Multi-inverter parallel state observer fractional order control method and device

Abstract

The invention provides a multi-inverter parallel state observer fractional order control method and a fractional order control device, and relates to the technical field of inverters. Constructing a fractional order current regulator of each inverter based on the mathematical model of the fractional order inverter; constructing a Romberg state observer to obtain predicted power at the next moment; current deviations of the d axis and the q axis of the output end of each inverter at the next moment are obtained through calculation; calculating to obtain respective self-adaptive virtual impedance at the next moment; performing fractional order adaptive virtual impedance d-axis and q-axis control by taking the current deviation of the d-axis and q-axis of the output end of each inverter at the next moment and the adaptive virtual impedance of each inverter at the next moment as the input of a fractional order current regulator; and generating an SPWM control signal at the next moment based on the d-axis component and the q-axis component of the actual output voltage of each inverter at the next moment. According to the invention, the dynamic and static performance of the multi-inverter parallel system is greatly improved, and the bus voltage is ensured to be basically constant.

Description

Technical Field

[0001] This application relates to the field of inverter technology, and in particular to a fractional-order control method and device for a state observer of multiple inverters connected in parallel. Background Technology

[0002] Parallel connection of multiple inverters is the main method and key equipment for solving the power generation output of different power sources in microgrids. The large-scale connection of unbalanced loads brings more severe challenges to the parallel connection of multiple inverters. In addition, the parallel connection of multiple inverters with unequal capacity requires stronger bus voltage and grid frequency balancing capabilities, better power distribution characteristics and certain anti-interference capabilities under unbalanced load conditions.

[0003] Existing traditional droop control and virtual synchronous generator control strategies are mostly applicable to special cases such as ideal loads with the same impedance of equal-capacity inverters. In reality, the inverter's own impedance and the line impedance are unlikely to be exactly the same, and single-phase and three-phase unbalanced loads are common. Traditional droop control algorithms have stringent conditions for precise power allocation and high requirements for the connected load. To improve the dynamic stability of the system, common control strategies include line impedance detection and calculation, introducing other compensation quantities, and high-frequency filtering. However, these methods are computationally complex, difficult to operate, and their effects are not ideal.

[0004] In practical applications of multi-inverter parallel operation, single-phase loads and three-phase unbalanced loads are more common, and the inverter capacities vary. However, microgrids based on droop control cannot fully meet the needs of this situation. Therefore, the commonly used control method is to add virtual impedance to change the characteristics of the line impedance, thereby enhancing the adaptability to droop control. However, the current method of adding virtual impedance to change the line impedance characteristics still cannot well meet the control requirements when multiple inverters with unequal capacities are connected in parallel to drive unbalanced loads. There are still static reactive power errors, line impedance differences between multiple inverters with unequal capacities, and large bus voltage fluctuations. Summary of the Invention

[0005] In view of the above problems, this application proposes a state observer fractional-order control method and a state observer fractional-order control device for multiple inverters in parallel, which overcomes the shortcomings of the prior art.

[0006] In a first aspect, embodiments of this application provide a state observer fractional-order control method for multiple inverters in parallel, comprising: constructing a fractional-order current regulator for each inverter based on a mathematical model of the fractional-order inverters; A Romberg state observer is constructed, and the grid-connected voltage and current values ​​at the current moment are used to predict the internal state variables of the multi-inverter grid-connected system in combination with the Romberg state observer to obtain the predicted power at the next moment. Based on the d-axis and q-axis components of the given current and grid-connected current on the load side of each inverter in the predicted power at the next moment, the current deviations of the d-axis and q-axis at the output of each inverter at the next moment are calculated. Based on the positive and negative sequence power values ​​of each inverter in the predicted power of the next moment, the adaptive virtual impedance of each inverter in the next moment is calculated. The current deviation of the d-axis and the current deviation of the q-axis at the output of each inverter at the next moment, and the adaptive virtual impedance of each inverter at the next moment are used as the input of the fractional-order current regulator to perform fractional-order adaptive virtual impedance d-axis and q-axis control, and output the d-axis component and q-axis component of the actual output voltage of each inverter at the next moment. Based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, an SPWM control signal is generated for the next moment and the operation of each inverter is controlled. At the next moment, the static error of reactive power when multiple inverters with unequal capacity are connected in parallel is eliminated, so that the difference in line impedance between multiple inverters with unequal capacity gradually decreases until it tends to be consistent.

[0007] Optionally, a fractional-order current regulator for each inverter is constructed based on the mathematical model of the fractional-order inverter, including: A circuit for constructing a three-phase fractional-order LCL inverter is introduced by using fractional-order inductors and fractional-order capacitors. A mathematical model of the abc coordinate coefficients of the three-phase fractional-order LCL inverter is established using the switching function description method. Based on the abc coordinate coefficient mathematical model, the DQ coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is established, and then the dq coordinate coefficient mathematical model is established. Based on the dq coordinate coefficient mathematical model, a fractional integral order is introduced to construct a fractional current regulator for each inverter. This fractional current regulator includes a d-axis fractional current regulator and a q-axis fractional current regulator.

[0008] Optionally, the abc coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0009] In the above formula, L1 represents the fractional-order inductor on the inverter side, and its order is... L2 represents the fractional-order inductance on the grid side, with its order being... C represents the fractional capacitance on the AC side, and its order is 1. order , , The range of values ​​for i is (0, 2). 1_abc Indicates the inverter output current, i 2_abcThis represents the grid-side current, i.e., the grid-connected current, where t represents the time constant, and u... abc Indicates the AC voltage measurement of the inverter, U C_abc U represents the fractional-order filter capacitor voltage. g_abc Indicates the three-phase grid voltage at the grid connection point, u oo’_abc Ps represents the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. abc u dc Indicates the transformation coefficients of the 32 coordinate system; The DQ coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0010] After simplifying the above equation, the corresponding simplified model expression is as follows:

[0011] In the above formula, C 2s / 3s The coefficients of the bi- and tri-transformations, i 1_DQ The D and Q components of the inverter input current, i 2_DQ The D and Q components of the inverter output current, u i_DQ The D and Q components of the inverter's AC side voltage, u c_DQ The D and Q components of the fractional capacitor voltage, s DQ Represents the D and Q components of the switching function, u dc U represents the DC side voltage. g_DQ The D and Q components of the three-phase grid voltage of the inverter, u oo’_DQ The D and Q components represent the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. The dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0012] In the above formula, Let d and q represent the input current of the inverter, respectively. These represent the d and q components of the inverter output current, respectively. These represent the d and q components of the three-phase grid voltage of the inverter, respectively. These represent the d and q components of the fractional capacitor voltage, respectively. These represent the d and q components of the AC side voltage of the inverter, respectively. Let d and q be the components of the switching function, respectively; C 3s / 2s C represents the coefficients of the 3x2 transformation. 2r / 2s Represents the transformation coefficients from stationary coordinates to rotating coordinates; Based on the dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, the fractional-order integral order λ is introduced to construct the fractional-order current regulator.

[0013] Optionally, based on the d-axis and q-axis components of the given current and grid-connected current on the load side of each inverter in the predicted power at the next moment, the current deviations of the d-axis and q-axis at the output of each inverter at the next moment are calculated, including: The magnitude I of the inverter load-side given current in the predicted power at the next moment. * ,and , After multiplication, the corresponding given values ​​of the d-axis current and q-axis current are obtained respectively. , ,in The load-side voltage phase is measured after passing through the phase-locked loop; For grid-connected current After sampling coefficients and coordinate transformation, the d-axis and q-axis components are obtained respectively. and ; The calculation yields the current deviation along the d-axis at the output of each inverter at the next moment. Current deviation along the q-axis at the next moment .

[0014] Optionally, based on the positive and negative sequence power values ​​of each inverter in the predicted power at the next moment, the adaptive virtual impedance of each inverter is calculated, including: The reactive power in the positive and negative sequence power values ​​of each inverter in the predicted power at the next moment is subtracted to obtain the power difference value ΔQ; The power difference ΔQ is combined with the integral term K v The product operation is performed on / S to obtain the adaptive virtual impedance Z at the next time step. v =K v / S*ΔQ, where K v This represents the adaptive adjustment coefficient, and S represents the calculation transformation method used to convert the power value into a resistance value.

