Data-driving-based weak connection type inverter network construction voltage support control method and data-driving-based weak connection type inverter network construction voltage support control system

Abstract

The invention relates to a weak connection type inverter network construction voltage support control method and system based on data driving, and the method comprises the following steps: building a mathematical model of a weak connection type micro-grid inverter under a two-phase static coordinate system for the topology of the weak connection type micro-grid inverter, constructing a discrete equation of the inverter in combination with a forward Euler method, and carrying out the network construction voltage support control of the weak connection type micro-grid inverter. On the basis of a continuous control set model prediction method, a network construction capacitor voltage prediction equation during multi-vector application is designed; establishing a data driving model according to the mathematical model of the weak connection type micro-grid inverter under the two-phase static coordinate system and the grid construction capacitor voltage prediction equation, updating parameters of the data driving model by combining an adaptive tracking algorithm, and predicting the data driving capacitor voltage; the influence of noise on a data driving model when the noise exists in voltage and current data in the weak connection type microgrid inverter is considered, and the influence of the noise on the data driving model is introduced in the process of predicting the voltage of a data driving capacitor.

Description

Technical Field

[0001] The present invention relates to a data-driven weak-connected inverter network voltage support control method and system, belonging to the technical field of power control. Background Art

[0002] A weakly connected microgrid is a distributed power system that can be flexibly connected or disconnected from the main power grid. It usually includes local renewable energy generation (such as solar and wind power), inverter systems, energy storage systems (such as batteries), and load management systems. In a weakly connected microgrid, the inverter can convert the DC power generated by the renewable energy generation system into AC power through an LC filter, providing stable AC output for the energy storage and load systems, and achieving grid voltage support.

[0003] Due to its superior dynamic performance and multi-objective control characteristics, the finite control set model predictive control (FCS-MPC) method has been widely used in weakly connected microgrid inverters to achieve efficient and accurate grid voltage support control. The FCS-MPC method designs a finite control set based on the LC filter inverter topology, uses its discrete model to predict the future state, and determines the optimal state from the finite control set through a cost function. However, the traditional FCS-MPC method lacks a modulation stage, which results in significant ripples in the inverter grid voltage and fluctuating switching frequency.

[0004] In order to improve the grid voltage support performance of the traditional FCS-MPC method in weakly connected microgrid inverter applications, researchers have developed a variety of continuous control set model predictive control (CCS-MPC) methods. These methods use multiple voltage vectors in each control cycle and calculate their duration to reduce voltage fluctuations and stabilize the switching frequency. Although the CCS-MPC method can effectively improve the inverter grid voltage support performance, the method is very sensitive to the difference between the LC filter model parameters and the actual parameters. If the model parameters of the LC filter do not match the actual parameters, the grid voltage prediction will be inaccurate, which will cause grid voltage distortion and even damage the weakly connected microgrid inverter.

[0005] In order to improve the parameter robustness of traditional model predictive control, researchers proposed a model-free predictive control (MFPC) method, which uses various data of current state and past behavior to predict the future state of the system without relying on an accurate model. Existing MFPC methods are mainly divided into three categories: ① MFPC method based on super-local model; ② MFPC method based on autoregressive model with exogenous input; ③ MFPC method based on voltage-current gradient. However, the first two MFPC methods require the use of system data of a large number of control cycles to achieve accurate modeling, which affects the dynamic performance of the network voltage control of the weakly connected microgrid inverter. At the same time, the modeling accuracy is sensitive to system noise. The third method relies on accurate voltage-current gradients and requires system modeling to achieve real-time gradient updates. This method is also sensitive to system noise. These problems will also affect the network voltage support performance of the weakly connected microgrid inverter. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the present invention proposes a data-driven weak-connected inverter network voltage support control method and system.

[0007] The technical solution of the present invention is as follows:

[0008] On the one hand, the present invention proposes a data-driven weak-connected inverter network voltage support control method, comprising the following steps:

[0009] Aiming at the topology of weakly connected microgrid inverter, a mathematical model of weakly connected microgrid inverter in two-phase stationary coordinate system is established. The discrete equation of the inverter is constructed by combining the forward Euler method. Based on the model prediction method of continuous control set, the grid capacitor voltage prediction equation is designed when multi-vector is applied.

[0010] According to the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation, a data-driven model is established, and the parameters of the data-driven model are updated in combination with the adaptive tracking algorithm to predict the data-driven capacitor voltage.

[0011] Considering the influence of noise on the data-driven model when there is noise in the voltage and current data of the weakly connected microgrid inverter, the influence of noise on the data-driven model is introduced in the process of predicting the data-driven capacitor voltage to ensure the anti-noise ability and dynamic performance of the inverter grid voltage support control.

