 35kV collector, a 220kV bus, a transformer, a sending-end converter station and a receiving-end converter station. The sub-wind farm is connected through a 35kV collector, and then connected to the 220kV bus through a step-up transformer. The 220kV bus is connected to the sending-end converter station to convert AC power into DC power, which is transmitted over long distances through a DC transmission line. When it reaches the load side, the DC transmission line is connected to the receiving-end converter station to convert DC power into AC power to supply power to the load. The sending-end converter station and the receiving-end converter station use MMC composed of half-bridge sub-modules as converters, and their topology is as follows: Figure 2 shown.

[0077] The MMC is an abbreviation for modular multilevel converter.

[0078] Figure 2 SM in the code represents a submodule, and U dc is the DC bus voltage, u pa is the voltage of the upper bridge arm of phase a, u na is the bridge arm voltage of phase a, L0 is the bridge arm reactance, i pa is the upper arm current of phase a, i na is the lower arm current of phase a, a, b, c represent three-phase A, B and C, and j is used to represent a, b, c in the following text.

[0079] Furthermore, the circulating current i in the MMC converter in the sending-end converter station Wg_cirj It can be expressed as formula (1). Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. Therefore, the circulating current can also be expressed as formula (2).

[0080]

[0081]

[0082] In the formula, I Wg_dc is the DC component in the converter bridge arm commutation, I Wg_k is the amplitude of the kth harmonic component in the bridge arm commutation. The harmonic components in the circulating current are basically even-numbered harmonic components. Wg_pj ,i Wg_nj They are the upper arm current and lower arm current of the MMC converter in the sending-end converter station respectively. The subscript Wg represents the power grid on the wind farm side.

[0083] The quasi-proportional resonant controller can have a large gain at a specified harmonic frequency, and its transfer function can be expressed as:

[0084]

[0085] In the formula, ω r is the resonant frequency, ω c is the cut-off frequency, K r is the resonance coefficient of the quasi-proportional resonant controller.

[0086] Taking the suppression of the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current as an example, the circulating current suppression controller is as follows: Figure 3 shown.

[0087] The control process of the circulating current suppressor is:

[0088] After the circulating current passes through a second-order low-pass filter to obtain the low-frequency component, a subtraction operation is performed to obtain the high-frequency component. PI controller 1 tracks the low-frequency component, QR controller 1 suppresses the 2-fold frequency component, QR controller 2 suppresses the 4-fold frequency component, and QR controller 3 suppresses the 6-fold frequency component. That is, QR controller 1, QR controller 2, and QR controller 3 suppress the 2nd harmonic component, 4th harmonic component, and 6th harmonic component in the circulating current, respectively.

[0089] Furthermore, the circulating current i in the MMC converter in the receiving end converter station G_cirj It is expressed as formula (4). Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. Therefore, the circulating current is also expressed as formula (5).

[0090]

[0091]

[0092] In the formula, I G_dcis the DC component in the converter bridge arm commutation, I G_k is the amplitude of the kth harmonic component in the bridge arm commutation; the harmonic components in the circulating current are basically even-numbered harmonic components; i G_pj is the upper arm current of the MMC converter in the sending-end converter station, i G_nj is the lower arm current of the MMC converter in the sending-end converter station.

[0093] Taking the suppression of the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current as an example, the circulating current suppression controller is as follows: Figure 4 shown. Figure 4 After the medium circulating current passes through a second-order low-pass filter to obtain the low-frequency component, a subtraction operation is performed to obtain the high-frequency component. The PI controller 2 tracks the low-frequency component, the QR controller 4 suppresses the 2-fold frequency component, the QR controller 5 suppresses the 4-fold frequency component, and the QR controller 6 suppresses the 6-fold frequency component. That is, the QR controller 4, the QR controller 5, and the QR controller 6 suppress the 2nd harmonic component, the 4th harmonic component, and the 6th harmonic component in the circulating current, respectively.