[0015] Optionally, the current deviation of the d-axis and the current deviation of the q-axis at the output of each inverter at the next moment, and the adaptive virtual impedance of each inverter at the next moment, are used as inputs to the fractional-order current regulator to perform fractional-order adaptive virtual impedance d-axis and q-axis control, outputting the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, including: The difference between the d-axis reference voltage after introducing the adaptive virtual impedance at the next moment and the voltage value corresponding to the adaptive virtual impedance at the next moment is used to obtain the reference voltage of the inverter's output d-axis at the next moment. The input voltage outer loop closed-loop transfer function is obtained by subtracting the reference voltage of the output terminal d-axis at the next moment from the d-axis component of the actual output voltage of the inverter at the next moment. The first current value is obtained by multiplying the reference voltage of the output terminal d-axis at the next moment with the first transformation parameter; The sum of the first current value and the second current value is then subtracted from the third current value to obtain the current deviation of the inverter output terminal d-axis at the next moment. The second current value is the output value of the voltage outer loop closed-loop transfer function. The current deviation of the inverter output terminal d-axis at the next moment is used as the input of the fractional current regulator. The output value of the fractional current regulator is multiplied by the inverter fundamental pulse width modulation ratio coefficient to obtain the d-axis component of the inverter reference output voltage at the next moment. The difference between the d-axis component of the inverter reference output voltage at the next moment and the d-axis component of the inverter actual output voltage at the next moment is calculated, and then multiplied with the second transformation parameter. The resulting current value is then compared with the d-axis component of the unbalanced load current to obtain the third current value. The third current value and the third transformation parameter are multiplied to obtain the d-axis component of the actual output voltage of the inverter at the next moment. The reference voltage of the q-axis at the inverter output terminal at the next moment is obtained by subtracting the voltage value corresponding to the adaptive virtual impedance at the next moment from the adaptive virtual impedance at the next moment. The difference between the reference voltage of the output terminal at the next q-axis moment and the q-axis component of the actual output voltage of the inverter at the next q-axis moment is then input into the voltage outer loop closed-loop transfer function. The fourth current value is obtained by multiplying the reference voltage of the output terminal q-axis at the next moment with the first transformation parameter; The sum of the fourth and fifth current values ​​is then subtracted from the sixth current value to obtain the current deviation of the inverter output terminal q axis at the next moment. The fifth current value is the output value of the voltage outer loop closed-loop transfer function. The current deviation of the inverter output terminal q-axis at the next moment is used as the input of the fractional current regulator. The output value of the fractional current regulator is multiplied by the inverter fundamental pulse width modulation ratio coefficient to obtain the q-axis component of the inverter reference output voltage at the next moment. The difference between the q-axis component of the inverter reference output voltage at the next moment and the q-axis component of the inverter actual output voltage at the next moment is calculated, and then multiplied with the second transformation parameter. The resulting current value is then compared with the q-axis component of the unbalanced load current to obtain the sixth current value. The sixth current value and the third transformation parameter are multiplied to obtain the q-axis component of the actual output voltage of the inverter at the next moment.

[0016] Optionally, based on the dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, the fractional-order integral order λ is introduced to construct the fractional-order current regulator, including: Based on the abc coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, and combined with the PID control principle, the closed-loop transfer function of the voltage outer loop is obtained. G v ( s ); The voltage outer loop closed-loop transfer function G v ( s Based on this, the fractional integral order λ is introduced to construct the fractional current regulator. Wherein, the voltage outer loop closed-loop transfer function G v ( s The expression for ) is as follows:

[0017] In the above formula, K vp , K vi Both represent the scaling factor of the virtual impedance; The fractional-order current regulator The expression is as follows:

[0018] In the above formula, K fpi This represents the proportionality coefficient of the current loop. K fii This represents the integral coefficient of the current loop.

[0019] Optionally, based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, an SPWM control signal for the next moment is generated, including: The SPWM modulation signal of the inverter's d-axis at the next moment is obtained by subtracting the d-axis component of the actual output voltage of the inverter from the d-axis component of the fractional capacitor voltage. The SPWM modulation signal of the inverter's q-axis at the next moment is obtained by subtracting the q-axis component of the actual output voltage of the inverter from the q-axis component of the fractional capacitor voltage. The SPWM modulation signal of the inverter's d-axis at the next moment and the SPWM modulation signal of the inverter's q-axis at the next moment together constitute the SPWM control signal at the next moment.

[0020] Optionally, a Romberg state observer is constructed, including: Based on LCL type inverter The mathematical model in the coordinate system is used to calculate the discrete prediction model of the inverter-side current of each inverter. Among them, the LCL type inverter is in The mathematical model in the coordinate system is expressed as follows:

[0021] Discretizing the above, we obtain the following equation:

[0022] And thus obtain The expression for the output current prediction model in the coordinate system is as follows:

[0023] In the above formula, for The predicted value of the grid-side current at time k+1; for The current on the inverter side of the shaft is k The predicted value at time +1; for The grid voltage of the axis is k The sampled value at time; for The voltage of the shaft filter capacitor is at k The sampled value at time; for The current on the grid side of the axis is k The sampled value at time; for The current on the inverter side of the shaft is k The sampled value at time; The system sampling period; When multiple inverters are running stably, and the voltage of the filter capacitor is stable, the following expression holds:

[0024] The discrete prediction model for the inverter-side current can be obtained by calculation using the above formula, and its expression is as follows:

[0025] Based on the inverter voltage and current equations, the state-space model of the LCL inverter circuit is constructed as follows:

[0026] Selecting the state vector Take the input vector ,in: , , ,

[0027] Then we have:

[0028] Selecting state vectors from the state-space model The resulting exact discrete-time model changes as follows:

[0029] in: , ,

[0030] ,

[0031] In the above formula, u g Indicates the grid connection voltage. u 0 indicates the AC side voltage of the inverter. i 1k This indicates the inverter output current. u Ck Indicates the three-phase grid voltage at the grid connection point. i 2k This represents the measured current, where T represents the time constant. Indicates the grid-connected voltage frequency. Indicates the resonant frequency; Substituting the coefficients, we get:

[0032]

[0033]

[0034] in, This is the resonant frequency of the filter; The feedback value in the precise discrete-time model x d ( k+1) The expression is as follows:

[0035] Among them, let , , , Then there is ; Based on the above parameter values, the values ​​of each matrix in the state-space model expression are calculated. Ignoring grid voltage disturbances, the equations for the Romberg state observer are constructed as follows:

[0036] The above transformation is as follows:

[0037] in, Let be the feedback gain matrix of the Romberg state observer. Then the above equation becomes the following equation:

[0038] To obtain the error equation of the Romberg state observer, the following equation is given:

[0039] Define error vector Then the error equation becomes the following:

[0040] The dynamic performance of the error vector depends on the matrix. eigenvalues.

[0041] Secondly, embodiments of this application provide a state observer fractional-order control device for multiple inverters connected in parallel, comprising: Construct a regulator module to build a fractional current regulator for each inverter based on the mathematical model of the fractional inverter; The prediction module is used to predict the internal state variables of the multi-inverter grid-connected system using the current grid-connected voltage value and the Romberg state observer, so as to obtain the predicted power at the next moment. The current deviation module is used to calculate the current deviation of each inverter output terminal at the next moment based on the d-axis and q-axis components of the given current and grid-connected current on the load side of each inverter in the predicted power at the next moment. The adaptive virtual impedance module is used to calculate the adaptive virtual impedance of each inverter at the next moment based on the positive and negative sequence power values ​​of each inverter in the predicted power at the next moment. The virtual impedance control module is used to take the d-axis current deviation and q-axis current deviation of each inverter output terminal at the next moment, and the adaptive virtual impedance of each inverter at the next moment as inputs to the fractional-order current regulator, to perform fractional-order adaptive virtual impedance d-axis and q-axis control, and output the d-axis component and q-axis component of the actual output voltage of each inverter at the next moment. The control signal module is used to generate the SPWM control signal for the next moment based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, and control the operation of each inverter. In the next moment, it eliminates the static error of reactive power when multiple inverters with unequal capacity are connected in parallel, so that the difference in line impedance between multiple inverters with unequal capacity gradually decreases until it tends to be consistent.