[0012] As a preferred implementation, the model prediction method based on the continuous control set is designed to design the network capacitor voltage prediction equation when the multi-vector application is specifically:

[0013]

[0014] Among them, v mb cαβ_x (k+1) is the predicted value of the capacitor voltage at time k+1 in the two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ (k) is the capacitor voltage measurement value at time k in the two-phase stationary coordinate system; g mb αβ_x (k) is the capacitor voltage gradient at time k in the two-phase stationary coordinate system, which is equal to the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding;

[0015] Among them, g mb αβ_x (k) is expressed as:

[0016]

[0017] Among them, i mb iα β _x (k+1) is the predicted value of the inverter side current at time k+1 in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ is the capacitor voltage on the inverter side in the two-phase stationary coordinate system; t is time; i iαβ (k) is the inverter side current at time k in the two-phase stationary coordinate system; C is the filter capacitor model parameter.

[0018] As a preferred implementation, the steps of establishing a data-driven model according to the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system and the grid capacitor voltage prediction equation are specifically as follows:

[0019] The capacitor voltage gradient at time k in the two-phase stationary coordinate system is transformed into:

[0020]

[0021] in, is the capacitor voltage gradient after transformation, i lαβ (k) is the load side current in the two-phase stationary coordinate system, L is the filter inductance model parameter, T s is the control period, v vsαβ is the virtual voltage vector of the inverter side in the two-phase stationary coordinate system;

[0022] According to the capacitor voltage gradient at time k in the two-phase stationary coordinate system after the above transformation, the data-driven model is established as follows:

[0023]

[0024] in, is the capacitor voltage gradient based on data at time k in the two-phase stationary coordinate system, where:

[0025]

[0026] Among them, A(k) is the set of known and unknown variables, B(k) is the set of system disturbances, and I αβ (k) is the current set at time k in the two-phase stationary coordinate system; V αβ (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponding;

[0027] The matrix form of the data-driven model is expressed as:

[0028]

[0029] in:

[0030]

[0031] Among them, y(k) is the input variable of the data-driven model, is the dynamic variable of the data-driven model, are the estimated variables for the data-driven model.

[0032] As a preferred implementation, the step of updating the parameters of the data-driven model in combination with the adaptive tracking algorithm to predict the data-driven capacitor voltage includes:

[0033] According to the adaptive tracking algorithm, the set of known and unknown variables A(k) and the system disturbance set B(k) in the data-driven model are estimated accordingly. The recursive matrix G(k) of the adaptive tracking algorithm is expressed as:

[0034]

[0035] in, express The estimated value of ; λ is the adaptive adjustment factor, expressed as:

[0036]

[0037] Among them, λ min is the minimum value of the adaptive adjustment factor, γ is a positive coefficient, and e0 is the error boundary value;

[0038] according to Get the data driving voltage gradient g corresponding to the voltage vector on the inverter side in the two-phase stationary coordinate systemmf αβ_x (k):

[0039]

[0040] Among them, V αβ_x (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding;

[0041] According to the data drive voltage gradient g mf αβ_x (k) and the virtual voltage vector, and obtain the capacitor voltage prediction value based on data driving.

[0042] As a preferred implementation, the method for introducing the influence of noise on the data driving model during the process of predicting the data driving capacitor voltage is specifically as follows:

[0043] The data driving voltage gradient is transformed into a data driving voltage gradient considering noise suppression, specifically:

[0044]

[0045] in, In order to consider the data driving voltage gradient of noise suppression, ρ is an adaptive adjustment coefficient, which is adjusted according to the noise size, and 0<ρ<1.

[0046] On the other hand, the present invention also proposes a data-driven weak-connected inverter network voltage support control system, comprising:

[0047] The mathematical model building module is used to establish the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system for the weakly connected microgrid inverter topology, and to construct the discrete equation of the inverter in combination with the forward Euler method. Based on the model prediction method of the continuous control set, the grid capacitor voltage prediction equation is designed when multi-vector applications are used;

[0048] The data-driven model building module is used to establish a data-driven model based on the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation, and update the parameters of the data-driven model in combination with the adaptive tracking algorithm to predict the data-driven capacitor voltage;

[0049] The noise elimination module is used to consider the influence of noise on the data-driven model when there is noise in the voltage and current data of the weakly connected microgrid inverter. The influence of noise on the data-driven model is introduced in the process of predicting the data-driven capacitor voltage to ensure the anti-noise ability and dynamic performance of the inverter grid voltage support control.

[0050] As a preferred implementation, the model prediction method based on the continuous control set is designed to design the network capacitor voltage prediction equation when the multi-vector application is specifically:

[0051]

[0052] Among them, v mb cαβ_x (k+1) is the predicted value of the capacitor voltage at time k+1 in the two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ (k) is the capacitor voltage measurement value at time k in the two-phase stationary coordinate system; g mb αβ_x (k) is the capacitor voltage gradient at time k in the two-phase stationary coordinate system, which is equal to the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding;

[0053] Among them, g mb αβ_x (k) is expressed as:

[0054]

[0055] Among them, i mb iα β _x (k+1) is the predicted value of the inverter side current at time k+1 in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ is the capacitor voltage on the inverter side in the two-phase stationary coordinate system; t is time; i iαβ (k) is the inverter side current at time k in the two-phase stationary coordinate system; C is the filter capacitor model parameter.

