A single-phase five-level rectifier non-delay power observation control method

Abstract

The application discloses a single-phase five-level rectifier non-delay power observation control method, which comprises the following steps: constructing orthogonal virtual voltage signals by using an improved generalized integral algorithm; constructing orthogonal virtual current signals according to a virtual current signal reconstruction algorithm; obtaining a non-delay power observer according to an instantaneous power calculation method; constructing a power prediction model of the single-phase five-level rectifier by using a forward Euler discretization method, designing an evaluation function to obtain a model prediction power control expression, and obtaining optimal modulation waves of the model prediction power control of the single-phase five-level rectifier in a dq coordinate system; the current inner loop of the single-phase five-level rectifier adopts model prediction power control based on the non-delay power observer, and the voltage outer loop adopts linear active disturbance rejection control; the single-phase five-level rectifier can realize fast tracking of the inner loop power, non-overshoot starting of the outer loop voltage and stronger anti-interference capability.

Description

Technical Field

[0001] This invention relates to the field of power electronics, and in particular to a time-delay-free power observation and control method for a single-phase five-level rectifier. Background Technology

[0002] Single-phase five-level rectifiers with coupled inductors have advantages such as fewer switching devices and no need for capacitor voltage balancing control, which simplifies modulation, facilitates control system design, and has broader application prospects. For rectifier control, a combined control method of voltage outer loop and current inner loop is generally adopted.

[0003] In recent years, model predictive power control (MMPC) has been increasingly applied to the inner loop control of rectifiers due to its advantages such as simplicity of implementation, flexible control, and ability to achieve multi-objective optimization control. However, compared to three-phase rectifiers, single-phase rectifiers lack a degree of freedom and cannot directly perform coordinate transformations, thus preventing the direct calculation of instantaneous power based on instantaneous reactive power theory. To achieve instantaneous power calculation, a virtual signal orthogonal to the grid-side voltage and current must be constructed. Traditional methods for constructing virtual orthogonal signals include the 1 / 4-cycle delay algorithm, the second-order generalized integral algorithm, and the Hilbert transform algorithm. However, these algorithms inevitably introduce a certain delay when constructing the virtual signal, which limits the power calculation speed of the inner loop when using model predictive power control for single-phase rectifiers, thus degrading the system's dynamic performance. Summary of the Invention

[0004] To address the poor control performance of model predictive power control (MMDC) in existing single-phase five-level rectifiers due to power calculation delays, this paper proposes a delay-free power observation control method for single-phase five-level rectifiers. This method derives the model predictive power control algorithm based on a mathematical model of the single-phase five-level rectifier. A virtual signal reconstruction (VSR) algorithm is proposed, combined with an improved second-order generalized integral (ISOGI) algorithm to construct a delay-free power observer. Building upon traditional model predictive power control, a model predictive power control for delayless power observer (MPPCFDPO) method is proposed for the current inner loop control. Furthermore, the voltage outer loop employs linear active disturbance rejection control (ADC), thereby eliminating the dynamic performance degradation caused by power calculation delays in the current inner loop model prediction and enhancing the anti-interference capability of the voltage outer loop.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for time-delay power observation and control of a single-phase five-level rectifier is disclosed. The inner current loop of the single-phase five-level rectifier employs model predictive power control based on a time-delay power observer, while the outer voltage loop employs linear active disturbance rejection control. This achieves rapid power tracking within the inner loop and overshoot-free start-up of the outer loop voltage. The time-delay power observation and control method includes the following steps:

[0007] S1: Obtain the grid-side voltage of the single-phase five-level rectifier, and use the improved generalized integral algorithm to construct an orthogonal virtual voltage signal to eliminate the influence of DC components and high-order harmonics in the input signal on the construction of the virtual voltage signal;

[0008] S2: Obtain the grid-side current of the single-phase five-level rectifier and construct an orthogonal virtual current signal based on the virtual current signal reconstruction algorithm;

[0009] S3: Obtain a time-delay power observer based on the instantaneous power calculation method to acquire the active and reactive power of the single-phase five-level rectifier;

[0010] S4: Based on the power model of a single-phase five-level rectifier in the dq rotating coordinate system, the forward Euler discretization method is used to construct the power prediction model of the single-phase five-level rectifier. The evaluation function is designed to obtain the model prediction power control expression, and the optimal modulation wave of the model prediction power control of the single-phase five-level rectifier in the dq coordinate system is obtained.

[0011] S5: Design an active disturbance rejection controller based on the system equation of the voltage outer loop, and obtain the reference active power value of the active disturbance rejection controller output. The linear active disturbance rejection controller includes a linear differential tracker, a linear extended state observer, and a linear state error feedback control law.

