A grid voltage observation method based on dual quasi-resonant filter

Abstract

The present invention discloses a grid voltage observation method based on dual quasi-resonant filters. The method comprises a sliding mode observer, a series-connected dual quasi-resonant filter, and an adaptive compensation link. The sliding mode observer is used to perform preliminary grid voltage observation, the first quasi-resonant filter is used to suppress grid voltage sliding mode noise and DC offset, and the second quasi-resonant filter and the adaptive compensation link are used to achieve grid voltage observation that is resistant to grid frequency offset. After adaptive compensation, the observed grid voltage is ultimately obtained. The present invention uses a quasi-resonant filter instead of a dual low-pass filter to overcome the effects of sampled DC offset on grid voltage observation. Furthermore, the present invention uses two identical quasi-resonant filters in series to suppress sliding mode noise and compensate for grid voltage phase and amplitude, thereby achieving accurate grid voltage observation without requiring a known grid angular frequency, thus overcoming the effects of grid frequency offset.

Description

Technical Field

[0001] The present invention relates to the technical field of power electronics, and in particular to a grid voltage observation method based on a dual quasi-resonance filter. Background Art

[0002] Grid-connected inverters are key devices that connect renewable energy generation systems, such as solar and wind power, to the power grid. In recent years, many researchers have conducted research on advanced control technologies for grid-connected inverters to improve their operational reliability and stability.

[0003] Common grid-connected inverter control technologies are mainly divided into direct power control and vector control. In practical applications, these control methods require the use of multiple high-precision voltage and current sensors to collect signals such as grid voltage and current. However, the use of sensors not only increases system hardware costs but also reduces system operational reliability. Therefore, how to reduce the use of sensors and achieve current sensorless or voltage sensorless control of grid-connected inverters has become a focus of widespread attention in industry and academia.

[0004] To achieve grid voltage observation, the paper [Power Predictive Control of Sensorless PWM Rectifier Based on Adaptive Neural Network Observation [J]. Proceedings of the CSEE, 2021, 41(03):1135-1146.] designed a neural network-based grid voltage observer, which can accurately achieve grid voltage observation. However, this type of observer requires online adaptive optimization of multiple parameters. The paper [Observer-based grid-voltage sensorless synchronization and control of a VSI-LCL tied to an unbalanced grid [J]. IEEE Transactions on Industrial Electronics, 2019, 66(7):4972-4981.] proposed an adaptive state observer for estimating the positive and negative sequence components of the grid voltage under unbalanced grid conditions. However, the design of this observer is relatively complex. Considering the advantages of sliding-mode observers, such as strong robustness and simple design, a method for observing unbalanced grid voltage based on a high-order sliding-mode observer is proposed in the paper [Sliding-mode observer based voltage-sensorless model predictive power control of PWM rectifier under unbalanced grid conditions[J].IEEE Transactions on Industrial Electronics, 2018, 65(7):5550-5560.]. This method not only observes the positive and negative sequence components of the grid voltage but also observes and suppresses DC bias. However, this method is sensitive to grid frequency offset.

[0005] In order to solve the problem that grid voltage observation is sensitive to grid frequency offset, the literature [Guo Leilei, et al. Model predictive control of grid-connected inverter without grid voltage sensor [J]. Transactions of China Electrotechnical Society, 2020, 35(12): 2612-2622.] proposed a grid voltage observation method based on sliding mode observer and dual low-pass filter (LPF). This method realizes accurate grid voltage observation without knowing the precise grid angular frequency by designing a dual LPF series structure. However, the LPF lacks the ability to suppress DC bias, which makes this method sensitive to DC bias and amplifies the influence of DC bias. The literature [A method for observing back electromotive force of permanent magnet synchronous motor, application number: 201910376079.0] proposed a method for observing back electromotive force of permanent magnet synchronous motor. The dual LPF method is used to compensate for the back electromotive force phase delay and amplitude attenuation caused by the first LPF according to the influence of the second LPF on the back electromotive force phase and amplitude. It can achieve accurate observation of back electromotive force, but this method cannot eliminate the influence of DC bias.

[0006] Unlike LPFs, quasi-resonant filters have bandpass filtering characteristics, which not only suppress high-frequency harmonics but also eliminate the effects of DC bias. Therefore, quasi-resonant filters are widely used in fields such as harmonic and DC bias suppression in grid-connected inverters. Summary of the Invention

[0007] In order to realize grid voltage observation of grid-connected inverters and overcome the influence of grid frequency offset and DC bias, the present invention proposes a grid voltage observation method based on a dual quasi-resonant filter, which realizes DC bias suppression and adopts a dual quasi-resonant filter in series instead of a single quasi-resonant filter, thereby realizing accurate grid voltage observation without knowing the precise grid angular frequency and overcoming the influence of grid frequency offset.