[0056] As a preferred implementation, the steps of establishing a data-driven model according to the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system and the grid capacitor voltage prediction equation are specifically as follows:

[0057] The capacitor voltage gradient at time k in the two-phase stationary coordinate system is transformed into:

[0058]

[0059] in, is the capacitor voltage gradient after transformation, i lαβ (k) is the load side current in the two-phase stationary coordinate system, L is the filter inductance model parameter, T s is the control period, v vsαβis the virtual voltage vector of the inverter side in the two-phase stationary coordinate system;

[0060] According to the capacitor voltage gradient at time k in the two-phase stationary coordinate system after the above transformation, the data-driven model is established as follows:

[0061]

[0062] in, is the capacitor voltage gradient based on data at time k in the two-phase stationary coordinate system, where:

[0063]

[0064] Among them, A(k) is the set of known and unknown variables, B(k) is the set of system disturbances, and I αβ (k) is the current set at time k in the two-phase stationary coordinate system; V αβ (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponding;

[0065] The matrix form of the data-driven model is expressed as:

[0066]

[0067] in:

[0068]

[0069] Among them, y(k) is the input variable of the data-driven model, is the dynamic variable of the data-driven model, are the estimated variables for the data-driven model.

[0070] On the other hand, the present invention also proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements a data-driven weak-connected inverter network voltage support control method as described in any embodiment of the present invention.

[0071] On the other hand, the present invention further proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a data-driven weak-connected inverter network voltage support control method as described in any embodiment of the present invention.

[0072] Additional aspects and advantages of the present invention will be set forth in the following description, and some of them will be apparent from the description, or may be understood by practicing the present invention. In addition, the various aspects and advantages of the present invention may be realized and obtained by the method steps and combinations particularly pointed out in the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 This is a schematic diagram of a method flow chart of Embodiment 1 of the present invention;

[0074] Figure 2 A schematic diagram of a topology of a weakly connected microgrid inverter in an embodiment of the present invention;

[0075] Figure 3 It is a voltage vector diagram of a weakly connected microgrid inverter in an embodiment of the present invention;

[0076] Figure 4 It is a graph of action time error when parameter error exists in an embodiment of the present invention;

[0077] Figure 5 is a graph of capacitance voltage error when parameter error exists in an embodiment of the present invention;

[0078] Figure 6 A schematic diagram of data-driven model update and prediction in an embodiment of the present invention;

[0079] Figure 7 is a cost function diagram in the presence of noise in an embodiment of the present invention;

[0080] Figure 8 This is a data-driven prediction impact analysis diagram in the presence of noise in an embodiment of the present invention;

[0081] Fig. 9 This is a diagram for analyzing the effectiveness of the adaptive adjustment coefficient in an embodiment of the present invention;

[0082] Fig.10 It is a control block diagram of a method based on embodiment 1 of the present invention. DETAILED DESCRIPTION

[0083] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0084] It should be understood that the step numbers used in this document are only for convenience of description and are not intended to limit the order in which the steps are executed.

[0085] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0086] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0087] The term "and / or" means and includes any and all possible combinations of one or more of the associated listed items.

[0088] Embodiment 1:

[0089] For the grid voltage support control of weakly connected microgrid inverters, if the traditional model predictive control method is used, the system grid voltage support performance is very sensitive to the difference between the LC filter model parameters and the actual parameters. If the model parameters of the LC filter do not match the actual parameters, the grid voltage prediction will be inaccurate, which will cause grid voltage distortion and even damage the weakly connected microgrid inverter. If the traditional model-free predictive control method is used, its system modeling will be affected by the amount of system voltage and current data and system noise, thus affecting the grid voltage support performance of the weakly connected microgrid inverter.

[0090] This embodiment proposes a data-driven weak-connected inverter grid voltage support control method, establishes a data model of a weak-connected microgrid inverter, and combines an adaptive tracking algorithm to achieve real-time online update of the data model; at the same time, the present invention takes into account the impact of system noise on the data model, designs an adaptive adjustment coefficient, adjusts the data model gain in real time, and reduces the sensitivity of grid voltage support control to noise. The present invention eliminates the impact of model parameters and system noise at the same time, ensuring the grid voltage support control performance of the weak-connected microgrid inverter.

[0091] See Figure 1 The method provided in this embodiment specifically includes the following steps:

[0092] S100, for Figure 2 The weakly connected microgrid inverter topology shown in the figure is used to establish the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system. The discrete equation of the inverter is constructed by combining the forward Euler method. Based on the model prediction method of the continuous control set, the grid capacitor voltage prediction equation is designed when multi-vector is applied, and its parameter sensitivity is analyzed.

[0093] S200, based on the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation, a data-driven model is established, and combined with an adaptive tracking algorithm, real-time online update of the data-driven model is performed to achieve data-driven capacitor voltage prediction and eliminate the influence of model parameters on grid voltage support control;

[0094] S300. Analyze the influence of noise on the accuracy of the inverter data-driven model when the voltage and current data in the weakly connected microgrid inverter system are noisy. In the process of predicting the data-driven capacitor voltage, introduce the influence of noise on the data-driven model to ensure the anti-noise capability and dynamic performance of the inverter grid voltage support control.