[0012] S6: Transform the predicted power control optimal modulation wave of the single-phase five-level rectifier model in the dq coordinate system into the αβ coordinate system to obtain the voltage modulation wave u. αin The signal is fed into the space vector modulation module for space vector modulation, generating a PWM signal to control the switching transistors of the single-phase five-level rectifier.

[0013] Furthermore, the improved generalized integral algorithm differs from the traditional second-order generalized integral algorithm in that it adds a subtraction channel containing a low-pass filter, the transfer function of which is G. LPF As shown in equation (A-1):

[0014]

[0015] In equation (A-1), τ is a parameter related to the cutoff frequency of the low-pass filter. At this point, if the input signal (grid-side voltage u)... s If the output contains a DC component, then the quadrature virtual voltage signal u is... sβ It no longer contains a DC component.

[0016] Furthermore, the specific steps for constructing orthogonal virtual current signals based on the virtual current signal reconstruction algorithm include:

[0017] S31: Based on the grid-side voltage u of the single-phase five-level rectifier s Obtain the grid-side voltage u from the phasor diagrams in the αβ and dq coordinate systems. s Angle with α axis And the phase angle θ transformed by dq / αβ is calculated as shown in equation (A-2):

[0018]

[0019] In equation (A-2), δ is the angle between the grid-side voltage phasor and the d-axis;

[0020] S32: Since the phase-locked loop can track the phase of the grid-side voltage, when the system is stable, the grid-side voltage u s Angle with α axis The phase angle θ is equal to that of the dq / αβ transformation, and the grid-side voltage u s d-axis component u in the dq coordinate system sd and q-axis component u sq The calculation is as shown in equation (A-3):

[0021]

[0022] In formula (A-3), u sm Indicates the magnitude of the grid-side voltage;

[0023] Based on mathematical geometric relationships and trigonometric function relationships, as shown in equation (A-4), calculate cosθ and sinθ:

[0024]

[0025] In formula (A-4), u sα Indicates grid-side voltage u s The α-axis component in the αβ coordinate system, u sα =u s ;u sβ Indicates grid-side voltage u s The β-axis component in the αβ coordinate system, i.e., the orthogonal virtual voltage signal;

[0026] S33: Based on the grid-side current i of the single-phase five-level rectifier sThe spatial distribution relationship between the αβ and dq coordinate systems is shown in equation (A-5), and the orthogonal virtual current signal i is obtained. sβ :

[0027]

[0028] In equation (A-5), i sα Indicates grid-side current i s The α-axis component in the αβ coordinate system, i sα =i s i sβ Indicates grid-side current i s β-axis component in the αβ coordinate system; i sd Indicates grid-side current i s In the d-axis component of the dq coordinate system, i sq Indicates grid-side current i s q-axis components in the dq coordinate system;

[0029] From equation (A-5), we can obtain equation (A-6):

[0030] i sβ =i sd sinθ+i sq cosθ(A-6)

[0031] Substituting equation (A-4) into equation (A-6), the orthogonal virtual current signal i can be calculated as shown in equation (A-7). sβ :

[0032]

[0033] Furthermore, according to the single-phase instantaneous power calculation method, the calculation method for the active power and reactive power of the single-phase five-level rectifier is shown in equation (A-8):

[0034]

[0035] In formula (A-8), u sd i sd These represent the grid-side voltage u. s and grid-side current i s d-axis components in the dq coordinate system; u sq i sq These represent the grid-side voltage u. s and grid-side current i s The q-axis component in the dq coordinate system.

[0036] Furthermore, the power model of a single-phase five-level rectifier in the dq rotating coordinate system is shown in equation (A-9):

[0037]

[0038] In equation (A-9), L s Indicates the grid-side inductance, u din u qin Let represent the d-axis component and q-axis component of the optimal voltage modulation wave under model predictive control, respectively, and ω represent the angular frequency.

[0039] The power prediction model of a single-phase five-level rectifier is constructed using the forward Euler discretization method, as shown in equation (A-10):

[0040]

[0041] The design of the evaluation function is a crucial step in the model prediction algorithm. Its purpose is to minimize the error between the predicted power and the commanded power in order to obtain an accurate voltage modulation waveform u. αin .

[0042] Furthermore, the evaluation function is shown in equation (A-11):

[0043] J = [P] ref -P(k+1)] 2 +[Q ref -Q(k+1)] 2 (A-11)

[0044] In equation (A-11), P ref Q ref P(k+1) and Q(k+1) represent the given active power and given reactive power along the dq axis, respectively, and the predicted active power and predicted reactive power, respectively.