[0008] The technical solutions adopted by the present invention to solve the technical problems are as follows:

[0009] A grid voltage observation method based on a dual quasi-resonant filter, the steps of which are as follows:

[0010] Step 1: The mathematical model of the grid-connected inverter in the two-phase stationary αβ coordinate system is:

[0011] Where, e αg 、e βg are the two components of the three-phase grid voltage in the αβ coordinate system, u α 、u β and i α 、i β are the components of the inverter output voltage and grid current in the αβ coordinate system respectively; based on this, the following sliding mode observer is designed Where, is the grid current observed by the sliding mode observer; K is the sliding mode gain; sgn() is the sign function; based on this, the current observation error equation is obtained Where, is the component of the error between the current observation value and the actual value in the αβ coordinate system;

[0012] Step 2: After the sliding mode observer obtained in step 1 converges, The grid voltage can be obtained by eliminating some of the current observation error equations. The sign function causes the observed grid voltage to contain a large amount of sliding mode noise; if the current or voltage signal in the sliding mode observer obtained in step 1 contains a DC bias, the obtained grid voltage will contain a DC component;

[0013] Step 3: Use a quasi-resonant filter to suppress the sliding mode noise and DC bias contained in the grid voltage observed in step 2, so as to obtain the grid voltage obtained by the first quasi-resonant filter. Where s is the Laplace operator, ω c1 is the resonant bandwidth of the quasi-resonant filter;

[0014] Step 4: The grid voltage e obtained in step 3 αR1 、e βR1 The grid voltage is filtered again through the quasi-resonant filter to obtain the grid voltage.

[0015] Step 5: The grid voltage obtained in step 3 is passed through the first quasi-resonant filter to obtain the grid voltage e αR1 、e βR1 The equation can be obtained as e αR1 、e βR1 The amplitude E R1 and phase

[0016] Step 6: The grid voltage obtained in step 4 is passed through the second quasi-resonant filter to obtain the grid voltage e αR2 、e βR2 The equation can be obtained as e αR2 、e βR2 The amplitude E R2 and phase θ R2 ,

[0017] Step 7: According to the e obtained in step 5 αR1 、e βR1 The amplitude E R1 and phase θR1 , and e obtained in step 6 αR2 、e βR2 The amplitude E R2 and phase θ R2 , we can get the amplitude attenuation ΔE produced by the quasi-resonant filter R and phase shift

[0018] Step 8: When the grid frequency has no error, ΔE R and Δθ R will be 0; when there is an error in the grid frequency, ΔE R and Δθ R will not be 0, at this time, according to the amplitude attenuation ΔE obtained in step 7 R and phase shift Δθ R Perform phase amplitude compensation to obtain the final grid voltage e after compensation. αR3 、

[0019] The present invention realizes DC bias suppression and adopts dual quasi-resonant filters in series instead of a single quasi-resonant filter, thereby realizing accurate observation of grid voltage without knowing the precise grid angular frequency and overcoming the influence of grid frequency offset, thereby enhancing the robustness of the grid voltage observation method to grid frequency offset.

[0020] The beneficial effects that the present invention can produce are as follows:

[0021] 1) Under an ideal power grid, the present invention can achieve accurate observation of the grid voltage;

[0022] 2) When a DC bias is present, the present invention eliminates the influence of the DC bias;

[0023] 3) When the grid frequency deviates, the present invention uses dual quasi-resonant filters in series, thereby overcoming the influence of the grid frequency deviation;

[0024] 4) In a harmonic power grid, the present invention uses a quasi-resonant filter to suppress grid voltage harmonics, which helps to reduce the impact of grid voltage harmonics on current control. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 It is an overall block diagram of the present invention.

[0027] Figure 2 The grid voltage observation result obtained by using the present invention under an ideal grid;

[0028] Figure 3 is the inverter voltage u α When a 1V DC bias is suddenly injected into the grid, the α-phase grid voltage and its error waveform observed by the present invention are obtained;

[0029] Figure 4 is the inverter voltage u α When a 1V DC bias is suddenly injected into the grid, the grid voltage of phase α and its error waveform are observed using the method proposed in the literature [Guo Leilei, et al. Model predictive control of grid-connected inverter without grid voltage sensor [J]. Transactions of China Electrotechnical Society, 2020, 35(12): 2612-2622.]