[0094] Furthermore, the control strategy block diagram of the converter at the sending-end converter station is as follows: Figure 5 As shown. It is divided into a d-axis voltage control link and a q-axis voltage control link; wherein, the d-axis voltage control link includes an outer-loop PI controller 3, an inner-loop controller 5, and an output limiter of the inner-loop and outer-loop controllers; the q-axis voltage control link includes an outer-loop PI controller 4, an inner-loop controller 6, and an output limiter of the inner-loop and outer-loop controllers. In addition, the input information of the d-axis outer-loop controller is the d-axis voltage, and the output information is the d-axis current reference value, which is also the input information of the PI controller 5. The input information of the q-axis outer-loop controller is the q-axis voltage, and the output information is the q-axis current reference value, which is also the input information of the PI controller 5.

[0095] u Wg-d 、u Wg-q 、i Wg-d 、i Wg-q They are obtained by detecting the three-phase voltage and three-phase current of the 220kV power grid on the wind turbine side through a low-pass filter and then through coordinate transformation. The calculation process is as follows: Figure 6 shown.

[0096] The coordinate transformation expression can be expressed as:

[0097]

[0098] In the formula, u Wg-A-LP 、u Wg-B-LP 、u Wg-C-LP Represents the low-frequency component of the three-phase grid voltage on the wind farm side; i Wg-A-LP 、iWg-B-LP 、i Wg-C-LP Represents the low-frequency component of the three-phase grid current on the wind farm side; u Wg-α 、u Wg-β Represents the low-frequency component of the three-phase grid voltage on the wind farm side in the αβ axis; i Wg-α 、i Wg-β It represents the low-frequency component of the three-phase grid current on the wind farm side in the αβ axis; u Wg-d 、u Wg-q The low-frequency component of the three-phase grid voltage on the wind farm side in the dq axis, i Wg-d 、i Wg-q It represents the low-frequency component of the three-phase grid voltage on the wind farm side in the dq axis, θ Wg Represents the voltage vector angle of the grid on the wind farm side.

[0099] Furthermore, the modulation strategy of the MMC converter in the sending-end converter station adopts the nearest level approximation modulation algorithm, which takes into account the capacitor voltage balance of the MMC submodule, such as Figure 7 shown.

[0100] Figure 7 in, u Wg-MMC-α is the α-axis component of the three-phase output voltage of the bridge arm of the sending-end MMC converter on the αβ axis, u Wg-MMC-β is the β-axis component of the three-phase output voltage of the bridge arm of the sending-end MMC converter on the αβ axis. Wg-MMC-a 、u Wg-MMC-b 、u Wg-MMC-c are the components of the three-phase output voltage of the bridge arm of the sending-end MMC converter on the abc axis; u Wg-cirA 、u Wg-cirB 、u Wg-cirC They are respectively the components of the three-phase output voltage of the bridge arm of the MMC converter on the abc axis taking into account the circulating current components of the upper and lower bridge arms of the sending-end MMC converter.

[0101] Furthermore, the control strategy block diagram of the MMC in the receiving converter station is as follows: Figure 8 As shown, it is specifically divided into a DC bus voltage control loop and a reactive power control loop; the DC bus voltage loop is divided into a voltage outer loop and a current inner loop, the voltage outer loop adopts a PI controller 6, and the current inner loop adopts a PI controller 8; the reactive power loop is divided into a reactive power outer loop and a current inner loop, the reactive power outer loop adopts a PI controller 7, and the current inner loop adopts a PI controller 9.

[0102] The input of the voltage outer loop controller is the collected DC bus voltage u dc and DC bus voltage reference value u dc_ref The output is the reference value i of the three-phase current component on the d-axis of the AC side power grid of the receiving converter station.G _ d _ ref ; The inner loop controller is i G_d_ref The difference between the actual value of the three-phase current component on the d-axis of the AC side grid of the receiving converter station. The output is the output three-phase voltage component u of the MMC converter of the receiving converter station on the d-axis. G_MMC_d .

[0103] The input of the reactive power outer loop controller is the reactive power Q G and reactive power reference value Q G_ref The difference is the component reference value i of the three-phase current on the q axis of the AC side power grid of the receiving converter station. G_q_ref ; The input of the inner loop controller is i G_q_ref The component i of the three-phase current on the q axis of the AC side power grid of the receiving converter station G_q The output is the component u of the three-phase voltage on the dq axis of the AC side grid of the receiving converter station. G_ MMC _q .