[0045] To minimize the error between the commanded power and the predicted power, and to achieve error-free power tracking, u is calculated for the evaluation function respectively. din (k) and u qin Take the partial derivative of (k) and set it to 0, as shown in equation (A-12):

[0046]

[0047] Based on equations (A-10) to (A-12), we can obtain equation (A-13):

[0048]

[0049] The obtained single-phase five-level rectifier model predicts the optimal modulation waveform for power control. din (k) and u qin (k) As in equation (A-14):

[0050]

[0051] Furthermore, based on the principle of linear active disturbance rejection control algorithm, the calculation method for obtaining the reference active power value of the active disturbance rejection controller output is as shown in equation (A-15):

[0052] P ref =P0-z2 / b(A-15)

[0053] In equation (A-15), b is the compensation coefficient, and the selection of the value of b will affect the compensation strength of the linear active disturbance rejection control.

[0054] The function of the linear state error feedback control law is to amplify the error term, thereby accelerating the tracking speed of the DC side voltage and thus improving dynamic performance.

[0055] P0 is obtained by designing a linear state error feedback control law, as shown in equation (A-16):

[0056]

[0057] In equation (A-16), k represents the error feedback coefficient, x1 represents the transient process of arranging the DC side voltage setpoint, and z1 represents the observed value of the DC side output voltage of the rectifier.

[0058] To ensure that the DC-side output voltage can respond quickly and without overshoot during startup, a linear tracking differentiator needs to be designed for the DC-side reference voltage so that the output voltage can transition to the voltage reference value quickly and smoothly.

[0059] Furthermore, as shown in equation (A-17), x1 is obtained by designing a linear differential tracker:

[0060]

[0061] In equation (A-17), r represents an adjustable parameter, the value of which determines the speed of voltage tracking, and x2 represents the derivative of x1. Indicates the DC-side reference voltage;

[0062] The purpose of designing the extended state observer is to observe internal and external disturbances of the system and track the DC-side output voltage in real time.

[0063] Furthermore, as shown in equation (A-18), a linearly extended state observer is designed to obtain z1:

[0064]

[0065] In equation (A-18), z2 represents the observed value of the disturbance, β1 and β2 represent adjustable gain coefficients, β1 is related to the speed of the control response, and β2 is related to the control error. The values ​​of β1 and β2 are tuned according to the bandwidth of the extended state observer.

[0066] Compared with the prior art, the present invention has the following technical effects:

[0067] (1) The power observer based on virtual signal reconstruction has a faster power calculation capability than the traditional instantaneous power observer. It can track power mutations without delay, so that the power in the inner loop of the rectifier can be quickly restored to stability, realize the rapid tracking of the power in the inner loop of a single-phase five-rectifier, and improve the dynamic performance of the control system.

[0068] (2) The use of linear active disturbance rejection control instead of traditional PI controller in the voltage outer loop can achieve overshoot-free start-up of the DC side voltage of the rectifier, as well as rapid recovery and stable output of the DC side voltage when the load is disturbed and the grid voltage drops. Attached Figure Description

[0069] Figure 1 This is a topology diagram of the single-phase five-level rectifier of the present invention;

[0070] Figure 2 This is a schematic diagram of a single-phase five-level rectifier control system.

[0071] Figure 3 Here is a block diagram of the ISOGI algorithm;

[0072] Figure 4 For grid-side voltage u s Vector diagrams in different coordinate systems;

[0073] Figure 5 Here is a block diagram of the VSR algorithm;

[0074] Figure 6 Schematic diagram of a no-delay power observer;

[0075] Figure 7 This is a schematic diagram of the LADRC algorithm.

[0076] Figure 8 The steady-state waveform of the rectifier

[0077] Figure 9 This is a comparison chart of power calculation performance based on different algorithms;

[0078] Figure 10 THD diagram of grid-side current;

[0079] Figure 11 The output voltage U under different operating conditions and different control algorithms dc Waveform comparison chart;

[0080] Figure 12 A comparison chart of the inner-loop dynamic performance under different control algorithms;

[0081] Figure 13 The waveform diagram is shown in the power supply voltage drop test.

[0082] Figure 14 The output voltage U when the load suddenly increases dc and input current i s Waveform diagram. Detailed Implementation

[0083] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the invention, any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art to all other embodiments obtained without creative effort should be included within the protection scope of the present invention.