[0095] In one embodiment, step S100 specifically includes:

[0096] according to Figure 2 The topology diagram of the weakly connected microgrid inverter is shown in Figure 1. Its mathematical model in the αβ two-phase stationary coordinate system can be expressed as:

[0097]

[0098] Where: i iαβ 、i lαβ 、v xαβ 、v cαβ They are the inverter side current, load side current, inverter side voltage vector, and capacitor voltage in the αβ two-phase stationary coordinate system. xαβ x∈0~7, such as Figure 3 As shown, they are v 0αβ (0,0,0),v 1αβ (1,0,0),v 2αβ (1,1,0),v 3αβ (0,1,0),v 4αβ (0,1,1),v 5αβ (0,0,1),v 6αβ (1,0,1),v 7αβ (1,1,1). L is the filter inductor model parameter, and C is the filter capacitor model parameter.

[0099] According to the front-line Euler method and the system model shown in formula (1), the inverter side current can be discretized and expressed as:

[0100]

[0101] Where: i mb iαβ_x (k+1) is the inverter side current prediction value based on the model at time k+1 in the αβ two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the αβ two-phase stationary coordinate system.xαβ (k) corresponds. i iαβ (k) is the measured current value of the inverter side at time k in the αβ two-phase stationary coordinate system. cαβ (k) is the capacitor voltage measurement value at time k in the αβ two-phase stationary coordinate system. s To control the cycle.

[0102] At the same time, based on the system model of formula (1), the capacitor voltage prediction value based on the model can be expressed as:

[0103]

[0104] Where: v mb cαβ_x (k+1) is the capacitor voltage prediction value based on the model at time k+1 in the αβ two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ (k) corresponds. cαβ (k) is the measured value of the capacitor voltage at time k in the αβ two-phase stationary coordinate system. mb αβ_x (k) is the capacitor voltage gradient based on the model at time k in the αβ two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ (k) corresponds to. According to formula (1), the capacitor voltage gradient g based on the model mb αβ_x (k) can be expressed as:

[0105]

[0106] In order to improve the grid voltage performance and fix the system switching frequency, the inverter side voltage vector v in the αβ two-phase stationary coordinate system can be used xαβ The virtual voltage vector v on the inverter side in the synthetic αβ two-phase stationary coordinate system vsαβ , which can be expressed as:

[0107]

[0108] Where: v xαβ_1 is the real voltage vector on the first inverter side in the αβ two-phase stationary coordinate system; v xαβ_2 is the real voltage vector on the second inverter side in the αβ two-phase stationary coordinate system; v xαβ_0 is the zero-side voltage vector of the inverter in the αβ two-phase stationary coordinate system. t1 is the action time of the real voltage vector on the first inverter side. t2 is the action time of the real voltage vector on the second inverter side. t0 is the action time of the zero voltage vector on the inverter side.

[0109] Combining the capacitor voltage prediction value shown in formula (3) and the virtual voltage vector shown in formula (5), we can get:

[0110]

[0111] Where: v mb cαβ_s (k+1) is the capacitor voltage prediction value based on the model at time k+1 in the αβ two-phase stationary coordinate system, which is consistent with the virtual voltage vector v on the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponds to g. mb αβ_x1 (k) is the capacitor voltage gradient based on the model at time k in the αβ two-phase stationary coordinate system, which is consistent with the real voltage vector v on the first inverter side at time k in the αβ two-phase stationary coordinate system. xαβ_1 Corresponding. mb αβ_x2 (k) is the capacitor voltage gradient based on the model at time k in the αβ two-phase stationary coordinate system, which is consistent with the real voltage vector v on the second inverter side at time k in the αβ two-phase stationary coordinate system. xαβ_2 Corresponding. mb αβ_x0 (k) is the capacitor voltage gradient based on the model at time k in the αβ two-phase stationary coordinate system, which is consistent with the inverter side zero voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ_0 Corresponding.

[0112] According to the minimum solution conditions, the action times t1, t2 and t0 can be obtained:

[0113]

[0114] Where: J(t1,t2) is the cost function containing t1 and t2, which can be expressed as:

[0115]

[0116] Where: v ref cα (k+1) is the reference value of the capacitor voltage at time k+1 in the α stationary coordinate system, v ref cβ (k+1) is the reference value of the capacitor voltage at time k+1 in the β stationary coordinate system.

[0117] From (2) to (8), it can be seen that the action times t1, t2 and t0 in the virtual voltage vector synthesis are determined by the filter inductor model parameter L and the filter capacitor model parameter C. When the filter inductor model parameter L is equal to the actual filter inductor parameter L a Mismatch, filter capacitor model parameter C and filter capacitor actual parameter C aIf they do not match, the action times t1, t2 and t0 will be inaccurate. Figure 4 is the action time error t when the model parameters do not match the actual parameters 1err ,t 2err and t 0err It can be seen that the action time error t 1err ,t 2err and t 0err As the parameter error increases, it will affect the system grid voltage performance.

