Improved vector current control method for flexible direct current grid-connected system under weak power grid

Abstract

The invention discloses an improved vector current control method for a flexible direct current grid-connected system under a weak power grid, and the method comprises the steps: building a VSC-HVDC vector current control strategy model, enabling an outer-loop control loop to generate a current reference for an inner-loop current control loop, and independently controlling the active power and reactive power; an improved vector current control strategy is provided, and feedback control is carried out through frequency change; a small signal model of the flexible direct current grid-connected system after the IVCC is applied is deduced, and the accuracy of the small signal model is verified; coordinated control of different external control loops of the VSC-HVDC is provided, and the dynamic response of the system is improved; and testing the dynamic response under the fault of the model. According to the method, switching of complex modification and control is avoided, frequency and voltage dynamic response characteristics in the disturbance process can be considered through coordination control between the active power control loop and the voltage control loop, and reliable guarantee is provided for safe and stable operation of the VSC-HVDC system under the weak power grid.

Description

Technical Field

[0001] The invention belongs to the field of power system optimization control, and in particular relates to an improved vector current control method for a flexible direct current grid-connected system under a weak power grid. Background Art

[0002] As fossil energy is increasingly depleted, remote renewable energy such as offshore wind power is gradually penetrating into the power system, and the new power system is characterized by a high proportion of new energy. In recent years, in order to improve the high line loss and low transmission efficiency caused by the uneven distribution of remote renewable energy power generation systems, high-voltage direct current transmission has been widely used to replace traditional alternating current transmission, which has the ability to transmit large amounts of electricity over long distances.

[0003] Compared with voltage source converter based HVDC (VSC-HVDC) (flexible DC transmission technology), HVDC based on linear rectifier converter always requires reactive power compensation, and when the short circuit ratio (SCR) is less than 2, instability may occur and the system loss is large. Therefore, VSC-HVDC has gradually been applied to power systems in recent years.

[0004] Power injection in VSC-HVDC is generally controlled by a phase-locked loop based on a vector current control strategy. Although this control strategy can independently control active and reactive power, it may bring some uncertain system oscillation effects in very weak grids. The fast phase-locked loop (PLL) will amplify the change of the common coupling point (PCC) voltage during disturbances, which will lead to tracking failure. This phenomenon occurs especially in weak grids.

[0005] In order to reduce the adverse effects of PLL on system stability, the existing optimization strategies can be improved from the following aspects: adjusting PLL control parameters, modifying PLL control structure, and adopting control methods without PLL. However, adjusting PLL control parameters cannot balance system stability and system dynamic response speed; some of the methods proposed by modifying PLL control structure are complex in design and difficult to implement; the control method without PLL may switch back to the classic vector current control strategy (vector current control, VCC) when a fault occurs, which brings additional burden to practical applications, and the response speed is slow and insufficient to cope with sudden current changes. Therefore, it is urgent to design a system control method with fast response and stability through flexible DC transmission in the context of weak power grid. Summary of the invention

[0006] Purpose of the invention: In view of the above problems, the present invention proposes an improved vector current control method for a flexible direct current grid-connected (VSC-HVDC) system under a weak power grid, improves the vector current control strategy with the system frequency as the feedback signal, and increases the gain boundary of the phase-locked loop when the system is stable; a coordinated control method for different external control loops of the VSC-HVDC is proposed to improve the dynamic response of the system in a simple and fast way.

[0007] Technical solution: To achieve the purpose of the present invention, the technical solution adopted by the present invention is: an improved vector current control method for a flexible direct current grid-connected system under a weak power grid, comprising the steps of:

[0008] Establish a VSC-HVDC vector current control strategy model, where the outer control loop generates current reference for the inner current control loop to achieve independent control of active and reactive power;

[0009] An improved vector current control strategy model IVCC based on VCC is constructed, and feedback control is performed through frequency changes to keep the frequency and voltage deviations within the threshold range;

[0010] A small signal model of the flexible direct current grid-connected system after using IVCC is constructed, including the AC network, low-pass filter structure, phase-locked loop structure, and internal and external control loops. VSC-HVDC realizes coordinated control in different external control loops.

[0011] Preferably, a VSC-HVDC vector current control strategy model is established, including a low-pass filter, an outer control loop, an inner current control loop, a current regulator and a phase-locked loop;

[0012] The synchronization angle of the VSC-HVDC vector current control strategy model is obtained by locking the phase-locked loop to the common coupling voltage point; the phase-locked loop output phase angle θ tracks the input phase angle θ0;

[0013] The VSC-HVDC vector current control strategy model converts the output signal of the current regulator into a three-phase value as the input of the PWM.

[0014] Preferably, the VSC-HVDC vector current control strategy model has a hierarchical cascade structure, the internal current regulator is decoupled in the dq reference system, and the active and reactive power are independently controlled;

[0015] The dynamic equation of the outer control loop is expressed as:

[0016]

[0017] Among them, e p and e v are defined as the active power and voltage offset, v tdm and vtqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qref are the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage of the common coupling point, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively;

[0018] The dynamic equation of the inner current control loop is expressed as:

[0019]

[0020] Among them, e i1d and e i1q are the differences between the d-axis component and the q-axis component of the inner loop current control reference current and the d-axis component and the q-axis component of the VSC output current after passing through the filter, respectively. cd and v cq are the d-axis and q-axis components of the output voltage of the inner current control loop, respectively. pid With K iid are the proportional parameter and integral parameter of the inner ring d axis, K piq With K iiq are the proportional parameter and integral parameter of the inner loop q-axis respectively, L1 is the line inductance, and ω is the PLL output angular frequency.