[0084] like Figure 1 The diagram shown is a topology diagram of a single-phase five-level rectifier. s L s These are the grid-side voltage and grid-side inductance, respectively. s For grid-side current, L b L c Two coupled inductors with mutual inductance M and i b i c These represent the currents flowing through the two coupled inductors, C is the DC-side supporting capacitor, R is the DC-side load, and U... dc The DC bus voltage is used for power switching transistors S1 to S6, and u ad Let T1, T2, and T3 represent the voltage between points a and d, i.e., the input voltage. In each bridge arm of the rectifier, the upper and lower switches operate complementaryly, with only two switching states. Let T1, T2, and T3 represent the switching states of the three bridge arms, respectively. The logic switching function is shown in equation (S-1):

[0085]

[0086] Based on the rectifier's operating state and combined with equation (S-1), we can obtain:

[0087]

[0088] In equation (S-2), S represents the logic switching function, u in This indicates the input voltage.

[0089] like Figure 2The diagram shows a schematic of a single-phase five-level rectifier control system. The inner current loop of the single-phase five-level rectifier employs model predictive power control based on a no-delay power observer, while the outer voltage loop employs linear active disturbance rejection control. This achieves rapid power tracking within the inner loop and overshoot-free start-up of the outer loop voltage. The no-delay power observation control method for the single-phase five-level rectifier specifically includes the following steps:

[0090] S1: Obtain the grid-side voltage of the single-phase five-level rectifier, and use the improved generalized integral algorithm to construct an orthogonal virtual voltage signal in the αβ coordinate system to eliminate the influence of DC components and higher harmonics in the input signal on the construction of the virtual voltage signal;

[0091] S2: Obtain the grid-side current of the single-phase five-level rectifier and construct an orthogonal virtual current signal in the αβ coordinate system according to the virtual current signal reconstruction algorithm;

[0092] S3: The virtual voltage and current signals in the αβ coordinate system are transformed to the dq coordinate system through αβ / dq transformation. The power observer without delay is obtained according to the instantaneous power calculation method to obtain the active power and reactive power of the single-phase five-level rectifier.

[0093] S4: Based on the power model of a single-phase five-level rectifier in the dq rotating coordinate system, the forward Euler discretization method is used to construct the power prediction model of the single-phase five-level rectifier. The evaluation function is designed to obtain the model prediction power control expression, and the optimal modulation wave of the model prediction power control of the single-phase five-level rectifier in the dq coordinate system is obtained.

[0094] S5: Design an active disturbance rejection controller based on the system equation of the voltage outer loop, and obtain the reference active power value of the active disturbance rejection controller output. The linear active disturbance rejection controller includes a linear differential tracker, a linear extended state observer, and a linear state error feedback control law.

[0095] S6: Transform the predicted power control optimal modulation wave of the single-phase five-level rectifier model in the dq coordinate system into the αβ coordinate system to obtain the voltage modulation wave u. αin The signal is fed into the space vector modulation module for space vector modulation, generating a PWM signal to control the switching transistors of the single-phase five-level rectifier.

[0096] like Figure 3 As shown, the difference between the improved generalized integral algorithm and the traditional second-order generalized integral algorithm lies in the addition of a subtraction channel containing a low-pass filter. The transfer function G of the low-pass filter is... LPF As shown in equation (S-3):

[0097]

[0098] In equation (S-3), τ is a parameter related to the cutoff frequency of the low-pass filter. At this point, if the input signal (grid-side voltage u)... s If the output contains a DC component, then the quadrature virtual voltage signal u is... sβ It no longer contains a DC component. This is because the orthogonal virtual voltage signal u... sβ If the frequency is 50Hz, then the cutoff frequency of the low-pass filter can be preferably 50Hz.

[0099] The specific steps for constructing orthogonal virtual current signals based on the virtual current signal reconstruction algorithm include:

[0100] S31: As Figure 4 As shown, based on the grid-side voltage u of the single-phase five-level rectifier s Obtain the grid-side voltage u from the phasor diagrams in the αβ and dq coordinate systems. s Angle with α axis And the phase angle θ transformed by dq / αβ is calculated as shown in equation (S-4):

[0101]

[0102] In equation (S-4), δ is the angle between the grid-side voltage phasor and the d-axis;

[0103] S32: Since the phase-locked loop can track the phase of the grid-side voltage, when the system is stable, the grid-side voltage u s Angle with α axis The phase angle θ is equal to that of the dq / αβ transformation, and the grid-side voltage u s d-axis component u in the dq coordinate system sd and q-axis component u sq The calculation is as shown in equation (S-5):

[0104]

[0105] In equation (S-5), u sm Indicates the magnitude of the grid-side voltage;

[0106] according to Figure 4 As shown, cosθ and sinθ are calculated using mathematical geometric relationships and trigonometric function relationships, as shown in equation (S-6):

[0107]