[0118] From (2) to (8), we can see that the capacitor voltage prediction value v based on the model at time k+1 in the αβ two-phase stationary coordinate system is mb cαβ_s (k+1) is also determined by the filter inductor model parameter L and the filter capacitor model parameter C. When the filter inductor model parameter L is equal to the actual filter inductor parameter L a Mismatch, filter capacitor model parameter C and filter capacitor actual parameter C a When the capacitor voltage is not matched, the predicted value v mb cαβ_s (k+1) is not accurate. Figure 5 is the three-phase capacitor voltage prediction error v when the model parameters do not match the actual parameters caerr 、v cberr and v ccerr It can be seen that the three-phase capacitor voltage prediction error v caerr 、v cberr and v ccerr As the parameter error increases, it will affect the system grid voltage performance.

[0119] In one embodiment, step S200 is specifically:

[0120] Figure 6 This is the process of establishing, updating and predicting the data-driven model of the weakly-connected microgrid inverter proposed in this embodiment.

[0121] According to formulas (2) to (4), the capacitor voltage gradient g based on the model at time k in the αβ two-phase stationary coordinate system is mb αβ_x (k) can be further expressed as:

[0122]

[0123] According to formula (9), the data-driven model of the weakly connected microgrid inverter can be established and expressed as:

[0124]

[0125] Where: g mf αβ_x(k) is the capacitor voltage gradient based on data at time k in the αβ two-phase stationary coordinate system. Where:

[0126]

[0127] Where: A(k) is the set of known and unknown variables, B(k) is the set of system disturbances, I αβ (k) is the current set at time k in the αβ two-phase stationary coordinate system, V αβ (k) is the voltage set at time k in the αβ two-phase stationary coordinate system, which is consistent with the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponding.

[0128] According to A(k) and B(k), the data model of the weakly connected microgrid inverter can be obtained without calculating according to the mathematical model of the system. The matrix form of the data-driven model of the weakly connected microgrid inverter can be expressed as:

[0129]

[0130] in:

[0131]

[0132] Where: y(k) is the input variable of the inverter data-driven model, is the dynamic variable of the inverter data-driven model, are the estimated variables for the inverter data-driven model.

[0133] According to the adaptive tracking algorithm, the set of known and unknown variables A(k) and the system disturbance set B(k) in the inverter data-driven model can be estimated accordingly. The recursive matrix G(k) of the adaptive tracking algorithm can be expressed as:

[0134]

[0135] Where: express The estimated value of . λ is the adaptive adjustment factor, which can be expressed as:

[0136]

[0137] Where: min is the minimum value of the adaptive adjustment factor, γ is a positive coefficient, and e0 is the error boundary value.

[0138] According to the estimated values ​​derived from (13) to (14) The voltage vector v on the inverter side in the αβ two-phase stationary coordinate system can be obtained: 0αβ ~v 7αβThe corresponding data driving voltage gradient g mf αβ_x (k), as follows:

[0139]

[0140] Where: V αβ_x (k) is the voltage set at time k in the αβ two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ (k) Corresponding

[0141] According to the data driving voltage gradient g shown in formula (16): mf αβ_x And the virtual voltage vector shown in formula (5), the capacitor voltage prediction value v based on data drive can be obtained mf cαβ_s (k+1):

[0142]

[0143] Where: v mf cαβ_s (k+1) is the capacitor voltage prediction value based on data drive at time k+1 in the αβ two-phase stationary coordinate system, which is consistent with the inverter side virtual voltage vector v at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponds to g. mf αβ_x1 (k) is the capacitor voltage gradient based on data drive at time k in the αβ two-phase stationary coordinate system, which is consistent with the first inverter side real voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ_1 Corresponding. mf αβ_x2 (k) is the capacitor voltage gradient based on data drive at time k in the αβ two-phase stationary coordinate system, which is consistent with the real voltage vector v on the second inverter side at time k in the αβ two-phase stationary coordinate system. xαβ_2 Corresponding. mf αβ_x0 (k) is the capacitor voltage gradient based on data drive at time k in the αβ two-phase stationary coordinate system, which is consistent with the inverter side zero voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ_0 Corresponding.

[0144] In one embodiment, step S300 is specifically:

[0145] The influence of noise on the accuracy of the inverter data-driven model in the weakly connected microgrid inverter system is analyzed separately.

[0146] Generally, the voltage and current data sampling of weakly connected microgrid inverters is easily affected by sensor accuracy, sampling circuit and switching noise. The inverter side current, load side current and capacitor voltage with noise in the αβ two-phase stationary coordinate system can be expressed as:

[0147]

[0148] Where: i iαβ_me 、v cαβ_me 、i lαβ_me They are the inverter side current, capacitor voltage, and load side current with noise in the αβ two-phase stationary coordinate system. i noise , Δv c noise , Δi l noise They are inverter side current noise, capacitor voltage noise, and load side current noise respectively.