[0021] Preferably, the improved vector current control strategy model IVCC for feedback control based on frequency change uses the grid frequency ω to describe the model tracking speed. IVCC adds a feedback link to improve the reference value of the outer loop, and uses ω as the feedback signal. The dynamic equation of the outer loop control loop is as follows:

[0022]

[0023] Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qrefare the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage, K D1 and K D2 are the damping coefficients of the active power control loop and the voltage control loop, △ω m is the frequency deviation after IVCC feedback loop filtering, satisfying △ω is the grid frequency deviation, T ω is the feedback loop filtering time, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively.

[0024] Preferably, a small signal model of the flexible direct current grid-connected system after using IVCC is constructed, and the model is expressed as:

[0025]

[0026] in, is the differential of x, u is the input variable, and is determined by the active power reference value P ref and voltage reference value V ref Composition; x is composed of 17 variables x 01 ,x 02 ,x 03 ,x 04 ,x 05 ,x1,x2,x3,x4,θ,ω,i 1d ,i 1q ,i 2d ,i 2q ,v td ,v tq The state vector composed of 01 ,x 02 ,x 03 ,x 04 ,x 05 is the state quantity related to the filter circuit, as shown in formula (5); x1, x2, x3, x4 are the state quantities related to the control loop, as shown in formula (6); θ, ω are the phase angle of the phase-locked loop voltage and the tracking frequency, respectively, i 1d ,i 1q are the d-axis and q-axis components of the VSC output current, respectively. 2d ,i 2q are the d-axis and q-axis components flowing into the infinite busbar, v td ,v tq is the d-axis and q-axis components of the common coupling point voltage, i 1d ,i 1q ,i 2d ,i 2q,v td ,v tq Together they constitute the voltage and current state quantities of the AC network;

[0027]

[0028] Among them, T mvd and T mvq are the time constants of the low-pass filter filtering the d-axis and q-axis voltage components, respectively, mid and T miq are the time constants of the low-pass filter filtering the d-axis and q-axis current components, △ω is the grid frequency deviation, T ω The time for feedback loop filtering.

[0029] Preferably, for the small signal model, according to the damping coefficient K D1 and K D2 When the eigenvalue of the dominant oscillation mode changes, the damping ratio is introduced to adjust the damping coefficient K D1 and K D2 The feasible domain of is determined, and the damping ratio ζ is expressed as follows:

[0030]

[0031] Wherein, σ is the real part of the underdamped oscillation mode, and ω is the imaginary part of the underdamped oscillation mode.

[0032] Preferably, the coordinated control of VSC-HVDC in different external control loops includes:

[0033] Damping coefficient K of active power control loop based on genetic algorithm D1 and the damping coefficient K of the voltage control loop D2 To optimize the parameter combination, first define the number of individuals in the population. Each individual in the population contains two genes, namely K D1 and K D2 , and initialize the population;

[0034] Design the objective function F of the coordinated control model and the fitness function value F obtained in each iteration fit as follows:

[0035] F(i,j)=λ1(S f (i,j)+S V (i,j))+λ2S s (i,j) (8),

[0036] F fit (i)=minF(i,j),j=1,2,…,m (9),

[0037] Among them, i is the number of population iterations, j represents the position of the individual in the population, λ1 and λ2 are defined as the weights of the dynamic response quality and the power consumption of the dynamic process, respectively, S f , S V and S s The three performance indicators of the objective function of coordinated control are S f and S V are the maximum deviations of the frequency and the voltage at the common coupling point after the disturbance, S s is the average change of active power and reactive power within t seconds after the disturbance, and m is the number of individuals in a population;

[0038] By calling the coordinated control model, the objective function is calculated. After each iteration, the coordinated control model performs selection, crossover and mutation operations on the population based on the genetic algorithm to optimize the individuals in the population and enter the next iteration.

[0039] When the fitness function F fit When convergence is achieved, the coordinated control iteration ends and the optimal solution K is output. D1 and K D2 , and obtain the optimal curve with the smallest maximum deviation between frequency and voltage and the smallest power change after the system is disturbed.

[0040] Preferably, according to the requirements for the dynamic response of the system, the constraints on the objective function F are set as follows:

[0041]

[0042] Among them, U N is the rated voltage. When the simulation result meets the constraint condition, the objective function calculation result F(i, j) is assigned. Otherwise, a penalty value is assigned to the objective function. After the calculation of F of m individuals is completed, the minimum F(i, j) is found according to formula (9) as the fitness function F of the current iteration. fit .

[0043] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0044] The present invention is to solve the problem that when VSC-HVDC is connected to a weak power grid, the voltage is greatly affected by the current. At this time, when the phase-locked loop gain is large, the q-axis voltage component v tq The change will be amplified, resulting in phase-locked loop tracking failure and system oscillation. An improved vector current control method for VSC-HVDC system under weak power grid is designed. It is proposed to use frequency change as feedback to control the smooth change of outer loop reference to achieve the goal of optimizing control and alleviating oscillation.

[0045] The small signal model proposed in this invention has sufficient accuracy. When VSC injects or absorbs power into a weak grid, the VCC strategy without damping control will become unstable if the phase-locked loop gain is large. If the phase-locked loop gain is reduced, stability can be achieved. On this basis, damping control with frequency change as feedback is added, and the system response becomes significantly faster, and the frequency and common coupling point voltage deviation is very small, proving that the damping coefficient K D1 and K D2 It will greatly affect the stability of the system. After a system failure, the AC side current, active power and common coupling point voltage can be restored to stability in a short time, ensuring fast response while taking into account stability.