[0108] In equation (S-6), u sα Indicates grid-side voltage u s The α-axis component in the αβ coordinate system, u sα =u s ;u sβ Indicates grid-side voltage u sThe β-axis component in the αβ coordinate system, i.e., the orthogonal virtual voltage signal;

[0109] S33: Based on the grid-side current i of the single-phase five-level rectifier s The spatial distribution relationship between the αβ and dq coordinate systems is shown in equation (S-7), and the orthogonal virtual current signal i is obtained. sβ :

[0110]

[0111] In equation (S-7), i sα Indicates grid-side current i s The α-axis component in the αβ coordinate system, i sα =i s i sβ Indicates grid-side current i s β-axis component in the αβ coordinate system; i sd Indicates grid-side current i s In the d-axis component of the dq coordinate system, i sq Indicates grid-side current i s q-axis components in the dq coordinate system;

[0112] From equation (S-7), we can obtain equation (S-8):

[0113] i sβ =i sd sinθ+i sq cosθ(S-8)

[0114] Substituting equation (S-6) into equation (S-8), the orthogonal virtual current signal i can be calculated as shown in equation (S-9). sβ :

[0115]

[0116] like Figure 5 As shown, the implementation block diagram of the β-axis virtual current signal can be obtained through equation (S-9).

[0117] According to the single-phase instantaneous power calculation method, the calculation method for the active power and reactive power of the single-phase five-level rectifier is shown in equation (S-10):

[0118]

[0119] In formula (S-10), u sd i sd These represent the grid-side voltage u. s and grid-side current i s d-axis components in the dq coordinate system; u sq i sqThese represent the grid-side voltage u. s and grid-side current i s The q-axis component in the dq coordinate system.

[0120] The virtual current signal along the β-axis can be constructed using equation (S-9). Based on the above analysis and combined with the instantaneous power calculation equation (S-10), a power observer without delay can be obtained, such as... Figure 6 As shown, this allows for the instantaneous power of a single-phase five-level rectifier to be observed without delay.

[0121] The power model of a single-phase five-level rectifier in the dq rotating coordinate system is shown in equation (S-11):

[0122]

[0123] In formula (S-11), L s Indicates the grid-side inductance, u din u qin Let represent the d-axis component and q-axis component of the optimal voltage modulation wave under model predictive control, respectively, and ω represent the angular frequency.

[0124] The power prediction model of a single-phase five-level rectifier is constructed using the forward Euler discretization method, as shown in equation (S-12):

[0125]

[0126] The design of the evaluation function is a crucial step in the model prediction algorithm. Its purpose is to minimize the error between the predicted power and the commanded power in order to obtain an accurate voltage modulation waveform u. αin .

[0127] The evaluation function is shown in equation (S-13):

[0128] J = [P] ref -P(k+1)] 2 +[Q ref -Q(k+1)] 2 (S-13)

[0129] In equation (S-13), P ref Q ref P(k+1) and Q(k+1) represent the given active power and given reactive power along the dq axis, respectively, and the predicted active power and predicted reactive power, respectively.

[0130] To minimize the error between the commanded power and the predicted power, and to achieve error-free power tracking, u is calculated for the evaluation function respectively. din (k) and u qin Take the partial derivative of (k) and set it to 0, as shown in equation (S-14):

[0131]

[0132] Based on equations (S-12) to (S-14), we can obtain equation (S-15):

[0133]

[0134] The obtained single-phase five-level rectifier model predicts the optimal modulation waveform for power control. din (k) and u qin (k) As in equation (S-16):

[0135]

[0136] To overcome the problems of excessive start-up overshoot and poor anti-interference capability of traditional PI controllers used in the outer voltage loop of a single-phase five-level rectifier, a linear auto disturbance rejection control (LADRC) device is introduced into the outer voltage loop to improve the control performance of the DC-side voltage of the single-phase five-level rectifier. The DC-side voltage equation of the single-phase five-level rectifier is a first-order equation; therefore, the outer voltage loop can be designed as a first-order system. For example... Figure 7 The diagram shown is a schematic of the LADRC algorithm, where LTD is the linear differential tracker, LESO is the linear extended state observer, and LSEF is the linear state error feedback control law.

[0137] Based on the principle of linear active disturbance rejection control algorithm, the calculation method for obtaining the reference active power value of the active disturbance rejection controller output is as shown in equation (S-17):

[0138] P ref =P0-z2 / b(S-17)

[0139] In equation (S-17), b is the compensation coefficient, and the selection of the value of b will affect the compensation strength of the linear active disturbance rejection control.

[0140] The function of the linear state error feedback control law is to amplify the error term, thereby accelerating the tracking speed of the DC side voltage and thus improving dynamic performance.