[0149] Substituting formula (18) into (13), we can obtain:

[0150]

[0151] Where: w v and w i are the noise disturbances of voltage and current respectively. In order to represent w v and w i The influence of voltage noise intensity δ v and the current noise intensity δ i It can be expressed as

[0152]

[0153] Where: v me and i me are the measured voltage and current respectively. ac and v ac are the actual voltage and current respectively. ref and i ref They are the references for voltage and current respectively.

[0154] Figure 7 is the cost function when there are different Chengdu noises on the inverter side current, load side current, and capacitor voltage. It can be observed that the cost function value changes with the voltage noise intensity δ v This will cause the capacitor voltage v to increase significantly with the increase of the data drive prediction. cαβ In addition, the cost function value is affected by the current noise intensity δ i Therefore, the effect of noise on the capacitor voltage v cαβ Compared with the influence of noise on the inverter side current iiαβ and load current i lαβ The impact of can be ignored.

[0155] The adaptive adjustment coefficient of the weakly connected microgrid inverter is designed to illustrate the effectiveness of the adjustment coefficient, and the data model gain is adjusted in real time according to the system error to ensure the anti-noise ability and dynamic performance of the inverter grid voltage support control:

[0156] A method for designing an adaptive adjustment coefficient of a weakly connected microgrid inverter is proposed to reduce the impact of voltage noise on data-driven predictive control. The data-driven voltage gradient g shown in (16) mf αβ_x In (k), an adaptive adjustment coefficient ρ is designed. At this time, the data driving voltage gradient g' is considered to suppress noise. mf αβ_x (k) can be expressed as

[0157]

[0158] Where ρ is the adaptive adjustment coefficient, 0<ρ<1. When the noise is large, the adaptive adjustment coefficient ρ can be reduced; otherwise, the adaptive adjustment coefficient ρ can be increased.

[0159] According to the data driving voltage gradient g' taking into account the noise suppression in (21), mf αβ_x The virtual voltage vector shown in (k) and (5) can be used to obtain the predicted value v' of the data drive capacitor voltage taking into account noise suppression. mf cαβ_s (k+1):

[0160]

[0161] Where: v' mf cαβ_s (k+1) is the predicted value of the data drive capacitor voltage at time k+1 taking into account noise suppression in the αβ two-phase stationary coordinate system, which is consistent with the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponds to g' mf αβ_x1 (k) is the data driving capacitor voltage gradient with noise suppression at time k in the αβ two-phase stationary coordinate system, which is equal to the first inverter side real voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ_1 Corresponding. g' mf αβ_x2 (k) is the voltage gradient of the data driving capacitor at time k in the αβ two-phase stationary coordinate system with noise suppression, which is equal to the real voltage vector v on the second inverter side at time k in the αβ two-phase stationary coordinate system. xαβ_2 Corresponding. g' mfαβ_x0 (k) is the data driving capacitor voltage gradient with noise suppression at time k in the αβ two-phase stationary coordinate system, which is equal to the inverter side zero voltage vector v at time k in the αβ two-phase stationary coordinate system. xαβ_0 Corresponding.

[0162] To ensure that the data drive capacitor voltage prediction value v' is taken into account for noise suppression mf cαβ_s (k+1) can track the reference value v ref cαβ (k+1), design the cost function G and express it as:

[0163]

[0164] Figure 8 The data-driven grid voltage support control of weakly connected microgrid inverters without adaptive adjustment coefficients is demonstrated. Figure 8 In the k-1 control cycle, v cα (k-1) represents the actual capacitor voltage in the α stationary coordinate system, and v cα_me (k-1) represents the capacitor voltage with noise in the α stationary coordinate system. According to the measured capacitor voltage v cα_me (k-1), predict the capacitor voltage of k control cycle to track the voltage reference value v ref cα The dotted line represents the prediction result in k-1 control cycle. The actual voltage in k-1 control cycle is shown as a solid line. Then, in k control cycle, v cα (k) and v cα_me (k). In the k control cycle, the voltage should be close to the voltage reference value v ref cα However, due to the limited range of duty cycle d, the adjustment range of capacitor voltage is limited. In the k control cycle, when d reaches 0, the change trend of capacitor voltage is represented by the dotted line. The actual voltage in the k control cycle is represented by the solid line. In the k+1 control cycle, v cα (k+1) and v cα_me (k+1). In the k+1 control cycle, when d reaches 1, the change trend of the capacitor voltage is also represented by the dotted line. In summary, noise will increase voltage fluctuations and affect the voltage performance of the data-driven network voltage support control.

[0165] Fig. 9 The cost function of data-driven predictive control with and without adaptive regulation coefficients is shown, considering different voltage noise intensities δ vWhen ρ = 1, noise suppression does not work, and when ρ < 1, noise suppression starts to work. When noise is suppressed and ρ = 1, as the voltage noise intensity δ v As the cost function value increases, it shows an upward trend. This shows that in the absence of an adaptive adjustment coefficient, the increase in voltage noise leads to an increase in the cost function value, reflecting the deterioration of the network voltage performance. When noise suppression starts to work, that is, ρ<1, the cost function value decreases significantly, indicating that the impact of noise on the network voltage performance is effectively suppressed.