[0042] This application proposes a state observer fractional-order control method for multiple inverters in parallel, and puts forward a PI control method based on variable virtual impedance. λ The D-adaptive control method mainly uses the integral element to eliminate the static error of reactive power, while adding an adjustable parameter λ to improve the shortcomings of traditional control strategies, such as being too simple and having poor steady-state performance.

[0043] Design a fractional-order current regulator A Romberg state observer was constructed, and a novel approach was proposed to use the obtained accurate next-moment predicted power for subsequent calculations. The calculated current deviation was then fed into a fractional-order current regulator to form a feedback loop that modulates the SPWM pulse signal for the next moment.

[0044] Simultaneously, a multi-loop control strategy is designed, creatively proposing to use the obtained accurate next-time predicted power for the calculation of adaptive virtual impedance. An adaptive virtual impedance outer-loop control strategy is proposed, which calculates the difference in reactive power after positive and negative sequence separation for each inverter at the next time step, and then adds an integral term to the difference to obtain the virtual impedance as a fractional-order current regulator. The input is used to perform fractional-order adaptive virtual impedance d-axis and q-axis control, outputting the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, generating the SPWM control signal for the next moment. This achieves zero steady-state error tracking of reactive power, gradually reducing the line impedance difference between multiple inverters with unequal capacity until they tend to be consistent, greatly improving the dynamic and static performance of the multi-inverter parallel system and ensuring that the bus voltage remains basically constant. The fractional-order control method proposed in this application has broad application prospects and high practicality. Attached Figure Description

[0045] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart of the state observer fractional-order control method for multiple inverters in parallel proposed in this application embodiment; Figure 2 This is a schematic diagram of the circuit structure of a three-phase fractional-order LCL inverter exemplified in the embodiments of this application; Figure 3 This is a block diagram illustrating the fractional-order adaptive virtual impedance d-axis control in the embodiments of this application; Figure 4 This is a schematic diagram illustrating the control principle when two inverters are connected in parallel to different load conditions, as exemplified in the embodiments of this application. Figure 5 This is a block diagram of a fractional-order control device for parallel connection of multiple inverters with unequal capacity, as described in the embodiments of this application. Detailed Implementation

[0046] The embodiments of this application will now be described in detail. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0047] This application proposes a fractional-order control method for a state observer in parallel multi-inverter configuration, referencing... Figure 1 The flowchart shown includes the following steps: Step 101: Construct the fractional current regulator for each inverter based on the mathematical model of the fractional inverter.

[0048] The fractional-order control method proposed in this application firstly requires constructing a fractional-order current regulator for each inverter based on the mathematical model of the fractional-order inverter.

[0049] In one embodiment of this application, the mathematical model of the fractional-order inverter preferably involves introducing fractional-order inductors and capacitors to construct the circuit of a three-phase fractional-order LCL inverter, and then using the switching function description method to establish the abc coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter. Based on this, a DQ coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is then established, followed by a dq coordinate coefficient mathematical model. Finally, based on the dq coordinate coefficient mathematical model, a fractional-order integral order is introduced to construct the fractional-order current regulator for each inverter. This fractional-order current regulator includes a d-axis fractional-order current regulator and a q-axis fractional-order current regulator.

[0050] Reference Figure 2 The circuit diagram shown is of a three-phase fractional-order LCL inverter. L1 represents the fractional-order inductor on the inverter side, and its order is... L2 represents the fractional-order inductance on the grid side, with its order being... C represents the fractional capacitance on the AC side, and its order is 1. order , , The range of values ​​for is (0, 2).

[0051] Based on the above mathematical model analysis, the mathematical model of the three-phase fractional-order LCL inverter in the stationary coordinate system, that is, the mathematical model of the abc coordinate coefficients of the three-phase fractional-order LCL inverter, can be expressed as follows:

[0052] In the above formula, i 1_abc Indicates the inverter output current, i 2_abc The grid-side current, i.e., the grid-connected current, is represented by t, which represents the practical constant, and u. abc Indicates the AC voltage measurement of the inverter, U C_abc U represents the fractional-order filter capacitor voltage. g_abc Indicates the three-phase grid voltage at the grid connection point, u oo’_abc Ps represents the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. abc u dc This represents the transformation coefficients of the 32 coordinate system.

[0053] Converting the above equation into a DQ coordinate coefficient mathematical model expression, we obtain the following expression for the DQ coordinate coefficient mathematical model of a three-phase fractional-order LCL inverter:

[0054] After simplifying the above equation, the corresponding simplified model expression is as follows:

[0055] In the above formula, C2s / 3s represents the coefficients of the two-three transformation, and i 1_DQ The D and Q components of the inverter input current, i 2_DQ The D and Q components of the inverter output current, u i_DQ The D and Q components of the inverter's AC side voltage, u c_DQ The D and Q components of the fractional capacitor voltage, s DQ Represents the D and Q components of the switching function, u dc U represents the DC side voltage. g_DQ The D and Q components of the three-phase grid voltage of the inverter, u oo’_DQ The D and Q components represent the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor; Then, converting it into an expression in the form of a dq coordinate coefficient mathematical model, we obtain the following expression for the dq coordinate coefficient mathematical model of a three-phase fractional-order LCL inverter:

[0056] In the above formula, Let d and q represent the input current of the inverter, respectively. These represent the d and q components of the inverter output current, respectively. These represent the d and q components of the three-phase grid voltage of the inverter, respectively. These represent the d and q components of the fractional capacitor voltage, respectively. These represent the d and q components of the AC side voltage of the inverter, respectively. Let d and q be the components of the switching function, respectively; C 3s / 2s C represents the coefficients of the 3x2 transformation. 2r / 2s This represents the transformation coefficient from stationary coordinates to rotating coordinates.

[0057] Based on the dq coordinate coefficient mathematical model of a three-phase fractional-order LCL inverter, a fractional-order integral order λ is introduced to construct a fractional-order current regulator.

[0058] Step 102: Construct a Romberg state observer and use the current grid-connected voltage and current values ​​to predict the internal state variables of the multi-inverter grid-connected system, thereby obtaining the predicted power for the next moment.

[0059] Constructing a Romberg state observer includes: Based on LCL type inverter The mathematical model in the coordinate system is used to calculate the discrete prediction model of the inverter-side current of each inverter. Among them, LCL type inverters are The mathematical model in the coordinate system is expressed as follows:

[0060] Discretizing the above, we obtain the following equation:

[0061] And thus obtain The expression for the output current prediction model in the coordinate system is as follows:

[0062] In the above formula, for The predicted value of the grid-side current at time k+1; for The current on the inverter side of the shaft is kThe predicted value at time +1; for The grid voltage of the axis is k The sampled value at time; for The voltage of the shaft filter capacitor is at k The sampled value at time; for The current on the grid side of the axis is k The sampled value at time; for The current on the inverter side of the shaft is k The sampled value at time; The system sampling period; When multiple inverters are running stably, and the voltage of the filter capacitor is stable, the following expression holds:

[0063] The discrete prediction model for the inverter-side current can be obtained by calculation from the above formula, and its expression is as follows:

[0064] Based on the inverter voltage and current equations, the state-space model of the LCL inverter circuit is constructed as follows:

[0065] Selecting the state vector Take the input vector ,in: , , ,

[0066] Then we have:

[0067] Select state vectors in the state-space model The resulting exact discrete-time model changes as follows:

[0068] in: , ,

[0069] ,

[0070] In the above formula, u g Indicates the grid connection voltage. u 0 indicates the AC side voltage of the inverter.i 1k This indicates the inverter output current. u Ck Indicates the three-phase grid voltage at the grid connection point. i 2k This represents the measured current, where T represents the time constant. Indicates the grid-connected voltage frequency. Indicates the resonant frequency; Substituting the coefficients, we get:

[0071]

[0072]

[0073] in, This is the resonant frequency of the filter; Feedback value in an exact discrete-time model x d ( k +1) The expression is as follows:

[0074] Among them, let , , , Then there is ; Based on the above parameter values, the values ​​of each matrix in the state-space model expression are calculated. Ignoring grid voltage disturbances, the equations for the constructed Romberg state observer are as follows:

[0075] The above transformation is as follows:

[0076] in, Let be the feedback gain matrix of the Romberg state observer. Then the above equation becomes the following equation:

[0077] To obtain the error equation for the Romberg state observer, the following equation is given:

[0078] Define error vector Then the error equation becomes the following:

[0079] The dynamic performance of the error vector depends on the matrix. eigenvalues.