[0030] Figure 5 The grid voltage observed by the present invention when the grid frequency is 45 Hz;

[0031] Figure 6 The grid voltage waveform observed by the present invention is obtained when 10% of the 7th harmonic voltage is injected into the grid. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without creative work are within the scope of protection of the present invention.

[0033] like Figure 1 As shown, the present invention consists of a sliding mode observer, a series dual quasi-resonant filter, and an adaptive compensation link. The sliding mode observer is used to achieve preliminary observation of the grid voltage. The first quasi-resonant filter is used to suppress the grid voltage sliding mode noise and DC offset. The second quasi-resonant filter and the adaptive compensation link are used to achieve grid voltage observation that is resistant to grid frequency offset. After adaptive compensation, the observed grid voltage can be finally obtained. Specific embodiments are as follows:

[0034] A grid voltage observation method based on a dual quasi-resonant filter, the steps are as follows:

[0035] Step 1: The mathematical model of the grid-connected inverter in the two-phase stationary αβ coordinate system is:

[0036]

[0037] Where, eαg 、e βg are the two components of the three-phase grid voltage in the αβ coordinate system, u α 、u β and i α 、i β are the components of the inverter output voltage and grid current in the αβ coordinate system respectively;

[0038] According to formula (1), the following conventional sliding mode observer can be designed:

[0039]

[0040] Where, is the grid current observed by the sliding mode observer; K is the sliding mode gain; sgn() is the sign function; combining equations (1) and (2), the current observation error equation can be obtained:

[0041]

[0042] Where, is the component of the error between the current observation value and the actual value in the αβ coordinate system;

[0043] Step 2: After the sliding mode observer obtained in step 1 converges, After eliminating some of the current observation error equations obtained in step 1, the grid voltage can be obtained:

[0044]

[0045] The sign function will cause the observed grid voltage to contain a large amount of sliding mode noise. If the current or voltage signal in the sliding mode observer obtained in step 1 contains a DC bias, the grid voltage obtained in this step will contain a DC component.

[0046] Step 3: Use a quasi-resonant filter to suppress the high-frequency sliding mode noise and DC bias contained in the grid voltage observed in step 2, thereby obtaining the grid voltage e obtained by passing the grid voltage through the first quasi-resonant filter. αR1 、e βR1 :

[0047]

[0048] Where s is the Laplace operator, ω c1 is the resonant bandwidth of the quasi-resonant filter;

[0049] Step 4: The grid voltage e obtained in step 3 αR1 、e βR1After filtering again with the quasi-resonant filter, the grid voltage can be obtained. The grid voltage obtained by the second quasi-resonant filter is e αR2 、e βR2 :

[0050]

[0051] Step 5: The grid voltage obtained in step 3 is passed through the first quasi-resonant filter to obtain the grid voltage e αR1 、e βR1 The equation can be obtained as e αR1 、e βR1 The amplitude E R1 and phase θ R1 :

[0052]

[0053] Step 6: The grid voltage obtained in step 4 is passed through the second quasi-resonant filter to obtain the grid voltage e αR2 、e βR2 The equation can be obtained as e αR2 、e βR2 The amplitude E R2 and phase θ R2 :

[0054]

[0055] Step 7: According to the e obtained in step 5 αR1 、e βR1 The amplitude E R1 and phase θ R1 , and e obtained in step 6 αR2 、e βR2 The amplitude E R2 and phase θ R2 , we can get the amplitude attenuation ΔE produced by the quasi-resonant filter R and phase shift Δθ R :

[0056]

[0057] Step 8: When the grid frequency has no error, ΔE R and Δθ R Will be 0; when there is an error in the grid frequency, ΔE R and Δθ R will not be 0, at this time, according to the amplitude attenuation ΔE obtained in step 7 R and phase shift Δθ R Perform phase amplitude compensation to obtain the final grid voltage e after compensation. αR3 、e βR3 :

[0058]

[0059] Not only does it not require the precise grid angular frequency to be known, but it can also overcome the influence of grid frequency offset, thereby enhancing the robustness of the grid voltage observation method to grid frequency offset.

[0060] To verify the effectiveness of this invention, a platform experiment was conducted. The platform uses a DSP28335 digital signal processor as the main control chip, a bidirectional DC power supply to simulate the DC side, and an Ametek MX30 programmable AC power supply to simulate a three-phase grid, with a sampling frequency of 15kHz. The inverter and AC power supply are connected via a filter inductor. The experimental waveforms were recorded using a Myway PE-View9 host computer.

[0061] Detailed experimental parameters are shown in Table 1.