[0166] like Fig.10 FIG. 1 is a flow chart of the entire control process based on the method of this embodiment. It mainly includes four parts: virtual voltage vector synthesis, data drive voltage prediction and noise suppression, cost function evaluation and PWM wave evaluation. Fig.10 As shown, first, the voltage and current data of the weakly connected microgrid inverter are measured and virtual vector synthesis is performed; then, the voltage and current data are used for data-driven prediction and noise suppression; finally, the optimal virtual voltage vector applied in the next control cycle is evaluated according to the cost function.

[0167] Embodiment 2:

[0168] This embodiment provides a data-driven weak-connected inverter network voltage support control system, including:

[0169] A mathematical model building module is used to establish a mathematical model of a weakly connected microgrid inverter in a two-phase stationary coordinate system for a weakly connected microgrid inverter topology, and to construct a discrete equation of the inverter in combination with the forward Euler method, and to design a grid capacitor voltage prediction equation for multi-vector application based on a model prediction method of a continuous control set; this module is used to implement the function of step S100 in the first embodiment, and will not be repeated here;

[0170] A data-driven model building module is used to establish a data-driven model based on the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system and the grid capacitor voltage prediction equation, and update the parameters of the data-driven model in combination with an adaptive tracking algorithm to predict the data-driven capacitor voltage; this module is used to implement the function of step S200 in the first embodiment, which will not be repeated here;

[0171] The noise elimination module is used to consider the influence of noise on the data-driven model when there is noise in the voltage and current data of the weakly connected microgrid inverter, introduce the influence of noise on the data-driven model in the process of predicting the data-driven capacitor voltage, and ensure the anti-noise ability and dynamic performance of the inverter network voltage support control; this module is used to implement the function of step S300 in Example 1, which will not be repeated here.

[0172] Embodiment three:

[0173] This embodiment proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the data-driven weak-connected inverter network voltage support control method as described in any embodiment of the present invention is implemented.

[0174] Embodiment 4:

[0175] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the data-driven weak-connected inverter network voltage support control method as described in any embodiment of the present invention is implemented.

[0176] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c can be single or multiple.

[0177] Those of ordinary skill in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented in a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0178] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0179] In several embodiments provided in the present application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory; hereinafter referred to as: ROM), random access memory (Random Access Memory; hereinafter referred to as: RAM), disk or optical disk, and other media that can store program codes.

[0180] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A data-driven weak-connected inverter network voltage support control method, characterized in that: The following steps are involved: Aiming at the topology of weakly connected microgrid inverter, a mathematical model of weakly connected microgrid inverter in two-phase stationary coordinate system is established. The discrete equation of the inverter is constructed by combining the forward Euler method. Based on the model prediction method of continuous control set, the grid capacitor voltage prediction equation is designed when multi-vector is applied. According to the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation, a data-driven model is established, and the parameters of the data-driven model are updated in combination with the adaptive tracking algorithm to predict the data-driven capacitor voltage. Considering the influence of noise on the data-driven model when there is noise in the voltage and current data of the weakly connected microgrid inverter, the influence of noise on the data-driven model is introduced in the process of predicting the data-driven capacitor voltage to ensure the anti-noise ability and dynamic performance of the inverter grid voltage support control.

2. A data-driven weak-connected inverter network voltage support control method according to claim 1, characterized in that: The model prediction method based on the continuous control set is designed to predict the network capacitor voltage when using multi-vectors: Among them, v mb cαβ_x (k+1) is the predicted value of the capacitor voltage at time k+1 in the two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ (k) is the capacitor voltage measurement value at time k in the two-phase stationary coordinate system; g mb αβ_x (k) is the capacitor voltage gradient at time k in the two-phase stationary coordinate system, which is equal to the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; Among them, g mb αβ_x (k) is expressed as: Among them, i mb iαβ_x (k+1) is the predicted value of the inverter side current at time k+1 in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ is the capacitor voltage on the inverter side in the two-phase stationary coordinate system; t is time; i iαβ (k) is the inverter side current at time k in the two-phase stationary coordinate system; C is the filter capacitor model parameter.

3. A data-driven weak-connected inverter network voltage support control method according to claim 2, characterized in that: The steps of establishing the data-driven model according to the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation are as follows: The capacitor voltage gradient at time k in the two-phase stationary coordinate system is transformed into: in, is the capacitor voltage gradient after transformation, i lαβ (k) is the load side current in the two-phase stationary coordinate system, L is the filter inductance model parameter, T s is the control period, v vsαβ is the virtual voltage vector of the inverter side in the two-phase stationary coordinate system; According to the capacitor voltage gradient at time k in the two-phase stationary coordinate system after the above transformation, the data-driven model is established as follows: in, is the capacitor voltage gradient based on data at time k in the two-phase stationary coordinate system, where: Among them, A(k) is the set of known and unknown variables, B(k) is the set of system disturbances, and I αβ (k) is the current set at time k in the two-phase stationary coordinate system; V αβ (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponding; The matrix form of the data-driven model is expressed as: in: Among them, y(k) is the input variable of the data-driven model, is the dynamic variable of the data-driven model, and θ(k) is the estimated variable of the data-driven model.