[0046] Therefore, the strategy of the present invention can be improved on the basis of vector current control, avoiding other complex modifications and switching of other control structures during operation, and is easier to understand and implement in application; the accuracy of the small signal model under the strategy is derived and verified; the good response of the system in frequency dynamics and voltage dynamics is achieved, and the stability of the system can be taken into account; the coordinated control between the active power control loop and the voltage control loop is realized, and the feasibility of the model under weak power grid is improved, which has good engineering application value and provides reliable guarantee for the safe and stable operation of the VSC-HVDC system in a weak power grid environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a block diagram of the system and control strategy in Example 1.

[0048] Figure 2 1 is a phasor diagram of PLL related quantities in the first embodiment.

[0049] Figure 3 This is the response of the small signal model in Example 1 when the active power changes stepwise.

[0050] Figure 4 is the eigenvalue trajectory of the system in Example 1 (K D1 From -10 to 10, K D2 =0).

[0051] Figure 5 is the eigenvalue trajectory of the system in Example 1 (K D2 From -10 to 10, K D1 =0).

[0052] Figure 6 K in Example 1 D2 (K D1 ) when K is fixed D1 (K D2 ) is a feasible value.

[0053] Figure 7 The damping coefficients D and K in the first embodiment areD1 and K D2 function.

[0054] Figure 8 1 is a comparison of the responses of VCC and IVCC to a power step change in the VSC injection mode in Example 1.

[0055] Fig. 9 Comparison of the responses of VCC and IVCC to power step changes in the VSC absorption mode in Example 1.

[0056] Fig.10 Comparison of different control loops IVCC in Example 1.

[0057] Fig.11 This is a response curve of the method in Example 1 after a three-phase fault in the system. DETAILED DESCRIPTION

[0058] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0059] The present invention discloses an improved vector current control method for a VSC-HVDC system under a weak power grid, which is used to solve the problem that the voltage of the VSC-HVDC system is greatly affected by the current under the background of the weak power grid. When the phase-locked loop gain is large, the q-axis voltage component v of the common coupling point is tq The changes will be amplified, causing the phase-locked loop to fail in tracking and resulting in system oscillation problems.

[0060] The present invention establishes a VSC-HVDC classic vector current control strategy model, generates a current reference for the inner current control loop by the outer control loop, realizes independent control of active and reactive power, and then obtains an improved vector current control strategy (IVCC) model based on VCC. The model performs feedback control through frequency changes, verifies and analyzes the derived small signal model compared with the detailed model, and obtains the correctness of the small signal model. Through the coordinated control of the VSC-HVDC system proposed by the present invention in different external control loops, the dynamic response of the system can be improved in a simple and fast way, and the frequency dynamics and voltage dynamics in the disturbance process are taken into account, and the dynamic response of the model under fault is good. Based on the classic VCC topology, the present invention avoids complex modifications and control conversions, has good engineering practical value, and provides reliable guarantee for the safe and stable operation of the flexible direct current transmission line connected to the grid under weak power grids.

[0061] Embodiment 1:

[0062] This embodiment provides an improved vector current control method for a VSC-HVDC system under a weak power grid, and builds a detailed model based on the Matlab / Simulink platform for verification and analysis, including:

[0063] Step 1: Establish a VSC-HVDC classic vector current control strategy model, where the outer control loop generates a current reference for the inner current control loop to achieve independent control of active and reactive power. The specific steps are:

[0064] 1) According to the whole system framework Figure 1 The dynamic equations of the low-pass filter, the outer control loop, the inner current control loop and the phase-locked loop are established and described separately.

[0065] The voltage d and q axis components v at the PR coupling point in the AC network tq 、v td The noise is eliminated by low-pass filter, and v tqm and v tdm The coordinates are transformed from abc to dq and then input into the control structure.

[0066] The phase-locked loop structure obtains the output synchronization angle θ by locking the common coupling point voltage, and θ provides an angle reference for coordinate conversion.

[0067] The control structure is designed as a hierarchical cascaded dual-loop control structure. The outer control loop includes an active power control loop and a voltage control loop, the outputs of which are used as reference inputs of the current dq axis of the inner current control loop. The inner current control loop is decoupled in the dq reference system, so that the system can independently control active and reactive power; the voltage dq component output by the control loop is converted into the voltage abc component through coordinate transformation for PWM modulation (the VSC-HVDC vector current control strategy model converts the output signal of the current regulator into a three-phase value as the input of PWM), and the output of the PWM modulator is a series of pulse width modulation signals, which control the switching state of the VSC to achieve precise control of the output voltage.

[0068] 2) Combine the dynamic equations of each part to obtain the state space expression of the whole system.

[0069] The dynamic equation of the outer control loop is expressed as:

[0070]

[0071] Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqmare the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qref are the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage of the common coupling point, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively.

[0072] The dynamic equation of the inner current control loop is expressed as:

[0073]

[0074] Among them, e i1d and e i1q are the differences between the d-axis component and the q-axis component of the inner loop current control reference current and the d-axis component and the q-axis component of the VSC output current after passing through the filter, respectively. cd and v cq are the d-axis and q-axis components of the output voltage of the inner current control loop, respectively. pid With K iid are the proportional parameter and integral parameter of the inner ring d axis, K piq With K iiq are the proportional parameter and integral parameter of the inner loop q-axis respectively, L1 is the line inductance, and ω is the PLL output angular frequency.