[0080] The Romberg state observer described above can accurately obtain the predicted power at the next moment. It is understood that the predicted power naturally includes voltage, current and positive and negative sequence power values.

[0081] In traditional parallel control of multiple inverters, the single-cycle delay in the digital control system (i.e., a one-beat delay between the calculated control input and the actual output) leads to current tracking lag, causing current ripple and overshoot. Furthermore, traditional control systems are highly sensitive to the inverter's LCL filter parameters; parameter mismatch can cause steady-state error in the dq-axis current, affecting control accuracy.

[0082] The method of adding virtual impedance to change the line impedance characteristics requires the use of power. If the current is not accurate enough, it will affect the accuracy of the power, which will cause the method to fail to meet the control requirements when multiple inverters with unequal capacity are connected in parallel to drive unbalanced loads, and there will still be static reactive power error.

[0083] This application creatively proposes the construction of a Romberg state observer, which predicts the current state in real time at the next moment by constructing a mathematical model of multiple inverters in parallel, and uses the predicted value as a feedback signal to compensate for control delay. The Romberg state observer, through a disturbance compensation stage, treats parameter errors as system disturbances and estimates and compensates for them in real time. Through the observer's synchronous estimation of voltage and current, high-precision instantaneous power values ​​can be indirectly obtained, achieving zero steady-state error tracking of reactive power. This gradually reduces the line impedance differences between multiple inverters with unequal capacities until they become consistent, significantly improving the dynamic and static performance of the multi-inverter parallel system and ensuring that the bus voltage remains essentially constant. Furthermore, compared to traditional methods that require the addition of power sensors for direct measurement, this application only requires basic current / voltage sensors, eliminating the need for additional measurement tools and indirectly reducing costs.

[0084] Step 103: Based on the d-axis and q-axis components of the given current and grid-connected current on the load side of each inverter in the predicted power at the next moment, calculate the current deviation of the d-axis and q-axis at the output of each inverter at the next moment.

[0085] After the fractional-order current regulator of the inverter is constructed, the d-axis and q-axis current deviations at the output of each inverter can be calculated based on the d-axis and q-axis components of the given current on the load side of each inverter and the grid-connected current in the predicted power at the next moment. Specifically, this includes: First, the amplitude I of the inverter load-side current given in the predicted power for the next moment. * ,and , After multiplication, the corresponding given values ​​of the d-axis current and q-axis current are obtained respectively. , ,in The phase of the load-side voltage is measured after passing through the phase-locked loop; this calculation yields the d-axis component of the given current on the inverter's load side. and q-axis components .

[0086] Simultaneously, the grid-connected current After sampling coefficients and coordinate transformation, the d-axis components and q-axis components are obtained respectively. and This calculation yields the d-axis component of the grid-connected current. and q-axis components .

[0087] After obtaining the above quantities, the current deviation of the d-axis at the output of each inverter at the next moment can be calculated. Current deviation along the q-axis at the next moment .

[0088] Step 104: Calculate the adaptive virtual impedance of each inverter in the next moment based on the positive and negative sequence power values ​​of each inverter in the predicted power of the next moment.

[0089] The calculation of adaptive virtual impedance requires the positive and negative sequence power values ​​of each inverter in the predicted power at the next time step. After obtaining the positive and negative sequence power values, the reactive power in the positive and negative sequence power values ​​of the inverter at the next time step is subtracted to obtain the power difference ΔQ at the next time step; then, the power difference ΔQ at the next time step is compared with the integral term K. v The product operation is performed on / S to obtain the adaptive virtual impedance Z at the next time step. v =K v / S*ΔQ, where K v This represents the adaptive adjustment coefficient, and S represents the calculation transformation method used to convert the power value into a resistance value. In this way, the adaptive required impedance for the next moment can be obtained.

[0090] Step 105: Use the current deviation of the d-axis and the current deviation of the q-axis at the output of each inverter at the next moment, and the adaptive virtual impedance of each inverter at the next moment as inputs to the fractional-order current regulator, perform fractional-order adaptive virtual impedance d-axis and q-axis control, and output the d-axis component and q-axis component of the actual output voltage of each inverter at the next moment.

[0091] The aforementioned steps yield the current deviation on the d-axis and the current deviation on the q-axis at the output of each inverter at the next moment, as well as the adaptive virtual impedance of each inverter at the next moment. These are used as inputs to a fractional-order current regulator to perform fractional-order adaptive virtual impedance d-axis and q-axis control, and output the d-axis and q-axis components of the actual output voltage of each inverter at the next moment.

[0092] In one embodiment of this application, the method for fractional-order adaptive virtual impedance d-axis control is the same as the method for q-axis control. Taking the d-axis control method as an example, combined with... Figure 3 The block diagram of the fractional-order adaptive virtual impedance d-axis control shown includes the following specific methods: The d-axis reference voltage after introducing the adaptive virtual impedance at the next time step v * dref The voltage value corresponding to the adaptive virtual impedance at the next moment. v load By performing the difference, the reference voltage of the inverter's output d-axis at the next moment can be obtained. v dref Then, the reference voltage of the output terminal d-axis at the next moment. v dref The d-axis component of the inverter's actual output voltage at the next moment. v d After subtraction, the input voltage outer loop closed-loop transfer function G v (s).

[0093] Next, the reference voltage of the output d-axis at the next moment. v dref , and the first transformation parameter ( Figure 3 sC f Identifier, where C f The first current value is obtained by multiplying the first and second current values ​​(representing the filter inductance coefficient of the LCL inverter); the sum of the first and second current values ​​is then combined with the third current value. i cd The difference is calculated, and the resulting current value is the current deviation of the inverter output d-axis at the next moment. The second current value is the voltage outer loop closed-loop transfer function G. v The output value of (s).

[0094] The current deviation at the d-axis of the inverter output at the next moment. As a fractional-order current regulator Input, fractional-order current regulator The output value and the inverter fundamental pulse width modulation scaling factor K pwm After performing the product operation, the d-axis component of the inverter reference output voltage at the next moment is obtained.v do .

[0095] The d-axis component of the inverter reference output voltage at the next moment v do The d-axis component of the inverter's actual output voltage at the next moment v d After subtracting, then compare with the second transformation parameter ( Figure 3 Chinese The current value is obtained by multiplying the values ​​of the variable (where r represents the adjustable coefficient). i d d-axis component of unbalanced load current i od Perform the difference operation to obtain the third current value. i cd ( i cd That is, the d-axis component of the fractional capacitance. Finally, regarding the third current value... i cd With the third transformation parameter ( Figure 3 Chinese The product operation (using the identifier) ​​is performed to obtain the d-axis component of the actual output voltage of the inverter at the next moment. v d .

[0096] The q-axis control method in this application is the same as the d-axis control method, and will not be described in detail here. It includes: subtracting the voltage value corresponding to the adaptive virtual impedance at the next moment from the q-axis reference voltage after introducing the adaptive virtual impedance at the next moment, to obtain the q-axis reference voltage at the inverter output at the next moment.

[0097] The input voltage outer loop closed-loop transfer function is obtained by subtracting the reference voltage of the output terminal q-axis at the next moment from the q-axis component of the actual output voltage of the inverter at the next moment; the fourth current value is obtained by multiplying the reference voltage of the output terminal q-axis at the next moment with the first transformation parameter.