4. A data-driven weak-connected inverter network voltage support control method according to claim 3, characterized in that: The step of updating the parameters of the data-driven model in combination with the adaptive tracking algorithm and predicting the data-driven capacitor voltage includes: According to the adaptive tracking algorithm, the set of known and unknown variables A(k) and the system disturbance set B(k) in the data-driven model are estimated accordingly. The recursive matrix G(k) of the adaptive tracking algorithm is expressed as: in, represents the estimated value of θ(k); λ is the adaptive adjustment factor, which is expressed as: Among them, λ min is the minimum value of the adaptive adjustment factor, γ is a positive coefficient, and e0 is the error boundary value; according to Get the data driving voltage gradient g corresponding to the voltage vector on the inverter side in the two-phase stationary coordinate system mf αβ_x (k): Among them, V αβ_x (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; According to the data drive voltage gradient g mf αβ_x (k) and the virtual voltage vector, and obtain the capacitor voltage prediction value based on data driving.

5. A data-driven weak-connected inverter network voltage support control method according to claim 4, characterized in that: The method for introducing the influence of noise on the data driving model during the process of predicting the data driving capacitor voltage is specifically as follows: The data driving voltage gradient is transformed into a data driving voltage gradient considering noise suppression, specifically: in, In order to consider the data driving voltage gradient of noise suppression, ρ is an adaptive adjustment coefficient, which is adjusted according to the noise size, and 0<ρ<1.

6. A data-driven weak-connected inverter grid voltage support control system, characterized in that: include: The mathematical model building module is used to establish the mathematical model of the weakly connected microgrid inverter in a two-phase stationary coordinate system for the weakly connected microgrid inverter topology, and to construct the discrete equation of the inverter in combination with the forward Euler method. Based on the model prediction method of the continuous control set, the grid capacitor voltage prediction equation is designed when multi-vector applications are used; The data-driven model building module is used to establish a data-driven model based on the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation, and update the parameters of the data-driven model in combination with the adaptive tracking algorithm to predict the data-driven capacitor voltage; The noise elimination module is used to consider the influence of noise on the data-driven model when there is noise in the voltage and current data of the weakly connected microgrid inverter. The influence of noise on the data-driven model is introduced in the process of predicting the data-driven capacitor voltage to ensure the anti-noise ability and dynamic performance of the inverter grid voltage support control.

7. A data-driven weak-connected inverter network voltage support control system according to claim 6, characterized in that: The model prediction method based on the continuous control set is designed to predict the network capacitor voltage when using multi-vectors: Among them, v mb cαβ_x (k+1) is the predicted value of the capacitor voltage at time k+1 in the two-phase stationary coordinate system, which is consistent with the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ (k) is the capacitor voltage measurement value at time k in the two-phase stationary coordinate system; g mb αβ_x (k) is the capacitor voltage gradient at time k in the two-phase stationary coordinate system, which is equal to the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; Among them, g mb αβ_x (k) is expressed as: Among them, i mb iαβ_x (k+1) is the predicted value of the inverter side current at time k+1 in the two-phase stationary coordinate system, which is the same as the inverter side voltage vector v at time k in the two-phase stationary coordinate system. xαβ (k) corresponding; v cαβ is the capacitor voltage on the inverter side in the two-phase stationary coordinate system; t is time; i iαβ (k) is the inverter side current at time k in the two-phase stationary coordinate system; C is the filter capacitor model parameter.

8. A data-driven weak-connected inverter network voltage support control system according to claim 7, characterized in that: The steps of establishing the data-driven model according to the mathematical model of the weakly connected microgrid inverter in the two-phase stationary coordinate system and the grid capacitor voltage prediction equation are as follows: The capacitor voltage gradient at time k in the two-phase stationary coordinate system is transformed into: in, is the capacitor voltage gradient after transformation, i lαβ (k) is the load side current in the two-phase stationary coordinate system, L is the filter inductance model parameter, T s is the control period, v vsαβ is the virtual voltage vector of the inverter side in the two-phase stationary coordinate system; According to the capacitor voltage gradient at time k in the two-phase stationary coordinate system after the above transformation, the data-driven model is established as follows: in, is the capacitor voltage gradient based on data at time k in the two-phase stationary coordinate system, where: Among them, A(k) is the set of known and unknown variables, B(k) is the set of system disturbances, and I αβ (k) is the current set at time k in the two-phase stationary coordinate system; V αβ (k) is the voltage set at time k in the two-phase stationary coordinate system, which is the same as the virtual voltage vector v at the inverter side at time k in the αβ two-phase stationary coordinate system. vsαβ (k) corresponding; The matrix form of the data-driven model is expressed as: in: Among them, y(k) is the input variable of the data-driven model, is the dynamic variable of the data-driven model, and θ(k) is the estimated variable of the data-driven model.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements a data-driven weak-connected inverter network voltage support control method as described in any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements a data-driven weak-connected inverter network voltage support control method as described in any one of claims 1 to 5.

Images