[0075] The low-pass filter dynamic characteristic equation expression is as follows:

[0076]

[0077] Among them, T mvd and T mvq are the time constants of the low-pass filter filtering the d-axis and q-axis voltages, respectively, mid and T miq are the time constants of the low-pass filter filtering the d-axis and q-axis current components, i 1d and i 1q are the d-axis and q-axis components of the VSC output current, respectively. td and v tq are the d-axis and q-axis components of the voltage at the common coupling point, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qmThey are the d-axis and q-axis components of the VSC output current after passing through the filter.

[0078] Step 2: Propose an improved vector current control strategy model based on VCC, perform feedback control through frequency changes, and obtain smaller frequency and voltage deviations. The specific steps are:

[0079] 1) Analyze the tracking characteristics of PLL in weak power grid.

[0080] Figure 2 is the dynamic vector diagram of PLL, such as Figure 2 As shown, the PLL tries to catch up with the voltage phase angle θ0 at the common coupling point. In steady state, the PLL output phase angle θ = θ0. However, this is not the case during dynamic response. The PLL tracking performance is determined by the two control parameters K controlled by the PI. p and K i A larger PLL gain represents a faster tracking process and is preferred if regulation control of the output frequency is required.

[0081] In a strong power grid, a larger gain has a better effect on the system, v tq The change of v is relatively smooth; in contrast, in weak power grids, tq Very sensitive to small disturbances, the larger gain of the PLL will amplify v tq The change will lead to PLL tracking failure and cause system oscillation.

[0082] The dynamic characteristic equation of the PLL is expressed as follows:

[0083]

[0084] Among them, V t is the voltage amplitude at the common coupling point, θ0 is the phase angle of the voltage at the common coupling point, θ is the phase angle of the phase-locked loop output, ω is the angular frequency of the phase-locked loop output, K p and K i are proportional gain and integral gain respectively.

[0085] 2) Based on the improved vector current control strategy model of VCC, the frequency change is designed for feedback control.

[0086] In view of the problem that PLL fails to track in a weak power grid, the present invention proposes a more gentle way to adjust the control reference of the outer loop to achieve v tq That is, by using the deviation △ω of the filtered frequency as the feedback term, the damping coefficient K D1 and K D2 Add support to control the reference of the outer loop control. The modified outer loop input item becomes:

[0087]

[0088] Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, P0 is the rated active power, V0 is the rated voltage, K D1 and K D2 are the damping coefficients of the active power control loop and the voltage control loop, △ω m is the frequency deviation after IVCC feedback loop filtering, satisfying △ω is the grid frequency deviation, T ω The time for feedback loop filtering.

[0089] Compared with VCC, K is added to the expressions of each outer loop reference term D1 △ω m and K D2 △ω m Damping control term, if the PCC voltage changes very quickly, indicating a particularly fast tracking speed, the change of the reference term slows down, so the damping coefficient K in the feedback link D1 and K D2 As an outer loop parameter, it plays a role in suppressing changes and making the system transient response smoother. The response of the PCC voltage to disturbances such as power or reference voltage adjustment will not be as drastic as before.

[0090] In contrast, if the PCC voltage changes more smoothly, the IVCC will operate in a similar way to the classic VCC, i.e., in steady state, the outer loop references of IVCC and classic VCC are the same.

[0091] The improved vector current control strategy model IVCC with frequency change feedback control uses the grid frequency ω to describe the model tracking speed. IVCC adds a feedback link to improve the reference value of the outer loop and uses ω as the feedback signal. The dynamic equation of the outer loop control loop is as follows:

[0092]

[0093] Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qmare the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qref are the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage, K D1 and K D2 are the damping coefficients of the active power control loop and the voltage control loop, △ω m is the frequency deviation after IVCC feedback loop filtering, satisfying △ω is the grid frequency deviation, T ω is the feedback loop filtering time, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively.

[0094] Step 3: Derive the small signal model of the flexible direct current grid-connected system after using IVCC, and compare it with the system response of the Simulink detailed model to verify the accuracy of the small signal model; ensure that the modeling comparison between the small signal model and the Simulink detailed model is consistent, so as to facilitate the analysis of the subsequent model simulation effect. The specific steps are:

[0095] 1) Derive the small signal model of the improved vector current control strategy.

[0096] According to step 2, write the state space expression of IVCC and select 17 state variables, which are x 01 ,x 02 ,x 03 ,x 04 ,x 05 ,x1,x2,x3,x4,θ,ω,i 1d ,i 1q ,i 2d ,i 2q ,v td ,v tq The state vectors composed of x are related to the low-pass filter, the outer control loop, the inner current control loop, the phase-locked loop and the AC network. 01 ,x 02 ,x 03 ,x 04 ,x 05 is the state quantity related to the filter circuit; x1, x2, x3, x4 are the state quantities related to the control loop; θ, ω are the phase angle of the phase-locked loop voltage and the tracking frequency respectively; i 1d ,i 1q are the d-axis and q-axis components of the VSC output current, respectively. 2d ,i 2qare the d-axis and q-axis components flowing into the infinite busbar, v td ,v tq They are the d-axis and q-axis components of the voltage at the common coupling point, and the six together constitute the voltage and current state quantity of the AC network;

[0097]

[0098] Among them, T mvd and T mvq are the time constants of the low-pass filter filtering the d-axis and q-axis voltage components, respectively, mid and T miq are the time constants of the low-pass filter filtering the d-axis and q-axis current components, △ω is the grid frequency deviation, T ω The time for feedback loop filtering.