[0062] Table 1 System parameters

[0063]

[0064] Figure 2 The grid voltage observation result obtained by the method proposed in this invention under an ideal grid is shown in FIG.

[0065] Figure 3 is the inverter voltage u α When a 1V DC bias is suddenly injected into the grid, the α-phase grid voltage and its error waveform observed by the method proposed in the present invention are shown.

[0066] Figure 4 is the inverter voltage u α When a 1V DC bias is suddenly injected into the grid, the α-phase grid voltage and its error waveform are observed using the method proposed in the literature [Guo Leilei, et al. Model predictive control of grid-connected inverter without grid voltage sensor [J]. Transactions of the Chinese Society of Electrotechnical Engineering, 2020, 35(12): 2612-2622.

[0067] Figure 5 It is the grid voltage observed by the method proposed in the present invention when the grid frequency is 45 Hz.

[0068] Figure 6 The grid voltage waveform is observed using the method proposed in the present invention when 10% of the 7th harmonic voltage is injected into the grid.

[0069] The comparative experimental results show that:

[0070] 1) Under an ideal power grid, the method proposed in this invention can realize accurate observation of the power grid voltage.

[0071] 2) When DC bias is present, the method proposed in the present invention eliminates the influence of DC bias and solves the problem that the method proposed in the literature [Guo Leilei, et al. Model predictive control of grid-connected inverter without grid voltage sensor [J]. Transactions of the Chinese Society of Electrotechnical Engineering, 2020, 35(12): 2612-2622.] will amplify the influence of DC bias.

[0072] 3) When the grid frequency deviates, the method proposed in the present invention uses dual quasi-resonant filters in series to overcome the influence of the grid frequency deviation.

[0073] 4) Under harmonic power grid, the method proposed in the present invention can achieve grid voltage harmonic suppression due to the use of quasi-resonant filters, which helps to reduce the impact of grid voltage harmonics on current control.

Claims

1. A grid voltage observation method based on a dual quasi-resonant filter, characterized in that: Here are the steps: Step 1: The mathematical model of the grid-connected inverter in the two-phase stationary αβ coordinate system is: Where, e αg 、e βg are the two components of the three-phase grid voltage in the αβ coordinate system, u α 、u β and i α 、i β are the components of the inverter output voltage and grid current in the αβ coordinate system respectively; based on this, the following sliding mode observer is designed Where, is the grid current observed by the sliding mode observer; K is the sliding mode gain; sgn is the sign function; based on this, the current observation error equation is obtained Where, is the component of the error between the current observation value and the actual value in the αβ coordinate system; Step 2: After the sliding mode observer obtained in step 1 converges, The grid voltage can be obtained by eliminating some of the current observation error equations. The sign function causes the observed grid voltage to contain a large amount of sliding mode noise; if the current or voltage signal in the sliding mode observer obtained in step 1 contains a DC bias, the obtained grid voltage will contain a DC component; Step 3: Use a quasi-resonant filter to suppress the high-frequency sliding mode noise and DC bias contained in the grid voltage observed in step 2, thereby obtaining the grid voltage e obtained by passing the grid voltage through the first quasi-resonant filter. αR1 、e βR1 , Where s is the Laplace operator, ω c1 is the resonant bandwidth of the quasi-resonant filter; Step 4: The grid voltage e obtained in step 3 αR1 、e βR1 The grid voltage is filtered again by the quasi-resonant filter to obtain the grid voltage e after the second quasi-resonant filter. αR2 、eβR2, Step 5: The grid voltage obtained in step 3 is passed through the first quasi-resonant filter to obtain the grid voltage e αR1 、e βR1 The equation can be obtained as e αR1 、e βR1 The amplitude E R1 and phase θ R1 , Step 6: The grid voltage obtained in step 4 is passed through the second quasi-resonant filter to obtain the grid voltage e αR2 、e βR2 The equation can be obtained as e αR2 、e βR2 The amplitude E R2 and phase θ R2 , Step 7: According to the e obtained in step 5 αR1 、e βR1 The amplitude E R1 and phase θ R1 , and e obtained in step 6 αR2 、e βR2 The amplitude E R2 and phase θ R2 , we can get the amplitude attenuation ΔE produced by the quasi-resonant filter R and phase shift Δθ R , Step 8: When the grid frequency has no error, ΔE R and Δθ R will be 0; when there is an error in the grid frequency, ΔE R and Δθ R will not be 0, at this time, according to the amplitude attenuation ΔE obtained in step 7 R and phase shift Δθ R Perform phase amplitude compensation to obtain the final grid voltage e after compensation. αR3 、e βR3 ,

Images