[0141] P0 is obtained by designing a linear state error feedback control law, as shown in equation (S-18):

[0142]

[0143] In equation (S-18), k represents the error feedback coefficient, x1 represents the transient process of arranging the DC side voltage setpoint, and z1 represents the observed value of the DC side output voltage of the rectifier.

[0144] To ensure that the DC-side output voltage can respond quickly and without overshoot during startup, a linear tracking differentiator needs to be designed for the DC-side reference voltage so that the output voltage can transition to the voltage reference value quickly and smoothly.

[0145] As shown in equation (S-19), x1 is obtained by designing a linear differential tracker:

[0146]

[0147] In equation (S-19), r represents an adjustable parameter whose value determines the speed of voltage tracking, and x2 represents the derivative of x1. Indicates the DC-side reference voltage;

[0148] The purpose of designing the extended state observer is to observe internal and external disturbances of the system and track the DC-side output voltage in real time.

[0149] As shown in equation (S-20), a linearly extended state observer is designed to obtain z1:

[0150]

[0151] In equation (S-20), z2 represents the observed value of the disturbance, β1 and β2 represent adjustable gain coefficients, β1 is related to the speed of the control response, and β2 is related to the control error. The values ​​of β1 and β2 are tuned according to the bandwidth of the extended state observer.

[0152] like Figure 8 The figure shows the steady-state waveform of the rectifier under the MPPCFDPO control strategy, with a given active power of 10kW and a given reactive power of 0. It can be seen that the rectifier input can generate a five-level circuit, corresponding to the working principle of the five-level rectifier described earlier. Furthermore, the grid-side voltage and current of the rectifier can achieve in-phase operation, meaning the rectifier can achieve unity power factor operation. Simultaneously, the grid-side voltage and current exhibit good sinusoidal characteristics. Therefore, the rectifier demonstrates excellent steady-state performance under this algorithm.

[0153] To fairly compare the power calculation capabilities of traditional power observers and no-delay power observers, the virtual voltage signals in the simulation were constructed using the ISOGI algorithm, while the virtual current signals were constructed using the TD algorithm, SOGI algorithm, and the VSR algorithm proposed in this paper, respectively.

[0154] like Figure 9 The diagram shows a comparison of power calculation performance based on a traditional power observer and a no-delay power observer. The given active power abruptly changes from 10kW to 18kW within 0.25s. (Comparison) Figure 9 (a) Figure 9 (b) and Figure 9 (c) It can be seen that all three algorithms can construct the β-axis virtual current signal and track power. However, the traditional TD algorithm and SOGI algorithm cannot obtain the virtual current signal in time after a power change, resulting in inaccurate power calculation in the inner loop and a delay, which affects the dynamic performance of the system. The proposed VSR algorithm can construct the virtual current signal without delay after a given power change, making the inner loop power calculation performance superior and the dynamic performance better.

[0155] Figure 10 The Fourier analysis results of the grid-side current under the three algorithms show that, as Figure 10 (a) Figure 10 (b) When using the traditional TD algorithm and the SOGI algorithm respectively, the total harmonic distortion (THD) of the grid-side current is 3.05% and 3.02%, respectively. Figure 10 (c) The grid-side current THD is 3.01% when using the VSR algorithm. This result shows that the steady-state performance of the VSR algorithm is as good as that of the traditional TD algorithm and the SOGI algorithm.

[0156] To verify the effectiveness and control performance of the outer-loop LADRC algorithm, simulation models of MPPCFDPO+PI and MPPCFDPO+LADRC control systems were built in simulation software. For example... Figure 11 The figure shows the DC side voltage U under different operating conditions. dc and grid-side current i s The experimental waveform comparison Figure 11 (a) shows the start-up waveforms of the rectifier DC-side voltage when the two control algorithms are applied respectively. It can be seen that both algorithms can make the DC-side voltage stabilize. However, the PI control has overshoot and takes a long time to reach steady state. In contrast, the LADRC control does not have overshoot and can reach steady state very quickly. Therefore, LADRC can make the DC-side voltage converge to the given value more accurately and quickly, and has a better control effect. Figure 11 (b) Simulated waveforms of the DC-side voltage are given for the two control algorithms when the given voltage changes from 500V to 550V. As can be seen from the figure, both algorithms can eventually track the given value, but the LADRC algorithm tracks much faster than the PI algorithm, and its dynamic performance is superior. Figure 11 (c) The simulated waveform of DC side voltage during load change is given. It can be seen that the LADRC control algorithm has smaller parameter perturbations and stronger anti-interference ability compared with the PI algorithm.

