Parameter monitoring method for grid-connected three-phase LCL inverter facing digital twin modeling

Abstract

The application discloses a grid-connected three-phase LCL inverter parameter monitoring method for digital twin modeling, comprising: obtaining actual measurement data of the grid-connected three-phase LCL inverter; based on the actual measurement data, iteratively optimizing a monitoring parameter set of a digital twin model of the grid-connected three-phase LCL inverter based on an adaptive particle swarm algorithm, comprising: running the digital twin model under each monitoring parameter set to obtain corresponding multiple groups of model output data, calculating corresponding error objective functions, and taking the monitoring parameter set corresponding to the minimum objective function value as the optimal monitoring parameter set; in response to the error objective function value of the optimal monitoring parameter set being less than a function threshold value, the optimal monitoring parameter set is output as the final optimal monitoring parameter set for subsequent parameter monitoring; otherwise, iteratively optimizing is continued. The high-precision calibration of the digital twin model and the real inverter is realized, so that the key parameters in the real inverter which are difficult to directly measure or have high measurement cost can be calculated in real time.

Description

Technical Field

[0001] This invention relates to a method for monitoring parameters of a grid-connected three-phase LCL inverter oriented towards digital twin modeling, and belongs to the field of inverter parameter monitoring technology. Background Technology

[0002] In recent years, the energy industry has been transforming towards a more cost-effective, low-carbon, and sustainable future. The power sector, driven by the rapid decline in the cost of renewable energy use, particularly wind and solar power, has become a leader in this transformation. New models are being developed through the large-scale integration of distributed energy resources with energy storage systems and electric vehicles. This scenario is made possible by the application of power electronic converters, which act as the interface between renewable energy systems and the grid, ensuring efficient power flow. Among these, three-phase inverters, due to their widespread use in low- and medium-voltage applications, have become a core module of distributed generation systems. The inverter's filtering stage is used to reduce harmonic content in voltage and current waveforms, with LCL filters widely adopted due to their excellent harmonic suppression at high frequencies.

[0003] However, due to the nonlinear operation of inverters, the high-frequency operation of switching devices, and the effects of thermoelectric stress, long-term operation may lead to performance degradation of components such as capacitors, inductors, and power switches. This parameter drift can cause deviations between the mechanistic analytical model based on the original design parameters and the actual system operation, resulting in the controller failing to meet design specifications and even causing system failures. Therefore, condition monitoring of key inverter components is crucial. Understanding component degradation helps improve the accuracy of analytical models, extend system life, and enhance reliability. Currently, various methods have been proposed for condition monitoring of key components. While algorithms such as recursive least squares and Kalman filters can be used for parameter estimation, they are difficult to estimate multiple parameters simultaneously, and some methods require additional sensors. Deep learning technology uses artificial neural networks to estimate component degradation, but it relies on a large amount of data to train the model, and in practice, it is often difficult to obtain sufficient data support. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a parameter monitoring method for grid-connected three-phase LCL inverters oriented towards digital twin modeling. Based on the adaptive particle swarm optimization algorithm, the monitoring parameter set of the digital twin model of the grid-connected three-phase LCL inverter is iteratively optimized, realizing high-precision calibration between the digital twin model and the real inverter. This allows for real-time calculation of key parameters in the real inverter that are difficult to measure directly or are too costly to measure, thus achieving parameter monitoring.

[0005] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0006] This invention discloses a method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling, comprising:

[0007] Acquire actual measurement data of the grid-connected three-phase LCL inverter; the actual measurement data includes operating status data and current measurement data;

[0008] Based on the actual measurement data, the monitoring parameter set of the pre-constructed digital twin model of the grid-connected three-phase LCL inverter is iteratively optimized using the adaptive particle swarm optimization algorithm, and the final optimal monitoring parameter set is output.

[0009] A digital twin model is constructed based on the final optimal monitoring parameter set to monitor the parameters of the grid-connected three-phase LCL inverter.

[0010] The adaptive particle swarm optimization algorithm includes the following steps:

[0011] Initialize multiple sets of monitoring parameters;

[0012] The operational status data is used as the input signal for the digital twin model. The digital twin model is run under each set of monitoring parameters to obtain multiple sets of corresponding model output data.

[0013] Calculate the objective function of the error between each set of model output data and the current measurement data, and take the monitoring parameter set corresponding to the minimum objective function value as the optimal monitoring parameter set;

[0014] If the error objective function value of the optimal monitoring parameter set is less than a preset function threshold, then the optimal monitoring parameter set is taken as the final optimal monitoring parameter set; otherwise, the weighting factor is adjusted and multiple sets of monitoring parameter sets are updated to continue iterative optimization.