[0104] u G_d and u G_q It can be obtained by formula (10)-(11), i G_d and i G_q Through formula (12)-(13):

[0105]

[0106] In the formula, u G-A-LP 、u G-B-LP 、u G-C-LP 、i WG-A-LP 、i G-B-LP 、i G-C-LP for u G-A 、u Wg-B 、u Wg-C 、i G-A-LP 、i G-B 、i G-C It is obtained through low-pass filter 4 and low-pass filter 5.

[0107] Furthermore, the modulation strategy of the MMC converter in the receiving-end converter station adopts the nearest level approximation modulation algorithm, which takes into account the capacitor voltage balance of the MMC submodule, such as Fig.10 shown.

[0108] Fig.10 in, u G-MMC-α is the α-axis component of the three-phase output voltage of the bridge arm of the receiving-end MMC converter on the αβ axis, u G-MMC-β is the β-axis component of the three-phase output voltage of the bridge arm of the receiving-end MMC converter on the αβ axis; u G-MMC-a 、uG-MMC-b 、u G-MMC-c are the components of the three-phase output voltage of the bridge arm of the receiving-end MMC converter on the abc axis; u G-cirA 、u G-cirB 、u G-cirC They are the components of the three-phase output voltage of the bridge arm of the MMC converter on the abc axis, taking into account the circulating current components of the upper and lower bridge arms of the receiving-end MMC converter.

[0109] Furthermore, θ Wg ,θ G It is obtained through a three-phase synchronous phase-locked loop, and its calculation block diagram is as follows Fig.11 The sending end three-phase synchronous phase-locked loop adopts a PI controller 10, and the receiving end three-phase synchronous phase-locked loop adopts a PI controller 11.

[0110] Second aspect: The control strategy of the receiving-end converter station and the sending-end converter station includes 11 PI controllers, 6 QR controllers, 5 first-order low-pass filters, and 2 second-order low-pass filters. Each PI controller contains two parameters, each QR controller contains 3 parameters, each first-order low-pass filter contains one parameter, and each second-order low-pass filter contains 2 parameters. That is, there are 49 parameters that need to be set in the control system of the receiving-end converter station and the sending-end converter station. These 49 parameters can be optimized by the sand cat swarm optimization algorithm.

[0111] like Fig.13 As shown, the specific steps of step S3 are as follows:

[0112] Step 1: Initialize the location of the sand cat population;

[0113] Step 2: In the exploration phase, each cat in the group updates its next position based on the optimal solution, its current position, and its sensitivity range. The new iterative position is between the prey position and the current position.

[0114] Step 3: During the predation phase, the sand cat is set to approach the hunting position at a random angle and eventually obtain the optimal position.

[0115] Furthermore, population initialization can be expressed as:

[0116] X i = l bi +a i ×(u bi -l bi ) (14)

[0117] Where, X i is the position of the i-th sand cat individual, a i is a random number between 0 and 1, u bi represents the upper boundary of the position of the i-th sand cat individual, lbi Represents the lower boundary of the location of the i-th sand cat individual.

[0118] The position of a sand cat represents a combination of 49 parameters, so X i is a 1×49 array. bi , l bi It is also a 1×49 array.

[0119] Furthermore, the search phase can be expressed as:

[0120]

[0121] In the formula, S Mi is the auditory characteristic of the i-th sand cat individual and can be set to a constant, such as 2. Gi is the sensitivity range of the i-th sand cat, t represents the current iteration number, and T is the total iteration number.

[0122] To avoid the r of every sand cat Gi The sand cat swarm optimization algorithm tends to be consistent, causing it to fall into a local optimal solution. The sensitivity update formula of each sand cat is shown in formula (16).

[0123] r i (t) = a i × Gi (t) (16)

[0124] The i-th sand cat updates its iterative position according to formula (17).

[0125] Pos i (t+1)=r i (t)×(Posbc i (t)-rand(0,1)×Posc i (t)) (17)

[0126] In the formula, Posbc i (t) is the current optimal solution position, Posc i (t) is the position of the i-th sand cat in the t-th iteration, Pos i (t+1) is the iterative position of the i-th sand cat.