[0098] The sum of the fourth and fifth current values ​​is then subtracted from the sixth current value to obtain the current deviation of the inverter output terminal q axis at the next moment. The fifth current value is the output value of the voltage outer loop closed-loop transfer function.

[0099] The current deviation of the inverter output q-axis at the next moment is used as the input of the fractional-order current regulator. The output value of the fractional-order current regulator is multiplied by the inverter's fundamental pulse width modulation ratio coefficient to obtain the q-axis component of the inverter's reference output voltage at the next moment.

[0100] The difference between the q-axis component of the inverter reference output voltage at the next moment and the q-axis component of the inverter actual output voltage at the next moment is calculated, and then multiplied with the second transformation parameter. The resulting current value is then compared with the q-axis component of the unbalanced load current to obtain the sixth current value.

[0101] By multiplying the sixth current value with the third transformation parameter, the q-axis component of the inverter's actual output voltage at the next moment is obtained. v d .

[0102] In the entire control process described above, based on the dq coordinate coefficient mathematical model of a three-phase fractional-order LCL inverter, the fractional-order integral order λ is introduced, and the specific method for constructing a fractional-order current regulator includes: First, based on the abc coordinate coefficient mathematical model of a three-phase fractional-order LCL inverter, and combined with the PID control principle, the voltage outer loop closed-loop transfer function is obtained. G v ( s Then, in the voltage outer loop closed-loop transfer function G v ( s Based on this, a fractional integral order λ is introduced to construct a fractional current regulator.

[0103] Among them, the voltage outer loop closed-loop transfer function G v ( s The expression for ) is as follows:

[0104] In the above formula, K vp , K vi Both represent the scaling factor of the virtual impedance; Fractional current regulator The expression is as follows:

[0105] In the above formula, K fpi This represents the proportionality coefficient of the current loop. K fii This represents the integral coefficient of the current loop.

[0106] Step 106: Based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, generate the SPWM control signal for the next moment and control the operation of each inverter. At the next moment, eliminate the static error of reactive power when multiple inverters with unequal capacity are connected in parallel, so that the line impedance difference between multiple inverters with unequal capacity gradually decreases until it tends to be consistent.

[0107] After obtaining the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, the SPWM control signal for the next moment is generated based on these two quantities. The SPWM control signal controls the operation of the corresponding inverter, eliminates the static error of reactive power when multiple inverters with unequal capacity are connected in parallel, and gradually reduces the difference in line impedance between multiple inverters with unequal capacity until they become consistent.

[0108] In one embodiment of this application, a preferred method for generating the SPWM control signal for the next time step includes: The difference between the d-axis component of the actual output voltage of the inverter at the next moment and the d-component of the fractional capacitor voltage is obtained to get the SPWM modulation signal of the inverter's d-axis at the next moment; the difference between the q-axis component of the actual output voltage of the inverter at the next moment and the q-component of the fractional capacitor voltage is obtained to get the SPWM modulation signal of the inverter's q-axis at the next moment; the SPWM modulation signals of the inverter's d-axis and q-axis at the next moment together constitute the SPWM control signal for the next moment.

[0109] The above process refers to Figure 4 The schematic diagram of the control principle when two inverters are connected in parallel to different load conditions can be understood more intuitively. DG1 and DG2 represent two inverters respectively. The voltage and current dual closed-loop process includes adaptive virtual impedance ( Figure 3 For intuitive understanding and diagrammatic simplicity, the adaptive virtual impedance is highlighted with a solid line box (which can be combined with...). Figure 3 To understand this, the voltage and current dual closed-loop processes of both inverters are identical. Figure 4 The dashed boxes indicate the specific steps involved in the process. u gabc This represents the three-phase grid voltage of the inverter in the abc coordinate system. u cabc This represents the voltage of a fractional capacitor in the abc coordinate system.

[0110] Fractional current regulator Of the two output quantities u rD This refers to the d-axis component of the actual output voltage of the aforementioned inverter at the next moment. v d Naturally, this is understandable. u rQ This refers to the q-axis component of the inverter's actual output voltage at the next instant.

[0111] The d-axis component of the actual output voltage of the inverter at the next moment v d With the d-component of the fractional capacitor voltage ucD By performing the subtraction, the SPWM modulation signal of the inverter's d-axis at the next moment is obtained. u MD The q-axis component of the actual output voltage of the inverter at the next moment and the q-component of the fractional capacitor voltage. u cQ By performing the subtraction, the SPWM modulation signal of the inverter's q-axis at the next moment is obtained. u MQ Among them, the SPWM modulation signal of the inverter's d-axis at the next moment. u MD The SPWM modulation signal of the inverter's q-axis at the next moment u MQ Together they form the SPWM control signal for the next moment.

[0112] Based on the aforementioned state observer fractional-order control method for multiple inverters in parallel, this application also proposes a fractional-order control device for multiple inverters in parallel with unequal capacities, referring to... Figure 5 The block diagram shown includes: Construct regulator module 510, which is used to construct a fractional current regulator for each inverter based on the mathematical model of the fractional current inverter; The prediction module 520 is used to predict the internal state variables of the multi-inverter grid-connected system by using the current grid-connected voltage value and combining it with the Romberg state observer, so as to obtain the predicted power at the next moment. The current deviation module 530 is used to calculate the current deviation of each inverter output terminal at the next moment based on the d-axis component and q-axis component of the given current and grid-connected current on the load side of each inverter in the predicted power at the next moment. The adaptive virtual impedance module 540 is used to calculate the adaptive virtual impedance of each inverter at the next moment based on the positive and negative sequence power values ​​of each inverter in the predicted power at the next moment. The virtual impedance control module 550 is used to take the d-axis current deviation and q-axis current deviation of each inverter output terminal at the next moment, and the adaptive virtual impedance of each inverter at the next moment as the input of the fractional-order current regulator, to perform fractional-order adaptive virtual impedance d-axis and q-axis control, and output the d-axis component and q-axis component of the actual output voltage of each inverter at the next moment. The control signal module 560 is used to generate the SPWM control signal for the next moment based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, and control the operation of each inverter. In the next moment, it eliminates the static error of reactive power when multiple inverters with unequal capacity are connected in parallel, so that the line impedance difference between multiple inverters with unequal capacity gradually decreases until it tends to be consistent.

[0113] Optionally, the regulator module 510 is specifically used for: A circuit for constructing a three-phase fractional-order LCL inverter is introduced by using fractional-order inductors and fractional-order capacitors. A mathematical model of the abc coordinate coefficients of the three-phase fractional-order LCL inverter is established using the switching function description method. Based on the abc coordinate coefficient mathematical model, the DQ coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is established, and then the dq coordinate coefficient mathematical model is established. Based on the aforementioned dq coordinate coefficient mathematical model, a fractional integral order is introduced to construct a fractional current regulator for each inverter. This fractional current regulator includes a d-axis fractional current regulator and a q-axis fractional current regulator. The abc coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0114] In the above formula, L1 represents the fractional-order inductor on the inverter side, and its order is... L2 represents the fractional-order inductance on the grid side, with its order being... C represents the fractional capacitance on the AC side, and its order is 1. order , , The range of values ​​for i is (0, 2). 1_abc Indicates the inverter output current, i 2_abc The grid-side current, i.e., the grid-connected current, is represented by t, which represents the practical constant, and u. abc Indicates the AC voltage measurement of the inverter, U C_abc U represents the fractional-order filter capacitor voltage. g_abc Indicates the three-phase grid voltage at the grid connection point, u oo’_abc Ps represents the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. abc u dc Indicates the transformation coefficients of the 32 coordinate system; The DQ coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0115] After simplifying the above equation, the corresponding simplified model expression is as follows:

[0116] In the above formula, C 2s / 3s The coefficients of the bi- and tri-transformations, i 1_DQ The D and Q components of the inverter input current, i 2_DQ The D and Q components of the inverter output current, u i_DQThe D and Q components of the inverter's AC side voltage, u c_DQ The D and Q components of the fractional capacitor voltage, s DQ Represents the D and Q components of the switching function, u dc U represents the DC side voltage. g_DQ The D and Q components of the three-phase grid voltage of the inverter, u oo’_DQ The D and Q components represent the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. The dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter is expressed as follows:

[0117] In the above formula, Let d and q represent the input current of the inverter, respectively. These represent the d and q components of the inverter output current, respectively. These represent the d and q components of the three-phase grid voltage of the inverter, respectively. These represent the d and q components of the fractional capacitor voltage, respectively. These represent the d and q components of the AC side voltage of the inverter, respectively. Let d and q be the components of the switching function, respectively; C 3s / 2s C represents the coefficients of the 3x2 transformation. 2r / 2s This represents the transformation coefficient from stationary coordinates to rotating coordinates.