[0015] Furthermore, the monitoring parameters in the monitoring parameter set include inverter-side filter inductor, grid-side filter inductor, filter capacitor, and line equivalent load.

[0016] Furthermore, the digital twin model includes a discretized mathematical equation model and a closed-loop virtual controller;

[0017] The discretized mathematical equation model is used to output the calculated values ​​of the inverter-side output current, the grid-side output current, and the filter capacitor voltage at the next moment, based on the grid voltage measurement signals at the current moment and the switching modulation signals output by the closed-loop virtual controller.

[0018] The closed-loop virtual controller is used to perform coordinate system transformation on the calculated value of the grid-side output current and the grid voltage measurement signal at the current moment to obtain the grid-side output current and grid voltage in the dq coordinate system;

[0019] Then, by using discrete equations and space vector pulse width modulation techniques, the switching modulation signal for the next moment is obtained.

[0020] Furthermore, the method for constructing the discretized mathematical equation model is as follows:

[0021] Based on Kirchhoff's laws, a set of state-space equations for a grid-connected three-phase LCL inverter is established.

[0022] The state-space equations are discretized using the bilinear transformation method to obtain a discretized mathematical equation model.

[0023] Furthermore, the expression for the discretized mathematical equation model is as follows:

[0024] ,

[0025] In the formula, This represents the calculated output current of phase a inverter at time n+1;

[0026] This represents the calculated output current of the b-phase inverter at time n+1.

[0027] This represents the calculated output current of the c-phase inverter at time n+1.

[0028] This represents the calculated value of the output current on the grid side of phase a at time n+1;

[0029] This represents the calculated value of the output current on the grid side of phase b at time n+1;

[0030] This represents the calculated value of the c-phase grid-side output current at time n+1;

[0031] This represents the calculated value of the phase a filter capacitor voltage at time n+1;

[0032] This represents the calculated value of the phase b filter capacitor voltage at time n+1;

[0033] This represents the calculated value of the c-phase filter capacitor voltage at time n+1;

[0034] This represents the calculated output current of phase a inverter at time n.

[0035] This represents the calculated output current of phase b inverter at time n.

[0036] This represents the calculated output current of the c-phase inverter at time n.

[0037] This represents the calculated value of the output current on the grid side of phase a at time n;

[0038] This represents the calculated value of the output current on the grid side of phase b at time n;

[0039] This represents the calculated value of the c-phase grid-side output current at time n;

[0040] This represents the calculated value of the phase a filter capacitor voltage at time n;

[0041] This represents the calculated value of the phase b filter capacitor voltage at time n;

[0042] This represents the calculated value of the c-phase filter capacitor voltage at time n;

[0043] This represents the phase a switching modulation signal at time n+1;

[0044] This represents the phase b switching modulation signal at time n+1;

[0045] This represents the c-phase switching modulation signal at time n+1;

[0046] This represents the measured grid voltage signal of phase a at time n+1;

[0047] This represents the measured voltage signal of phase b of the power grid at time n+1;

[0048] This represents the measured voltage signal of phase c of the power grid at time n+1;

[0049] This represents the phase a switching modulation signal at time n;

[0050] This represents the phase b switching modulation signal at time n;

[0051] This represents the c-phase switching modulation signal at time n;

[0052] This represents the measured voltage signal of phase a of the power grid at time n;

[0053] This represents the measured voltage signal of phase b of the power grid at time n;

[0054] This represents the measured voltage signal of phase c of the power grid at time n;

[0055] Represents the first coefficient matrix; This represents the second coefficient matrix.

[0056] Furthermore, the discrete equations of the closed-loop virtual controller are as follows:

[0057] ,

[0058] ,

[0059] ,

[0060] ,

[0061] The d-axis output signal represents the discrete equation of the closed-loop virtual controller in the (k+1)th control cycle.

[0062] This represents the q-axis output signal of the discrete equations of the closed-loop virtual controller in the (k+1)th control cycle;

[0063] This represents the proportional control parameters of the PI controller;

[0064] This represents the reference value of the d-axis current in the dq coordinate system.

[0065] This represents the reference value of the q-axis current in the dq coordinate system;

[0066] The d-axis component of the grid-side output current in the dq coordinate system at time n is represented.

[0067] The q-axis component of the grid-side output current in the dq coordinate system at time n is represented.