[0099] Finally, the seventeenth-order state space model of the system is obtained, and the state space expression is as follows: in, is the differential of x, u is the input variable, and is determined by the active power reference value P ref and voltage reference value V ref Composition; x is composed of 17 variables x 01 ,x 02 ,x 03 ,x 04 ,x 05 ,x1,x2,x3,x4,θ,ω,i 1d ,i 1q ,i 2d ,i 2q ,v td ,v tq The state vector composed of . The above state space expression is linearized at the operating steady-state point, and the state equation expression is as follows: Among them, A is the state matrix, B is the input matrix, is the differential of the state variable increment, △x is the state variable increment, △u is the input variable increment. A is a 17×17 square matrix.

[0100] In this embodiment, the whole system mainly includes a low-pass filter, an outer loop control structure, an inner loop current control structure, a phase-locked loop structure and an AC network. The system block diagram is shown in FIG. Figure 1 In steady state, the VSC injects 500MW of active power into the weak grid and adjusts the voltage at the common coupling point to 1.0pu. The system parameters are shown in Table 1.

[0101] Substituting the parameters in Table 1 into the state matrix A, we can finally obtain a 17th-order constant coefficient state equation. The eigenvalue of the state matrix A can reflect the stability of the system.

[0102] Table 1 System parameters

[0103] System Parameters Numeric Rated voltage 230kV Rated frequency 60Hz VSC rated power 500MW Transmission line characteristic impedance 5.67 Short circuit ratio 1 Converter Reactance 0.15pu Capacitor Filter 0.15pu Active power control gain (0.5,50) Voltage Control Gain (0.5,50) Current Control Gain (2,100)

[0104] 2) In order to verify the accuracy of the small signal model, first use Matlab / Simulink to Figure 1 The system and control structure are used to build a detailed model that meets the scenario requirements. The system parameters of the detailed model are still based on Table 1. In steady state, the VSC injects 500MW of active power into the system.

[0105] 3) Observe the time domain response of the detailed model. First, adjust the voltage at the common coupling point to 1.0pu. At t=1.5s, the power reference value changes from the original 1.0pu to 0.99pu. The curve obtained by this time domain response is as follows: Figure 3 shown.

[0106] 4) In the same scenario, the time domain response of the small signal model is simulated according to the code. The curve of the time domain response is as follows: Figure 3 shown.

[0107] Figure 3 It shows that under the same operating scenario, the time domain response curves of the small signal model using the improved vector current control strategy are basically consistent with those of the detailed model. Under the condition that a certain error is allowed, the figure satisfies the consistency of the results.

[0108] Therefore, the small signal model derived by the present invention has sufficient accuracy.

[0109] 5) In K p =50,K i =250, calculate the following four system oscillation modes corresponding to damping control: (a) K D1 =0,K D2 =0,(b)K D1 =-5,K D2 =0,(c)K D1 =0,K D2 =5,(d)K D1 =-5,K D2 =5, the calculation results are shown in Table 2.

[0110] When K D1 =0,K D2 = 0, the dominant eigenvalue of the system is -4.565 ± j24.394, then the oscillation frequency of the small signal model can be calculated by 24.394 / 2 / π = 3.8896 Hz, which is the same as Figure 3 The time simulation results in (a) are basically consistent, and the other cases can be verified in the same way.

[0111] Therefore, based on the derived small signal model, the calculated eigenvalues ​​of the system are consistent with the time simulation results, verifying the accuracy of the small signal model.

[0112] Table 2 Characteristic roots of linearized system

[0113]

[0114] Step 4: Propose coordinated control of VSC-HVDC in different external control loops to improve the dynamic response of the system, including frequency dynamics and voltage dynamics, in a simple and fast way. The specific steps are:

[0115] 1) Draw the trajectory of the underdamped oscillation mode.

[0116] Based on the derived small signal model, the eigenvalues ​​of the calculated underdamped oscillation mode closest to the imaginary axis are plotted as the damping coefficient varies.

[0117] There are two specific scenarios:

[0118] (a) Active power control loop damping coefficient K D1 The voltage control loop damping coefficient K varies from -10 to 10. D2 = 0, observe the motion trajectory of the underdamped oscillation mode closest to the imaginary axis calculated by the small signal model, such as Figure 4 shown.

[0119] (b) Voltage control loop damping coefficient K D2 From -10 to 10, the active power control loop damping coefficient K D1 = 0, observe the motion trajectory of the underdamped oscillation mode closest to the imaginary axis calculated by the small signal model, such as Figure 5 shown.

[0120] It can be concluded from the eigenvalue calculation and the underdamped oscillation mode trajectory that when K D1 =0,K D2 = 0, the system is unstable. Figure 4 , Figure 5 In the indication, they all have positive real parts.

[0121] For scenario (a), when K is fixed D2 =0, K D1 In the process from -10 to 10, the oscillation frequency keeps decreasing, the damping ratio keeps decreasing, the oscillation mode changes from the original underdamped oscillation mode to the amplified oscillation mode, and the system changes from a stable state to an unstable state, which is not conducive to the stability of the system. Therefore, in this scenario, a negative K is needed. D1 To ensure K D2 When =0, all eigenvalues ​​are on the left half plane of the s-plane.

[0122] For scenario (b), when K is fixed D1 =0, K D2 In the process from -10 to 10, the oscillation frequency first increases continuously and then changes suddenly. After the sudden change, it still has an increasing trend. The oscillation damping ratio continues to increase, and the oscillation mode changes from the original amplification oscillation mode to the under-damped oscillation mode. The system changes from the original unstable state to the stable state, which is conducive to the stability of the system. Therefore, in this scenario, a positive K is required. D2 To ensure K D1 When =0, all eigenvalues ​​are on the left half plane of the s-plane.

[0123] 2) Analyze the feasible value of the damping coefficient under the positive damping state.