[0118] Based on the dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, the fractional-order integral order λ is introduced to construct the fractional-order current regulator.

[0119] Optionally, the current deviation module 530 is specifically used for: The magnitude I of the inverter load-side given current in the predicted power at the next moment. * ,and , After multiplication, the corresponding given values ​​of the d-axis current and q-axis current are obtained respectively. , ,in The load-side voltage phase is measured after passing through the phase-locked loop; For grid-connected current After sampling coefficients and coordinate transformation, the d-axis and q-axis components are obtained respectively. and ; The calculation yields the current deviation along the d-axis at the output of each inverter at the next moment. Current deviation along the q-axis at the next moment .

[0120] Optionally, the adaptive virtual impedance module 540 is specifically used for: The reactive power in the positive and negative sequence power values ​​of each inverter in the predicted power at the next moment is subtracted to obtain the power difference value ΔQ; The power difference ΔQ is combined with the integral term K v The product operation is performed on / S to obtain the adaptive virtual impedance Z at the next time step. v =K v / S*ΔQ, where K v This represents the adaptive adjustment coefficient, and S represents the calculation transformation method used to convert the power value into a resistance value.

[0121] Optionally, the virtual impedance control module 550 is specifically used for: The difference between the d-axis reference voltage after introducing the adaptive virtual impedance at the next moment and the voltage value corresponding to the adaptive virtual impedance at the next moment is used to obtain the reference voltage of the inverter's output d-axis at the next moment. The input voltage outer loop closed-loop transfer function is obtained by subtracting the reference voltage of the output terminal d-axis at the next moment from the d-axis component of the actual output voltage of the inverter at the next moment. The first current value is obtained by multiplying the reference voltage of the output terminal d-axis at the next moment with the first transformation parameter; The sum of the first current value and the second current value is then subtracted from the third current value to obtain the current deviation of the inverter output terminal d-axis at the next moment. The second current value is the output value of the voltage outer loop closed-loop transfer function. The current deviation of the inverter output terminal d-axis at the next moment is used as the input of the fractional current regulator. The output value of the fractional current regulator is multiplied by the inverter fundamental pulse width modulation ratio coefficient to obtain the d-axis component of the inverter reference output voltage at the next moment. The difference between the d-axis component of the inverter reference output voltage at the next moment and the d-axis component of the inverter actual output voltage at the next moment is calculated, and then multiplied with the second transformation parameter. The resulting current value is then compared with the d-axis component of the unbalanced load current to obtain the third current value. The third current value and the third transformation parameter are multiplied to obtain the d-axis component of the actual output voltage of the inverter at the next moment. The reference voltage of the q-axis at the inverter output terminal at the next moment is obtained by subtracting the voltage value corresponding to the adaptive virtual impedance at the next moment from the adaptive virtual impedance at the next moment. The difference between the reference voltage of the output terminal at the next q-axis moment and the q-axis component of the actual output voltage of the inverter at the next q-axis moment is then input into the voltage outer loop closed-loop transfer function. The fourth current value is obtained by multiplying the reference voltage of the output terminal q-axis at the next moment with the first transformation parameter; The sum of the fourth and fifth current values ​​is then subtracted from the sixth current value to obtain the current deviation of the inverter output terminal q axis at the next moment. The fifth current value is the output value of the voltage outer loop closed-loop transfer function. The current deviation of the inverter output terminal q-axis at the next moment is used as the input of the fractional current regulator. The output value of the fractional current regulator is multiplied by the inverter fundamental pulse width modulation ratio coefficient to obtain the q-axis component of the inverter reference output voltage at the next moment. The difference between the q-axis component of the inverter reference output voltage at the next moment and the q-axis component of the inverter actual output voltage at the next moment is calculated, and then multiplied with the second transformation parameter. The resulting current value is then compared with the q-axis component of the unbalanced load current to obtain the sixth current value. The sixth current value and the third transformation parameter are multiplied to obtain the q-axis component of the actual output voltage of the inverter at the next moment. Specifically, based on the dq coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, the fractional-order integral order λ is introduced to construct the fractional-order current regulator, including: Based on the abc coordinate coefficient mathematical model of the three-phase fractional-order LCL inverter, and combined with the PID control principle, the closed-loop transfer function of the voltage outer loop is obtained. G v ( s ); The voltage outer loop closed-loop transfer function G v ( s Based on this, the fractional integral order λ is introduced to construct the fractional current regulator. Wherein, the voltage outer loop closed-loop transfer function G v ( s The expression for ) is as follows:

[0122] In the above formula, Kvp , K vi Both represent the scaling factor of the virtual impedance; The fractional-order current regulator The expression is as follows:

[0123] In the above formula, K fpi This represents the proportionality coefficient of the current loop. K fii This represents the integral coefficient of the current loop.

[0124] Optionally, the control signal module 560 is specifically used for: The SPWM modulation signal of the inverter's d-axis at the next moment is obtained by subtracting the d-axis component of the actual output voltage of the inverter from the d-axis component of the fractional capacitor voltage. The SPWM modulation signal of the inverter's q-axis at the next moment is obtained by subtracting the q-axis component of the actual output voltage of the inverter from the q-axis component of the fractional capacitor voltage. The SPWM modulation signal of the inverter's d-axis at the next moment and the SPWM modulation signal of the inverter's q-axis at the next moment together constitute the SPWM control signal at the next moment.

[0125] Optionally, the prediction module 520 constructs the Romberg state observer using the following methods: Based on LCL type inverter The mathematical model in the coordinate system is used to calculate the discrete prediction model of the inverter-side current of each inverter. Among them, the LCL type inverter is in The mathematical model in the coordinate system is expressed as follows:

[0126] Discretizing the above, we obtain the following equation:

[0127] And thus obtain The expression for the output current prediction model in the coordinate system is as follows:

[0128] In the above formula, for The predicted value of the grid-side current at time k+1; for The current on the inverter side of the shaft is k The predicted value at time +1; for The grid voltage of the axis is k The sampled value at time; for The voltage of the shaft filter capacitor is at k The sampled value at time; for The current on the grid side of the axis is k The sampled value at time; for The current on the inverter side of the shaft is k The sampled value at time; The system sampling period; When multiple inverters are running stably, and the voltage of the filter capacitor is stable, the following expression holds:

[0129] The discrete prediction model for the inverter-side current can be obtained by calculation using the above formula, and its expression is as follows:

[0130] Based on the inverter voltage and current equations, the state-space model of the LCL inverter circuit is constructed as follows:

[0131] Selecting the state vector Take the input vector ,in: , , ,

[0132] Then we have:

[0133] Selecting state vectors from the state-space model The resulting exact discrete-time model changes as follows:

[0134] in: , ,

[0135] ,

[0136] In the above formula, u g Indicates the grid connection voltage. u 0 indicates the AC side voltage of the inverter. i 1k This indicates the inverter output current.u Ck Indicates the three-phase grid voltage at the grid connection point. i 2k This represents the measured current, where T represents the time constant. Indicates the grid-connected voltage frequency. Indicates the resonant frequency; Substituting the coefficients, we get:

[0137]

[0138]

[0139] in, This is the resonant frequency of the filter; The feedback value in the precise discrete-time model x d ( k +1) The expression is as follows:

[0140] Among them, let , , , Then there is ; Based on the above parameter values, the values ​​of each matrix in the state-space model expression are calculated. Ignoring grid voltage disturbances, the equations for the Romberg state observer are constructed as follows:

[0141] The above transformation is as follows:

[0142] in, Let be the feedback gain matrix of the Romberg state observer. Then the above equation becomes the following equation:

[0143] To obtain the error equation of the Romberg state observer, the following equation is given:

[0144] Define error vector Then the error equation becomes the following:

[0145] The dynamic performance of the error vector depends on the matrix. eigenvalues.