[0068] This represents the d-axis output signal of the integral control section in the (k+1)th control cycle;

[0069] This represents the q-axis output signal of the integral control section in the (k+1)th control cycle;

[0070] The angular frequency of the mains voltage;

[0071] This indicates the inverter-side inductance of a three-phase LCL inverter;

[0072] This represents the grid-side inductance of a three-phase LCL inverter;

[0073] Represents the d-axis component of the grid voltage in the dq coordinate system at time n;

[0074] Represents the q-axis component of the grid voltage in the dq coordinate system at time n;

[0075] This represents the d-axis output signal of the integral control section in the k-th control cycle;

[0076] This represents the q-axis output signal of the integral control section in the k-th control cycle;

[0077] This represents the integral control parameters of the PI controller;

[0078] This indicates the control cycle of the closed-loop virtual controller.

[0079] Furthermore, running the digital twin model under each set of monitoring parameters includes:

[0080] Obtain the true sampling period for current measurement data;

[0081] Based on the actual sampling period, configure the solution step size of the digital twin model to ensure that the actual sampling period is an integer multiple of the solution step size;

[0082] For any set of monitoring parameters, the working status data is used as the input signal of the digital twin model, the corresponding digital twin model is run, and the original model output data is collected with the solution step size.

[0083] The original model output data is preprocessed synchronously using filter middleware to obtain synchronized model output data.

[0084] Furthermore, the expression for the model's output data is as follows:

[0085] ,

[0086] In the formula, This represents the output data of the j-th synchronized model; This represents the original model output data; This indicates the actual sampling period of the current measurement data; Indicates the step size for solving; This represents the actual number of samples for the current measurement data.

[0087] Furthermore, the model output data includes the calculated value of the inverter-side output current and the calculated value of the grid-side output current, and the difference between the calculated value of the inverter-side output current and the calculated value of the grid-side output current is the calculated value of the filter capacitor output current.

[0088] The current measurement data includes the inverter-side output current measurement value and the grid-side output current measurement value. The difference between the inverter-side output current measurement value and the grid-side output current measurement value is the filter capacitor output current measurement value.

[0089] Furthermore, the expression for the error objective function is as follows:

[0090] ,

[0091] In the formula, Represent the objective function of the error;

[0092] This represents the calculated output current of the j-th phase a-filter capacitor in the digital twin model.

[0093] This represents the actual measured output current value of the j-th phase a filter capacitor;

[0094] This represents the calculated output current of the j-th phase b-filter capacitor in the digital twin model.

[0095] This represents the actual measured output current value of the j-th phase b filter capacitor;

[0096] This represents the calculated output current of the j-th c-phase filter capacitor from the digital twin model.

[0097] This represents the actual measured output current value of the j-th c-phase filter capacitor;

[0098] This represents the actual number of samples for the current measurement data.

[0099] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0100] The present invention provides a parameter monitoring method for grid-connected three-phase LCL inverters based on digital twin modeling. First, it iteratively optimizes the monitoring parameter set of the digital twin model of the grid-connected three-phase LCL inverter based on the adaptive particle swarm optimization algorithm to achieve high-precision calibration between the digital twin model and the real inverter. This allows for real-time calculation of key parameters in the real inverter that are difficult to measure directly or are too costly to measure, thus enabling parameter monitoring. Second, this method adopts a non-intrusive design, eliminating the need for complex calibration and additional circuitry. It can monitor the status of multiple key components of the grid-connected three-phase LCL inverter, reducing damage to the real system and lowering the cost of parameter monitoring. Attached Figure Description

[0101] Figure 1 This is a flowchart of a method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling, provided in an embodiment of the present invention.

[0102] Figure 2 This is a circuit topology diagram of a grid-connected three-phase LCL inverter provided in an embodiment of the present invention;

[0103] Figure 3 This is a structural block diagram of the closed-loop virtual controller provided in an embodiment of the present invention;

[0104] Figure 4 This is a digital twin model structure diagram of a grid-connected three-phase LCL inverter provided in an embodiment of the present invention;

[0105] Figure 5 This is a flowchart illustrating the structure of a grid-connected three-phase LCL inverter parameter monitoring method for digital twin modeling provided in this embodiment of the invention.

[0106] Figure 6 These are the key parameter monitoring results of the grid-connected three-phase LCL inverter provided in this embodiment of the invention;

[0107] Figure 7 This is a comparison diagram of the output characteristics of the digital twin model of the grid-connected three-phase LCL inverter and the real system provided in this embodiment of the invention;

[0108] Figure 8 This is the present invention. Figure 7 Enlarged view of the output characteristics comparison chart for response times between 0.388s and 0.398s. Detailed Implementation

[0109] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0110] like Figure 1As shown, this embodiment provides a method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling, including:

[0111] Obtain actual measurement data from the grid-connected three-phase LCL inverter; the actual measurement data includes operating status data and current measurement data.