[0124] By introducing the oscillation damping ratio, K D1 and K D2 To analyze the stability of the system more thoroughly, the damping ratio of the general oscillation mode is defined as Where σ and ω are the real and imaginary parts of the oscillation mode λ = σ + jω, respectively. Similar to 1), this step only focuses on the oscillation damping coefficient of the dominant characteristic root.

[0125] Calculate K D1 and K D2 When one damping coefficient is constant, all feasible values ​​of the other damping coefficient can be found so that the oscillation damping ratio Is positive.

[0126] Figure 6 (a) and Figure 6 (b) in the figure gives different K D2 K D1 By observing the feasible value range, we can conclude that the larger K D2 and a smaller K D1 This contributes to a wider feasible area, which results in more choices of damping coefficients.

[0127] 3) Analyze the coordinated control from different external control loops.

[0128] First, set the proportional integral parameter K of the system p =130,K i =650; then, according to the small signal model calculation, the damping ratio of the underdamped oscillation mode closest to the imaginary axis is plotted as K D1 and K D2 Functions such as Figure 7 The corresponding damping ratio in the unstable case is defined as -1. Figure 7 Represents unstable combinations of damping coefficients, including the case with no damping control.

[0129] For example, when K D1 = -5 and K D2 =5, the damping ratio reaches 0.57. However, when K D1 = -10 and K D2 = 0, the damping ratio is only 0.01. Therefore, coordination between the damping supports of the two outer rings is necessary.

[0130] Depend on Figure 7 From (a) in the figure, it can be concluded that in the stable region, when K D1 When it is small, the damping ratio is K D2 More sensitive, as shown by the step shape. On the contrary, when K D1 When it is larger, the damping ratio is K D2 The sensitivity is low, such as Figure 7 As shown in (a) in . Figure 7 From (b) in the figure, we can conclude that K D2 = 0, the damping ratio is much smaller than K D1 = 0. This means that when a larger PLL gain is used, the support of an external voltage control loop is necessary; generally speaking, a larger amplitude K D1 and K D2 A larger damping ratio is possible because it represents a faster settling time. However, it should be noted that K D1 and K D2 The absolute value of cannot be too large. On the one hand, the increase in the damping ratio is limited. On the other hand, the converter is required to provide greater power.

[0131] The coordinated control steps of VSC-HVDC in different external control loops are as follows:

[0132] 1) Define the number of individuals in the population m; each individual in the population contains 2 genes, that is, K D1 and K D2 , and initialize the population according to the set specific range.

[0133] 2) The VSC-HVDC connected weak grid model using IVCC is called to calculate the objective function of the initialized population. The objective function F is as follows:

[0134] F(i,j)=λ1(S f (i,j)+S V (i,j))+λ2S s (i,j),

[0135] Where i is the number of population iterations, and j represents the position of the individual in the population. λ1 and λ2 are defined as the weights of the dynamic response quality and the power consumption of the dynamic process, respectively. f , S V and Ss The three performance indicators used as reference for the objective function of coordinated optimization control. Specifically, S f and S V are the maximum deviations of the frequency and the voltage at the common coupling point after the disturbance, S s It is the average change of active power and reactive power within t seconds after the disturbance.

[0136] According to the requirements for the dynamic response of the system, the constraints on F are set as follows:

[0137]

[0138] Among them, U N is the rated voltage. When the simulation result meets the constraint conditions, the system assigns the objective function calculation result F(i,j), otherwise it assigns a larger penalty value to the objective function.

[0139] 3) Calculate the fitness function value F obtained by iteration fit , F fit as follows:

[0140] F fit (i)=minF(i,j),(j=1,2,…,m),

[0141] Where m is the number of individuals in a population. After the calculation of F for m individuals is completed, the minimum F(i,j) is found according to formula (9) as the fitness function F of the current iteration. fit .

[0142] 4) Based on the genetic algorithm, the population is selected, crossed and mutated to optimize the individuals in the population and complete a population iteration. fit If it does not converge, proceed to the next iteration.

[0143] 5) When the fitness function F fit The iteration ends when convergence is achieved, and the optimal solution K is output D1 and K D2 , and obtain the optimal curve with the smallest maximum deviation between frequency and voltage and the smallest power change after the system is disturbed.

[0144] Step 5: Compare the dynamic performance of classic VCC and IVCC, compare the impact of different outer loop controls on the IVCC system response, compare the performance indicators of IVCC after coordinating parameters with other controls, and test the dynamic response under model faults. The specific steps are:

[0145] 1) In Matlab / Simulink Figure 1 The system shown is used to build a model.

[0146] 2) Compare classic VCC and IVCC.

[0147] (a) VSC injects power into the grid.

[0148] Set the power step change from 1.0pu to 0.95pu, Figure 8 The dynamic response curves of frequency, active power and voltage amplitude at the common coupling point during the change are shown. If a large PLL gain of 130 is used and there is no damping control, the system is unstable. If the PLL gain is reduced to 50, stability can be achieved. It is also shown that the large PLL gain of 130 and IVCC are respectively D1 = -5 and K D2 = 5. Compared with the classic VCC with reduced PLL gain, not only the frequency, active power and common coupling point voltage amplitude stabilization time is faster, but also the frequency and common coupling point voltage deviation is smaller, and the dynamic response performance is better.

[0149] (b) The VSC absorbs power from the grid.