[0146] In summary, this application proposes a state observer fractional-order control method for multi-inverter parallel operation, and introduces a PI control method based on variable virtual impedance. λ The D-adaptive control method mainly uses the integral element to eliminate the static error of reactive power, while adding an adjustable parameter λ to improve the shortcomings of traditional control strategies, such as being too simple and having poor steady-state performance.

[0147] Design a fractional-order current regulator A Romberg state observer was constructed, and a novel approach was proposed to use the obtained accurate next-moment predicted power for subsequent calculations. The calculated current deviation was then fed into a fractional-order current regulator to form a feedback loop that modulates the SPWM pulse signal for the next moment.

[0148] Simultaneously, a multi-loop control strategy is designed, creatively proposing to use the obtained accurate next-time predicted power for the calculation of adaptive virtual impedance. An adaptive virtual impedance outer-loop control strategy is proposed, which calculates the difference in reactive power after positive and negative sequence separation for each inverter at the next time step, and then adds an integral term to the difference to obtain the virtual impedance as a fractional-order current regulator. The input is used to perform fractional-order adaptive virtual impedance d-axis and q-axis control, outputting the d-axis and q-axis components of the actual output voltage of each inverter at the next moment, generating the SPWM control signal for the next moment. This achieves zero steady-state error tracking of reactive power, gradually reducing the line impedance difference between multiple inverters with unequal capacity until they tend to be consistent, greatly improving the dynamic and static performance of the multi-inverter parallel system and ensuring that the bus voltage remains basically constant. The fractional-order control method proposed in this application has broad application prospects and high practicality.

[0149] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0150] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0151] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.

Claims

1. A state observer fractional order control method for multi-inverter parallel connection, characterized in that, The method comprises the following steps: constructing a fractional order current regulator of each inverter based on a mathematical model of the fractional order inverter; constructing a Luenberger state observer, and predicting internal state quantities of the multi-inverter grid-connected system by using grid-connected voltage and current values at a current moment and the Luenberger state observer to obtain predicted power at a next moment; calculating current deviation of a d-axis and a q-axis of an output end of each inverter at the next moment according to d-axis and q-axis components of a given current and a grid-connected current of each inverter at the load side in the predicted power at the next moment; calculating self-adaptive virtual impedance of each inverter at the next moment according to positive and negative sequence power values of each inverter in the predicted power at the next moment; taking the current deviation of the d-axis and the current deviation of the q-axis of the output end of each inverter at the next moment and the self-adaptive virtual impedance of each inverter at the next moment as inputs of the fractional order current regulator to perform fractional order self-adaptive virtual impedance d-axis and q-axis control and output d-axis and q-axis components of an actual output voltage of each inverter at the next moment; generating SPWM control signals at the next moment based on the d-axis and q-axis components of the actual output voltage of each inverter at the next moment to control operation of each inverter, eliminating static error of reactive power when the unequal-capacity multi-inverters are connected in parallel at the next moment, and gradually reducing differences in line impedance between the unequal-capacity multi-inverters until the differences tend to be consistent.

2. The state observer fractional order control method according to claim 1, characterized in that, The method for constructing a fractional order current regulator of each inverter based on a mathematical model of the fractional order inverter comprises the following steps: introducing fractional order inductance and fractional order capacitance to construct a three-phase fractional order LCL type inverter circuit; establishing an abc coordinate system mathematical model of the three-phase fractional order LCL type inverter by using a switching function description method; based on the abc coordinate system mathematical model, establishing a DQ coordinate system mathematical model of the three-phase fractional order LCL type inverter, and further establishing a dq coordinate system mathematical model; based on the dq coordinate system mathematical model, introducing a fractional order integral order, and further constructing a fractional order current regulator of each inverter, wherein the fractional order current regulator comprises a d-axis fractional order current regulator and a q-axis fractional order current regulator.

3. The state observer fractional order control method according to claim 2, characterized in that, The abc coordinate system mathematical model of the three-phase fractional order LCL type inverter is expressed as follows: In the above formula, L1 represents the fractional-order inductor on the inverter side, and its order is... L2 represents the fractional-order inductance on the grid side, with its order being... C represents the fractional capacitance on the AC side, and its order is 1. order , , The range of values ​​for i is (0, 2). 1_abc Indicates the inverter output current, i 2_abc This represents the grid-side current, i.e., the grid-connected current, where t represents the time constant, and u... abc Indicates AC voltage measurement of the inverter, U C_abc U represents the fractional-order filter capacitor voltage. g_abc Indicates the three-phase grid voltage at the grid connection point, u oo’_abc Ps represents the voltage difference between the midpoint of the power grid and the midpoint of the filter capacitor. abc u dc Indicates the transformation coefficients of the 32 coordinate system; The DQ coordinate system mathematical model of the three-phase fractional order LCL type inverter is expressed as follows: After simplifying the above formula, the corresponding simplified model is expressed as follows: In the above formula, C 2s / 3s denotes the coefficients of the inverse Clarke transformation, i 1_DQ denotes the D, Q components of the inverter input current, i 2_DQ denotes the D, Q components of the inverter output current, u i_DQ denotes the D, Q components of the inverter ac-side voltage, u c_DQ denotes the D, Q components of the fractional order capacitor voltage, s DQ denotes the D, Q components of the switching function, u dc denotes the dc-side voltage, U g_DQ denotes the D, Q components of the inverter three-phase grid voltage, u oo’_DQ denotes the D, Q components of the voltage difference between the grid midpoint and the filter capacitor midpoint, The dq coordinate system mathematical model of the three-phase fractional order LCL type inverter is expressed as follows: In the above equations, denote the d, q components of the inverter input current, respectively, denote the d, q components of the inverter output current, respectively; denote the d, q components of the inverter three-phase grid voltage, respectively; denote the d, q components of the fractional order capacitor voltage, respectively; denote the d, q components of the inverter AC side voltage, respectively; denote the d, q components of the switching function, respectively; C 3s / 2s denote the coefficients of the two-to-three transformation, C 2r / 2s denote the transformation coefficients from stationary to rotating coordinates; Based on the dq coordinate system mathematical model of the three-phase fractional order LCL type inverter, the fractional order integral order λ is introduced, and the fractional order current regulator is further constructed.

4. The state observer fractional order control method according to claim 1, characterized in that, The method for calculating current deviation of a d-axis and a q-axis of an output end of each inverter at the next moment according to d-axis and q-axis components of a given current and a grid-connected current of each inverter at the load side in the predicted power at the next moment comprises the following steps: the amplitude I of the given current on the inverter load side for the predicted power at the next moment * , and , after multiplication, the given values of the corresponding d-axis current and q-axis current are obtained respectively , , wherein is the phase of the load side voltage measured after the phase-locked loop To grid current After sampling coefficient and coordinate transformation, d-axis component and q-axis component are obtained respectively And ; The operation obtains the current deviation of the d-axis of each inverter output end at the next time , the current deviation of the q-axis at the next time .

5. The state observer fractional order control method according to claim 1, characterized in that, The method for calculating self-adaptive virtual impedance of each inverter at the next moment according to positive and negative sequence power values of each inverter in the predicted power at the next moment comprises the following steps: The reactive power in the positive and negative sequence power values of each inverter in the predicted power of the next moment is subtracted to obtain a power difference value ΔQ; The power difference ΔQ is multiplied by an integral term K v to obtain the adaptive virtual impedance Z v = K v / S*ΔQ, where K v represents an adaptive adjustment coefficient, and S represents a calculation transformation mode for converting the power value into an impedance value.