[0127] Furthermore, the sign of the transition of the sand cat group from the search phase to the predation phase is R i , when |R i |>1, the sand cat enters the exploration phase; when |R i |≤1, the sand cat enters the predation stage. i It can be expressed as:

[0128] R i (t) = 2 × rGi (t)×rand(0,1)-r Gi (t) (18)

[0129] Furthermore, the predation stage formula can be expressed as:

[0130] Pos rndi (t)=|rand(0,1)×Posbc i (t)-Posc i (t)| (19)

[0131] Pos i (t+1)=Posbc i (t)-R i (t)×Pos rndi (t)×cos(rand(0,1)×2π) (20)

[0132] Furthermore, the Pos obtained by the sand cat group optimization algorithm is i (T) is the final optimization result, Pos i (T) is a 1×49 array, i.e., the values ​​of 49 parameters of the control systems of the receiving-end converter station and the sending-end converter station.

[0133] The present invention obtains the parameters that need to be set in the control strategy of the receiving-end converter station and the sending-end converter station of the offshore wind farm through a sand cat swarm optimization algorithm rather than through experience, which can reduce the parameter setting debugging time, improve the accuracy of parameter setting, and further improve the performance of the control system.

[0134] The above is only a preferred implementation mode of the present invention, which certainly cannot be used to limit the scope of rights of the present invention. It should be pointed out that for technicians in this technical field, any modification or equivalent replacement of the technical solution of the present invention without creative work does not depart from the protection scope of the technical solution of the present invention.

Claims

1. A method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm, characterized in that: S1: Construct the structure of the offshore wind farm converter station, which includes a sub-wind farm, a 35kV collector, a 220kV bus, a transformer, a sending-end converter station and a receiving-end converter station; the sub-wind farm is connected through the 35kV collector, and then connected to the 220kV bus through a step-up transformer. The 220kV bus is connected to the sending-end converter station to convert AC power into DC power, and then transmitted over long distances through DC transmission lines. When it reaches the load side, the DC transmission line is connected to the receiving-end converter station to convert DC power into AC power to supply power to the load; S2: Analyze the control systems of the sending-end converter station and the receiving-end converter station, and clarify the parameters that need to be set in the control system; S3: Use the sand cat swarm algorithm to optimize the parameters that need to be set; Wherein, the step S3 specifically includes the following steps: Step 1: Initialize the location of the sand cat population; Step 2: In the exploration phase, each cat in the group updates its next position based on the optimal solution, its current position, and its sensitivity range. The new iterative position is between the prey position and the current position. Step 3: During the predation phase, the sand cat is set to approach the hunting position at a random angle and eventually obtain the optimal position.

2. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: Circulating current in the MMC converter in the sending-end converter station Wg_cirj It is expressed as formula (1); Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. Wg_cirj It can also be expressed as formula (2): In the formula, I Wg_dc is the DC component in the converter bridge arm commutation, I Wg_k is the amplitude of the kth harmonic component in the bridge arm commutation; the harmonic components in the circulating current are basically even-numbered harmonic components; i Wg_pj is the upper arm current of the MMC converter in the sending-end converter station, i Wg_nj is the lower arm current of the MMC converter in the sending-end converter station, and the subscript Wg represents the power grid on the wind farm side; k represents the harmonic order, ω0 represents the fundamental angular frequency, represents the initial phase of the kth harmonic; The quasi-proportional resonant controller can have a large gain at a specified harmonic frequency, and its transfer function is expressed as: In the formula, ω r is the resonant frequency; ω c is the cut-off frequency, K r is the resonance coefficient of the quasi-proportional resonant controller; Taking the suppression of the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current in the control strategy of the sending-end converter station as an example, the circulating current suppression controller needs to include a second-order low-pass filter, a PI controller 1, and three QR controllers QR controller 1, QR controller 2, and QR controller 3 to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current respectively.

3. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 2, characterized in that: Circulating current in the MMC converter in the receiving converter station G_cirj It is expressed as formula (4); Since the high-frequency harmonic components of the upper bridge arm voltage and the lower bridge arm voltage in the MMC converter pass through the bridge arm reactance, harmonic components are introduced into the circulating current. G_cirj It can also be expressed as formula (5): In the formula, I G_dc is the DC component in the converter bridge arm commutation, I G_k is the amplitude of the kth harmonic component in the bridge arm commutation; the harmonic components in the circulating current are basically even-numbered harmonic components; i G_pj is the upper arm current of the MMC converter in the sending-end converter station, i G_nj is the lower arm current of the MMC converter in the sending-end converter station; Indicates the phase of the kth harmonic of the MMC converter in the receiving-end converter station; Taking the control strategy of the receiving-end converter station to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current as an example, the circulating current suppression controller needs to include a second-order low-pass filter, a PI controller 2, and three QR controllers QR controller 4, QR controller 5, and QR controller 6 to suppress the second harmonic component, the fourth harmonic component and the sixth harmonic component in the circulating current respectively.

4. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: The control strategy of the converter of the sending-end converter station is divided into a d-axis voltage control link and a q-axis voltage control link; wherein, the d-axis voltage control link includes an outer-loop PI controller 3, an inner-loop controller 5, and output limiters of the inner-loop and outer-loop controllers; the q-axis voltage control link includes an outer-loop PI controller 4, an inner-loop controller 6, and output limiters of the inner-loop and outer-loop controllers; in addition, the input information of the d-axis outer-loop controller is the d-axis voltage, and the output information is the d-axis current reference value, which is also the input information of the PI controller 5; the input information of the q-axis outer-loop controller is the q-axis voltage, and the output information is the q-axis current reference value, which is also the input information of the PI controller 5; The d-axis voltage and q-axis voltage are obtained by coordinate transformation of the a, b, and c three-phase voltages; the d-axis current and q-axis current are obtained by coordinate transformation of the a, b, and c three-phase currents; the coordinate transformation expression can be expressed as: In the formula, u Wg-A-LP 、u Wg-B-LP 、u Wg-C-LP They represent the low-frequency components of the three-phase grid voltage on the wind farm side; i Wg-A-LP 、i Wg-B-LP 、i Wg-C-LP They represent the low-frequency components of the three-phase grid current on the wind farm side; u Wg-α It represents the low-frequency component of the three-phase grid voltage on the wind farm side in the α-axis, u Wg-β Represents the low-frequency component of the three-phase grid voltage on the wind farm side in the β-axis; i Wg-α It represents the low-frequency component of the three-phase grid current on the wind farm side in the α-axis, i Wg-β represents the low-frequency component of the three-phase grid current on the wind farm side in the β axis; u Wg-d The d-axis component of the low-frequency component of the three-phase grid voltage on the wind farm side is represented by u Wg-q The q-axis component of the low-frequency component of the three-phase grid voltage on the wind farm side, i Wg-d 、i Wg-q Represents the low-frequency component of the three-phase grid current on the wind farm side in the dq axis; θ Wg Represents the voltage vector angle of the power grid on the wind farm side; u Wg-A-LP 、u Wg-B-LP 、u Wg-C-LP 、i Wg-A-LP 、i Wg-B-LP 、i Wg-C-LP for u Wg-A 、u Wg-B 、u Wg-C 、i Wg-A-LP 、i Wg-B 、i Wg-C Obtained through low-pass filter 1 and low-pass filter 2.

5. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: The modulation strategy of the MMC converter in the sending-end converter station adopts the nearest level approximation modulation algorithm, which takes into account the capacitor voltage balance of the MMC submodules.

6. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: The control system of the MMC in the receiving-end converter station is divided into a DC bus voltage control loop and a reactive power control loop; the DC bus voltage loop is divided into a voltage outer loop and a current inner loop, the voltage outer loop adopts a PI controller 6, and the current inner loop adopts a PI controller 8; the reactive power loop is divided into a reactive power outer loop and a current inner loop, the reactive power outer loop adopts a PI controller 7, and the current inner loop adopts a PI controller 9. The input of the voltage outer loop controller is the collected DC bus voltage u dc and DC bus voltage reference value u dc_ref The output is the reference value i of the three-phase current component on the d-axis of the AC side power grid of the receiving converter station. G_d_ref ; The inner loop controller is i G_d_ref The difference between the actual value of the three-phase current component on the d-axis of the AC side grid of the receiving converter station; its output is the component u of the output three-phase voltage on the d-axis of the MMC converter of the receiving converter station G_MMC_d ; The input of the reactive power outer loop controller is the reactive power Q G and reactive power reference value Q G_ref The difference is the component reference value i of the three-phase current on the q axis of the AC side power grid of the receiving converter station. G_q_ref , the input of the inner loop controller is i G_q_ref The component i of the three-phase current on the q axis of the AC side power grid of the receiving converter station G_q The output is the component u of the three-phase voltage on the dq axis of the AC side power grid of the receiving converter station. G_MMC_q ; u G_d and u G_q It can be obtained by formula (10)-(11), i G_d and i G_q Through formula (12)-(13): In the formula, u G-A-LP 、u G-B-LP 、u G-C-LP 、i WG-A-LP 、i G-B-LP 、i G-C-LP for u G-A 、u Wg-B 、u Wg-C 、i G-A-LP 、i G-B 、i G-C It is obtained through low-pass filter 4 and low-pass filter 5.

7. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: The modulation strategy of the MMC converter in the receiving-end converter station adopts the nearest level approximation modulation algorithm, which takes into account the capacitor voltage balance of the MMC submodule.

8. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: θ Wg and θ G It is obtained through a three-phase synchronous phase-locked loop, wherein the three-phase synchronous phase-locked loop at the sending end adopts a PI controller 10, and the three-phase synchronous phase-locked loop at the receiving end adopts a PI controller 11.

9. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 1, characterized in that: The control strategy of the receiving-end converter station and the sending-end converter station includes 11 PI controllers, 6 QR controllers, 5 first-order low-pass filters, and 2 second-order low-pass filters. Each PI controller contains two parameters, each QR controller contains 3 parameters, each first-order low-pass filter contains one parameter, and each second-order low-pass filter contains 2 parameters; that is, there are 49 parameters that need to be set in the control system of the receiving-end converter station and the sending-end converter station; these 49 parameters can be optimized by the sand cat swarm optimization algorithm.

10. The method for optimizing parameters of converter stations in offshore wind farms based on the sandcat swarm optimization algorithm according to claim 9, characterized in that: The population initialization is expressed as: X i =l bi +a i ×(u bi -l bi )(14) In the formula, X i is the position of the i-th sand cat individual, a i is a random number between 0 and 1, u bi represents the upper boundary of the position of the i-th sand cat individual, l bi represents the lower boundary of the position of the i-th sand cat individual; The position of a sand cat represents a combination of 49 parameters, so X i is a 1×49 array, u bi , l bi It is also a 1×49 array.

11. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 9, characterized in that: The search phase in step 2 is expressed as: In the formula, S Mi is the auditory characteristic of the i-th sand cat individual, set to a constant 2, r Gi is the sensitivity range of the i-th sand cat, t represents the current iteration number, and T is the total iteration number; To avoid the r of every sand cat Gi The sand cat swarm optimization algorithm tends to be consistent, causing it to fall into a local optimal solution. The sensitivity update formula of each sand cat is shown in formula (16): r i (t)=a i ×r Gi (t) (16) The i-th sand cat updates its iterative position according to formula (17); Pos i (t+1)=r i (t)×(Posbc i (t)-rand(0,1)×Posc i (t)) (17) In the formula, Posbc i (t) is the current optimal solution position, Posc i (t) is the position of the i-th sand cat in the t-th iteration, Pos i (t+1) is the iterative position of the i-th sand cat.

12. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 9, characterized in that: The symbol of the transition of the sand cat group from the search phase to the predation phase is R i , when |R i |>1, the sand cat enters the exploration phase; when |R i |≤1, the sand cat enters the predation stage; R i It is expressed as: R i (t)=2×r Gi (t)×rand(0,1)-r Gi (t) (18) 13. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 9, characterized in that: The formula for the predation stage in step 3 is expressed as: Pos rndi (t)=|rand(0,1)×Pos bci (t)-Pos ci (t)| (19) Pos i (t+1)=Pos bci (t)-R i (t)×Pos rndi (t)×cos(rand(0,1)×2π) (20)。 14. The method for optimizing parameters of a wind farm converter station based on a sand cat swarm algorithm according to claim 9, characterized in that: Pos calculated by the sand cat group optimization algorithm i (T) is the final optimization result, Pos i (T) is a 1×49 array, i.e., the values ​​of 49 parameters of the control systems of the receiving-end converter station and the sending-end converter station.

Images