[0157] Therefore, compared with MPPCFDPO+PI control, MPPCFDPO+LADRC can enable the outer loop voltage to have better tracking performance and anti-interference capability.

[0158] like Figure 12 The figure shows a comparison of the inner-loop power tracking performance of three algorithms under the condition of a 40% sudden change in active power P*. Figure 12 (a) It can be seen that the power observer based on the TD algorithm requires 5ms for the inner loop active power to re-track the given active power after a sudden change in the given active power. The reactive power requires the same amount of time to transition to a steady state and has a large power fluctuation. Figure 12 (b) In the power observer based on the SOGI algorithm, after a sudden change in given active power, the inner-loop active power takes approximately 7ms to re-track the given value, and the reactive power also takes the same amount of time to stabilize, exhibiting significant fluctuations in reactive power. Figure 12 (c) The time-delay power observer constructed based on the VSR algorithm can enable the inner loop active power to track the given value immediately after a given active power change, with almost no adjustment time required. Compared with the traditional algorithm, the time required is reduced by 5 to 7 ms, and the reactive power has almost no fluctuation and remains in a stable state.

[0159] Therefore, the time-delay-free power observer constructed based on the VSR algorithm enables model predictive power control to have superior dynamic performance compared to traditional power observers.

[0160] like Figure 13 The figure shows the experimental waveform during a sudden change in grid voltage, where the voltage source amplitude drops abruptly from 311V to 255V. Figure 13 (a) Figure 13 In (b), the outer voltage loop employs both the PI and LADRC algorithms, while the inner loop uses the MPPCFDPO algorithm. It can be seen that under both control algorithms, a voltage drop in the power supply will cause a decrease in the DC-side voltage U. dc Fluctuations, and at the same time, the grid-side current i s It will gradually increase to compensate for the impact of grid voltage drops on DC-side power. The difference lies in the DC-side voltage U during PI control. dc Voltage and grid-side current i s It takes 80ms to transition to steady state, which is quite time-consuming, and the DC-side voltage fluctuates by 8.8% relative to the given value. In contrast, with LADRC control, the DC-side voltage U... dc and grid-side current i s It can transition to steady state in just 24ms, while the DC side voltage fluctuation is 4.4%, which is 4.4% lower than that of PI control, and the control time is reduced by 56ms.

[0161] like Figure 14 The figure shows the DC side voltage U when the load changes abruptly. dc and grid-side current i sThe waveform shows the load resistance abruptly changing from 50Ω to 25Ω. Figure 14 (a) It can be seen that after the load mutation, when the PI algorithm is applied, U dc It took approximately 49ms to regain stability, with the voltage drop being 6.4% compared to the given value, and i s It also takes the same amount of time to stabilize, which is quite long; by Figure 14 (b) It can be seen that when using the LADRC algorithm, U dc It was able to stabilize in just about 9ms, with a voltage drop of 1.6%, which is 4.8% less than the voltage drop of the PI algorithm. Meanwhile, the grid-side current i... s It stabilized again with almost no time adjustment.

[0162] In summary, compared with MPPCFDPO+PI control, MPPCFDPO+LADRC can estimate and compensate for external disturbances in real time, giving the DC side voltage a better ability to resist load disturbances and power supply voltage fluctuations.