6. The state observer fractional order control method according to claim 3, characterized in that, The d-axis current deviation of each inverter output end at the next moment and the q-axis current deviation at the next moment, and the adaptive virtual impedance of each inverter at the next moment are taken as inputs of the fractional order current regulator to perform fractional order adaptive virtual impedance d-axis and q-axis control, and the d-axis component and the q-axis component of the actual output voltage of each inverter at the next moment are output, comprising: The d-axis reference voltage after introducing the adaptive virtual impedance at the next moment is subtracted from the voltage value corresponding to the adaptive virtual impedance at the next moment to obtain the reference voltage of the output end d-axis of the inverter at the next moment; The reference voltage of the output end d-axis at the next moment is input into the voltage outer loop closed loop transfer function after being subtracted from the d-axis component of the actual output voltage of the inverter at the next moment; The reference voltage of the output end d-axis at the next moment is multiplied by the first conversion parameter to obtain a first current value; The first current value, the second current value are summed and then subtracted from the third current value to obtain the current deviation of the output end d-axis of the inverter at the next moment, wherein the second current value is the output value of the voltage outer loop closed loop transfer function; The current deviation of the output end d-axis of the inverter at the next moment is taken as the input of the fractional order current regulator, and the output value of the fractional order current regulator is multiplied by the fundamental wave pulse width modulation proportionality coefficient to obtain the d-axis component of the reference output voltage of the inverter at the next moment; The d-axis component of the reference output voltage of the inverter at the next moment is subtracted from the d-axis component of the actual output voltage of the inverter at the next moment, and then multiplied by the second conversion parameter to obtain a current value, which is subtracted from the d-axis component of the unbalanced load current to obtain the third current value; The third current value is multiplied by the third conversion parameter to obtain the d-axis component of the actual output voltage of the inverter at the next moment; The q-axis reference voltage after introducing the adaptive virtual impedance at the next moment is subtracted from the voltage value corresponding to the adaptive virtual impedance at the next moment to obtain the reference voltage of the output end q-axis of the inverter at the next moment; The reference voltage of the output end q-axis at the next moment is input into the voltage outer loop closed loop transfer function after being subtracted from the q-axis component of the actual output voltage of the inverter at the next moment; The reference voltage of the output end q-axis at the next moment is multiplied by the first conversion parameter to obtain a fourth current value; The fourth current value, the fifth current value are summed and then subtracted from the sixth current value to obtain the current deviation of the output end q-axis of the inverter at the next moment, wherein the fifth current value is the output value of the voltage outer loop closed loop transfer function; The current deviation of the output end q-axis of the inverter at the next moment is taken as the input of the fractional order current regulator, and the output value of the fractional order current regulator is multiplied by the fundamental wave pulse width modulation proportionality coefficient to obtain the q-axis component of the reference output voltage of the inverter at the next moment; The q-axis component of the next moment of the inverter reference output voltage is subtracted from the q-axis component of the next moment of the inverter actual output voltage, and then multiplied by the second conversion parameter to obtain a current value, and the q-axis component of the unbalanced load current is subtracted to obtain the sixth current value; The sixth current value is multiplied by the third conversion parameter to obtain the q-axis component of the next moment of the inverter actual output voltage.

7. The state observer fractional order control method according to claim 6, characterized in that, Based on the dq coordinate system mathematical model of the three-phase fractional order LCL type inverter, the fractional order integral order λ is introduced, and then the fractional order current regulator is constructed, including: Based on the mathematical model of the three-phase fractional order LCL type inverter in abc coordinate system, combining with the PID control principle, the voltage outer loop closed-loop transfer function is obtained G v ( s ); On the basis of the voltage outer loop closed-loop transfer function G v s , the fractional order integral order λ is introduced, and then the fractional order current regulator is constructed.​ The expression of the voltage outer loop closed-loop transfer function G v ( s ) is as follows: In the above formulae, K vp , K vi both represent the proportional coefficient of the virtual impedance; The fractional current regulator The expression is as follows: In the above formula, K fpi represents the proportional coefficient of the current loop, K fii represents the integral coefficient of the current loop.

8. The state observer fractional order control method according to claim 7, characterized in that, Based on the d-axis component and the q-axis component of the next moment of the actual output voltage of each inverter, the next moment of the SPWM control signal is generated, including: The d-axis component of the next moment of the inverter actual output voltage is subtracted from the d component of the fractional order capacitor voltage to obtain the SPWM modulation signal of the inverter d-axis of the next moment; The q-axis component of the next moment of the inverter actual output voltage is subtracted from the q component of the fractional order capacitor voltage to obtain the SPWM modulation signal of the inverter q-axis of the next moment; The SPWM modulation signal of the inverter d-axis of the next moment and the SPWM modulation signal of the inverter q-axis of the next moment together constitute the SPWM control signal of the next moment.

9. The state observer fractional order control method according to claim 3, characterized in that, A Luenberger state observer is constructed, including: Based on the mathematical model of LCL type inverter under the coordinate system, the discrete prediction model of inverter-side current of each inverter is obtained by calculation. In the LCL type inverter The mathematical model in the coordinate system is expressed as follows: Discretization is performed on the above to obtain the following formula: Further, we obtain The output current prediction model expression under the coordinate system is as follows: In the above formula, for The predicted value of the grid-side current at time k+1; for The current on the inverter side of the shaft is k The predicted value at time +1; for The grid voltage of the axis is k The sampled value at time; for The voltage of the shaft filter capacitor is at k The sampled value at time; for The current on the grid side of the axis is k The sampled value at time; for The current on the inverter side of the shaft is k The sampled value at time; The system sampling period; When the multiple inverters are stably operated, the filter capacitor voltage is stable, and the following expression is obtained: The above formula can obtain the discrete prediction model of the inverter side current, which is expressed as follows: Based on the inverter voltage and current equation, the state space model of the LCL type inverter circuit is constructed as follows: Selecting a state vector , taking an input vector wherein: ， ， ， Then there is: selecting a state vector for the state space model The post-formed accurate discrete-time model changes as follows: Wherein: ， ， ， in the above formula, u g represents the grid voltage, u 0 represents the inverter ac side voltage, i 1k represents the inverter output current, u Ck represents the grid point three-phase grid voltage, i 2k represents the grid current, T represents the time constant, represents the grid voltage frequency, represents the resonance frequency; The coefficients are brought into to obtain: wherein, is the resonance frequency of the filter; The feedback value in the precise discrete-time model x d ( k +1) is expressed as follows: wherein let , , , , then ; According to the values of the above parameters, the values of each matrix in the state space model expression are calculated, and under the condition of ignoring the grid voltage disturbance, the equation of the constructed Luenberger state observer is as follows: The above transformation is as follows: wherein is the feedback gain matrix of the Luenberger state observer, if then the above equation becomes the following equation: In order to obtain the error equation of the Luenberger state observer, the following formula is provided: Defining the error vector The error equation then becomes the following: The dynamic behavior of the error vector depends on the eigenvalues of the matrix ​ 10. A multi-inverter parallel state observer fractional order control device, characterized in that, Including: A regulator module is constructed for constructing a fractional order current regulator of each inverter based on the mathematical model of the fractional order inverter; A prediction module is used to predict the internal state quantity of the multiple inverter grid-connected system by using the grid-connected voltage value at the current moment and combining the Luenberger state observer to obtain the predicted power at the next moment; A current deviation module is used to calculate the current deviation of the d-axis and the q-axis of the output end of each inverter at the next moment according to the d-axis component and the q-axis component of the given current and the grid-connected current of each inverter at the next moment; An adaptive virtual impedance module is used to calculate the adaptive virtual impedance at the next moment according to the positive and negative sequence power values of each inverter at the next moment. a virtual impedance control module, configured to take each inverter output end d-axis current deviation and q-axis next time current deviation, each inverter next time adaptive virtual impedance as inputs of the fractional order current regulator, perform fractional order adaptive virtual impedance d-axis and q-axis control, and output d-axis component and q-axis component of each inverter actual output voltage next time; a control signal module, configured to generate SPWM control signals next time based on the d-axis component and q-axis component of each inverter actual output voltage next time, and control operation of each inverter, so as to eliminate reactive power static error in the next time when the unequal capacity multi-inverters are connected in parallel, and gradually reduce line impedance difference between the unequal capacity multi-inverters until the line impedance difference tends to be consistent.

Images