[0150] Set the power step change from -0.75pu to -0.8pu, Fig. 9 The dynamic response curves of frequency, active power, and voltage amplitude at the common coupling point during the change are shown. When the PLL proportional gain is set to 6 and there is no damping control, the system becomes unstable. Stability can be achieved when a smaller gain of 5 is used. It is also shown that stability can be maintained even with a gain of 10 after applying IVCC. In addition, it is observed that the maximum frequency and PCC voltage amplitude deviations of the classical VCC with reduced gain are almost the same as those of IVCC. However, the settling time of IVCC is much faster. This proves that the proposed control strategy can improve the PLL gain when the VSC operates in absorption.

[0151] 3) Comparison of different outer control loops.

[0152] Fig.10 The system performance of power step change from 1.0pu to 0.95pu under three different POD-VCC control is compared, and the dynamic response curves of frequency, active power, common coupling point voltage amplitude and reactive power during the change process are shown. The damping coefficient of the IVCC outer loop control loop is set to K D1 = -10 and K D2 = 0, only the active power control loop from the outside is used. Although the system eventually reaches stability, the transient process is poor. This is also related to Figure 7 This is consistent with the conclusion of low damping ratio shown in (a). Fig.10 The damping coefficient setting of the outer loop control loop shows better dynamic performance, which shows the outer voltage loop (K D1 =0 and K D2=10), both external voltage circuits are equipped with damping control (K D1 = -5 and K D2 =5). This indicates that if a larger PLL gain is used, the damping control from the external voltage loop should be considered. Fig.10 As shown in Figure 1, damping control from only the outer voltage loop produces a smoother voltage curve compared to the damping control results from both outer loops. At the same time, both active power and reactive power drop faster. This is because a larger K D1 and K D2 .

[0153] 4) Performance comparison of the coordinated IVCC with other working conditions.

[0154] When the IVCC outer loop control parameters are coordinated and optimized in the power transmission mode, the fitness function converges in the fifth iteration, and the fitness function value of the optimal working condition is 1.1843. The output result K D1 =-9.78,K D2 =7.92. According to S f , S V and S s The three performance indicators are compared in the coordinated working condition, the three working conditions in the emission mode and the existing control MVCC. The results are shown in Table 3. Compared with MVCC, IVCC is not only f +S V There are significant advantages in the index S s This makes the fitness function of IVCC more superior. D1 =0,K D2 =10, it is superior in both the maximum deviation of frequency and voltage and the power change, so that it has the smallest fitness function and achieves the coordination purpose.

[0155] Table 3 Performance indicators of different working conditions in emission mode

[0156]

[0157]

[0158] 5) Dynamic response to three-phase fault.

[0159] This step uses the IVCC control strategy to test the fault ride-through capability of the system. First, a phase-to-ground fault is set at the common coupling point, the fault occurs at a time constant of t=2s, and the fault is preset to be cleared after the 10th 60 Hz cycle. Fig.11 The dynamic response curves of AC side current, active power and common coupling point voltage amplitude during the fault process are shown. Fig.11 It shows that after the control link detects a fault, the internal current reference is set to the reference value before the fault, but the control strategy does not need to be changed. Fig.11 The system shown can recover stability in a short time and its fault ride-through capability is excellent in very weak grids.

[0160] The method proposed in the present invention can be used to solve the problems such as system oscillation caused by PLL tracking failure in a VSC-HVDC system under a weak power grid, and has good dynamic performance.

[0161] The present invention discloses an improved vector current control method for a VSC-HVDC system under a weak power grid, which is used to solve the problem that the voltage is greatly affected by the current when the VSC-HVDC is connected to a weak power grid. When the phase-locked loop gain is large, v tq The changes will be amplified, resulting in the failure of phase-locked loop tracking, and then causing system oscillation problems and other problems. Based on the classical vector current control strategy, an improved vector current control strategy based on frequency feedback control is designed. Compared with the existing methods, the improved vector control strategy is based on the existing mature VCC strategy, avoiding other complex modifications and switching of other control structures during operation, and is easier to understand and implement in applications; it broadens the maximum limit of the phase-locked loop gain, realizes the coordinated control between the active power control loop and the voltage control loop, and improves the feasibility of the model under weak power grids, with good engineering application value; it achieves good response of the system in frequency dynamics and voltage dynamics, obtains smaller frequency and voltage deviations, has stronger recovery capabilities in the event of faults, and takes into account the stability of the system. Therefore, the present invention provides reliable guarantee for the safe and stable operation of the VSC-HVDC system under weak power grids.

[0162] The above description is only a preferred specific implementation manner of the present invention. It should be pointed out that the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An improved vector current control method for a flexible DC grid-connected system under a weak power grid, characterized in that: Includes steps: Establish a VSC-HVDC vector current control strategy model, where the outer control loop generates current reference for the inner current control loop to achieve independent control of active and reactive power; An improved vector current control strategy model IVCC based on VCC is constructed, and feedback control is performed through frequency changes to keep the frequency and voltage deviations within the threshold range; A small signal model of the flexible direct current grid-connected system after using IVCC is constructed, including the AC network, low-pass filter structure, phase-locked loop structure, and internal and external control loops. VSC-HVDC realizes coordinated control in different external control loops.

2. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 1 is characterized in that: Establish a VSC-HVDC vector current control strategy model, including a low-pass filter, an outer control loop, an inner current control loop, a current regulator, and a phase-locked loop; The synchronization angle of the VSC-HVDC vector current control strategy model is obtained by locking the phase-locked loop to the common coupling voltage point; the phase-locked loop output phase angle θ tracks the input phase angle θ0; The VSC-HVDC vector current control strategy model converts the output signal of the current regulator into a three-phase value as the input of the PWM.

3. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 2 is characterized in that: The VSC-HVDC vector current control strategy model has a hierarchical cascade structure, the internal current regulator is decoupled in the dq reference system, and the active and reactive power are independently controlled; The dynamic equation of the outer control loop is expressed as: Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qref are the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage of the common coupling point, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively; The dynamic equation of the inner current control loop is expressed as: Among them, e i1d and e i1q are the differences between the d-axis component and the q-axis component of the inner loop current control reference current and the d-axis component and the q-axis component of the VSC output current after passing through the filter, respectively. cd and v cq are the d-axis and q-axis components of the output voltage of the inner current control loop, respectively. pid With K iid are the proportional parameter and integral parameter of the inner ring d axis, K piq With K iiq are the proportional parameter and integral parameter of the inner loop q-axis respectively, L1 is the line inductance, and ω is the PLL output angular frequency.

4. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 3 is characterized in that: The improved vector current control strategy model IVCC with frequency change feedback control uses the grid frequency ω to describe the model tracking speed. IVCC adds a feedback link to improve the reference value of the outer loop and uses ω as the feedback signal. The dynamic equation of the outer loop control loop is as follows: Among them, e p and e v are defined as the active power and voltage offset, v tdm and v tqm are the d-axis and q-axis components of the voltage at the common coupling point after passing through the filter, i 1dm and i 1qm are the d-axis and q-axis components of the VSC output current after passing through the filter, i 1dref and i 1qref are the d-axis and q-axis reference currents of the inner loop current control, P0 is the rated active power, V0 is the rated voltage, K D1 and K D2 are the damping coefficients of the active power control loop and the voltage control loop, △ω m is the frequency deviation after IVCC feedback loop filtering, satisfying △ω is the grid frequency deviation, T ω is the feedback loop filtering time, K pp With K ip are the proportional parameter and integral parameter of the outer loop active power, K pv With K iv They are the proportional parameter and integral parameter of the outer loop voltage PI control respectively.

5. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 4 is characterized in that: The small signal model of the flexible DC grid-connected system after using IVCC is constructed, and the model is expressed as: in, is the differential of x, u is the input variable, and is determined by the active power reference value P ref and voltage reference value V ref Composition; x is composed of 17 variables x 01 ,x 02 ,x 03 ,x 04 ,x 05 ,x1,x2,x3,x4,θ,ω,i 1d ,i 1q ,i 2d ,i 2q ,v td ,v tq The state vector composed of 01 ,x 02 ,x 03 ,x 04 ,x 05 is the state quantity related to the filter circuit, as shown in formula (5); x1, x2, x3, x4 are the state quantities related to the control loop, as shown in formula (6); θ, ω are the phase angle of the phase-locked loop voltage and the tracking frequency respectively; i 1d ,i 1q are the d-axis and q-axis components of the VSC output current, respectively. 2d ,i 2q are the d-axis and q-axis components flowing into the infinite busbar, v td ,v tq is the d-axis and q-axis components of the common coupling point voltage, i 1d ,i 1q ,i 2d ,i 2q ,v td ,v tq Together they constitute the voltage and current state quantities of the AC network; Among them, T mvd and T mvq are the time constants of the low-pass filter filtering the d-axis and q-axis voltage components, respectively, mid and T miq are the time constants of the low-pass filter filtering the d-axis and q-axis current components, △ω is the grid frequency deviation, T ω The time for feedback loop filtering.

6. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 5 is characterized in that: For the small signal model, according to the damping coefficient K D1 and K D2 When the eigenvalue of the dominant oscillation mode changes, the damping ratio is introduced to adjust the damping coefficient K D1 and K D2 The feasible domain of is determined, and the damping ratio ζ is expressed as follows: Wherein, σ is the real part of the underdamped oscillation mode, and ω is the imaginary part of the underdamped oscillation mode.

7. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 1 is characterized in that: Coordinated control of VSC-HVDC in different external control loops, including: Damping coefficient K of active power control loop based on genetic algorithm D1 and the damping coefficient K of the voltage control loop D2 To optimize the parameter combination, first define the number of individuals in the population. Each individual in the population contains two genes, namely K D1 and K D2 , and initialize the population; Design the objective function F of the coordinated control model and the fitness function value F obtained in each iteration fit as follows: F(i,j)=λ1(S f (i,j)+S V (i,j))+λ2S s (i,j) (8), F fit (i)=minF(i,j),j=1,2,…,m (9), Among them, i is the number of population iterations, j represents the position of the individual in the population, λ1 and λ2 are defined as the weights of the dynamic response quality and the power consumption of the dynamic process, respectively, S f , S V and S s The three performance indicators of the objective function of coordinated control are S f and S V are the maximum deviations of the frequency and the voltage at the common coupling point after the disturbance, S s is the average change of active power and reactive power within t seconds after the disturbance, and m is the number of individuals in a population; By calling the coordinated control model, the objective function is calculated. After each iteration, the coordinated control model performs selection, crossover and mutation operations on the population based on the genetic algorithm to optimize the individuals in the population and enter the next iteration. When the fitness function F fit When convergence is achieved, the coordinated control iteration ends and the optimal solution K is output. D1 and K D2 , and obtain the optimal curve with the smallest maximum deviation between frequency and voltage and the smallest power change after the system is disturbed.

8. The improved vector current control method of the flexible DC grid-connected system under weak power grid according to claim 7 is characterized in that: According to the requirements for the dynamic response of the system, the constraints on the objective function F are set as follows: Among them, U N is the rated voltage. When the simulation result meets the constraint condition, the objective function calculation result F(i, j) is assigned. Otherwise, a penalty value is assigned to the objective function. After the calculation of F of m individuals is completed, the minimum F(i, j) is found according to formula (9) as the fitness function F of the current iteration. fit .

Images