Claims

1. A method for time-delay power observation and control of a single-phase five-level rectifier, characterized in that, The inner current loop of the single-phase five-level rectifier adopts model predictive power control based on a no-delay power observer, and the outer voltage loop of the single-phase five-level rectifier adopts linear active disturbance rejection control to achieve rapid tracking of the inner loop power and overshoot-free start-up of the outer loop voltage. The no-delay power observation control method includes the following steps: S1: Obtain the grid-side voltage of the single-phase five-level rectifier and construct an orthogonal virtual voltage signal using an improved generalized integral algorithm; S2: Obtain the grid-side current of the single-phase five-level rectifier and construct an orthogonal virtual current signal based on the virtual current signal reconstruction algorithm; The specific steps include: S21: Based on the grid-side voltage u of the single-phase five-level rectifier s Obtain the grid-side voltage u from the phasor diagrams in the αβ and dq coordinate systems. s Angle with α axis and dq / αβ transformation phase angle The calculation process is as shown in equation (1): ； （1） In equation (1), δ is the angle between the grid-side voltage phasor and the d-axis; S22: When the system is stable, the grid-side voltage u s Angle with α axis Phase angle with dq / αβ transformation They are equal. Based on mathematical geometric relationships and trigonometric function relationships, as shown in equation (2), calculate... as well as : ； （2） In equation (2), u sα Indicates grid-side voltage u s The α-axis component, u, in the αβ coordinate system sα =u s ;u sβ Indicates grid-side voltage u s The β-axis component in the αβ coordinate system, i.e., the orthogonal virtual voltage signal; S23: Based on the grid-side current i of the single-phase five-level rectifier s The spatial distribution relationship between the αβ and dq coordinate systems is shown in equation (3), and the orthogonal virtual current signal i is obtained. sβ : ； （3） In equation (3), i sα Indicates grid-side current i s The α-axis component in the αβ coordinate system, i sα =i s i sβ Indicates grid-side current i s β-axis components in the αβ coordinate system; Indicates grid-side current i s The d-axis component in the dq coordinate system Indicates grid-side current i s q-axis components in the dq coordinate system; S3: Obtain a time-delay power observer based on the instantaneous power calculation method to acquire the active and reactive power of the single-phase five-level rectifier; S4: Based on the power model of a single-phase five-level rectifier in the dq rotating coordinate system, the forward Euler discretization method is used to construct the power prediction model of the single-phase five-level rectifier. The evaluation function is designed to obtain the model prediction power control expression, and the optimal modulation wave of the model prediction power control of the single-phase five-level rectifier in the dq coordinate system is obtained. S5: Design an active disturbance rejection controller based on the system equation of the voltage outer loop, and obtain the reference active power value of the active disturbance rejection controller output. The linear active disturbance rejection controller includes a linear differential tracker, a linear extended state observer, and a linear state error feedback control law. S6: The optimal modulation wave for power control of the single-phase five-level rectifier model in the dq coordinate system is transformed into the αβ coordinate system to obtain the voltage modulation wave, which is then sent to the space vector modulation module for space vector modulation to generate a PWM signal to control the switching transistors of the single-phase five-level rectifier.

2. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 1, characterized in that, The difference between the improved generalized integral algorithm and the traditional second-order generalized integral algorithm lies in the addition of a subtraction channel containing a low-pass filter. The transfer function of the low-pass filter... As shown in equation (4): ； （4） In equation (4), These are parameters related to the cutoff frequency of the low-pass filter.

3. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 2, characterized in that, The calculation method for the active power and reactive power of the single-phase five-level rectifier is shown in equation (5): ； （5） In equation (5), u sd i sd These represent the grid-side voltage u. s and grid-side current i s The d-axis components in the dq coordinate system; u sq i sq These represent the grid-side voltage u. s and grid-side current i s The q-axis component in the dq coordinate system.

4. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 3, characterized in that, The power model of a single-phase five-level rectifier in the dq rotating coordinate system is shown in equation (6): ； （6） In equation (6), Indicates the grid-side inductance. , Let represent the d-axis and q-axis components of the optimal voltage modulation wave under model predictive control, respectively. Indicates angular frequency; The power prediction model of a single-phase five-level rectifier is constructed using the forward Euler discretization method, as shown in equation (7): （7）。 5. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 4, characterized in that, The evaluation function is shown in equation (8): ； （8） In equation (8), , These represent the given active power and given reactive power along the dq axis, respectively. , These represent the predicted active power and the predicted reactive power, respectively. Calculate the evaluation function respectively and The partial derivatives are set to 0, as in equation (9): ； （9） Based on equations (7) to (9), we can obtain equation (10): ；（10） The obtained single-phase five-level rectifier model predicts the optimal modulation waveform for power control. as well as As shown in equation (11): （11）。 6. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 5, characterized in that, The calculation method for obtaining the reference active power value of the active disturbance rejection controller output is as shown in equation (12): ； （12） In equation (12), b is the compensation coefficient. The selection of the value of b will affect the compensation strength of the linear active disturbance rejection control; the linear state error feedback control law is designed to obtain... As in equation (13): ； （13） In equation (13), k represents the error feedback coefficient, x1 represents the transient process of the DC side voltage setpoint arrangement, and z1 represents the observed value of the DC side output voltage of the rectifier.

7. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 6, characterized in that, As shown in equation (14), a linear differential tracker is designed to obtain x1: ； （14） In equation (14), r represents an adjustable parameter whose value determines the speed of voltage tracking, and x2 represents the derivative of x1. This indicates the DC-side reference voltage.

8. The method for non-delayed power observation and control of a single-phase five-level rectifier according to claim 7, characterized in that, As shown in equation (15), a linearly extended state observer is designed to obtain z1: ； （15） In equation (15), z2 represents the observed value of the disturbance, β1 and β2 represent adjustable gain coefficients, β1 is related to the speed of the control response, and β2 is related to the control error. The values ​​of β1 and β2 are tuned according to the bandwidth of the extended state observer.

Images