[0112] Based on actual measurement data, the monitoring parameter set of the pre-constructed digital twin model of the grid-connected three-phase LCL inverter is iteratively optimized using the adaptive particle swarm optimization algorithm, and the final optimal monitoring parameter set is output.

[0113] A digital twin model is constructed based on the final optimal set of monitoring parameters to monitor the parameters of the grid-connected three-phase LCL inverter.

[0114] The adaptive particle swarm optimization algorithm includes the following steps:

[0115] Initialize multiple sets of monitoring parameters;

[0116] Using operational status data as input signals to the digital twin model, the digital twin model is run under each set of monitoring parameters to obtain multiple sets of corresponding model output data.

[0117] Calculate the objective function of the error between each set of model output data and the current measurement data, and take the monitoring parameter set corresponding to the minimum objective function value as the optimal monitoring parameter set;

[0118] If the error objective function value of the optimal monitoring parameter set is less than a preset function threshold, then the optimal monitoring parameter set is taken as the final optimal monitoring parameter set; otherwise, the weighting factor is adjusted and multiple sets of monitoring parameter sets are updated to continue iterative optimization.

[0119] The technical concept of this invention is as follows: First, based on the adaptive particle swarm optimization algorithm, the monitoring parameter set of the digital twin model of the grid-connected three-phase LCL inverter is iteratively optimized to achieve high-precision calibration between the digital twin model and the real inverter, thereby calculating in real time the key parameters in the real inverter that are difficult to measure directly or are too costly to measure, and realizing parameter monitoring; Second, this method adopts a non-intrusive design, without the need for complex calibration and additional circuits, and can monitor the status of multiple key components of the grid-connected three-phase LCL inverter, reducing damage to the real system and lowering the cost of parameter monitoring.

[0120] Step 1: Obtain the actual measurement data of the grid-connected three-phase LCL inverter.

[0121] Actual measurement data includes operating status data and current measurement data.

[0122] Operating status data includes voltage measurement signals, such as DC side input voltage and grid-connected voltage.

[0123] The current measurement data includes the inverter-side output current measurement value and the grid-side output current measurement value.

[0124] Step 2: Based on the actual measurement data, iteratively optimize the monitoring parameter set of the pre-constructed digital twin model of the grid-connected three-phase LCL inverter using the adaptive particle swarm optimization algorithm, and output the final optimal monitoring parameter set.

[0125] 2.1 Construction of the digital twin model of the grid-connected three-phase LCL inverter.

[0126] Digital twin models include discretized mathematical equation models and closed-loop virtual controllers;

[0127] The discretized mathematical equation model is used to output the calculated values ​​of the inverter-side output current, grid-side output current, and filter capacitor voltage at the next time step, based on the grid voltage measurement signals at the current time and the switching modulation signals output by the closed-loop virtual controller.

[0128] The block diagram of the closed-loop virtual controller is as follows: Figure 3 As shown, a coordinate system transformation is performed on the calculated grid-side output current and the measured grid voltage signal at the current moment to obtain the grid-side output current and grid voltage in the dq coordinate system; then, the switching modulation signal for the next moment is obtained through discrete equations and space vector pulse width modulation technology. It should be noted that the closed-loop virtual controller in this method is applicable not only to discrete mathematical models but also to physical converters. By combining the discrete model and the closed-loop virtual controller, a complete digital twin model of a grid-connected three-phase LCL inverter is constructed, capable of simulating the output characteristics of an actual physical system. Specifically, the structure of the digital twin model of the grid-connected three-phase LCL inverter is as follows: Figure 4 As shown.

[0129] 2.1.1 Discretized mathematical equation model.

[0130] Taking the grid-connected three-phase LCL inverter as the research object, its circuit topology is as follows: Figure 2 As shown. Based on Kirchhoff's laws, the state-space equation of a three-phase LCL inverter in grid-connected mode is:

[0131] ,

[0132] In the formula,

[0133] These are the inverter-side output currents of phases a, b, and c of the three-phase LCL inverter, respectively.

[0134] These are the grid-side output currents of phases a, b, and c of the three-phase LCL inverter, respectively.

[0135] These are the filter capacitor voltages for phases a, b, and c, respectively.

[0136] These are the derivatives of the output currents on the inverter side of the three-phase LCL inverter, representing phases a, b, and c, respectively.

[0137] These are the derivatives of the grid-side output currents of phases a, b, and c of a three-phase LCL inverter, respectively.

[0138] These are the derivatives of the filter capacitor voltages for phases a, b, and c, respectively.

[0139] This refers to the DC voltage on the DC side of the inverter.

[0140] and These are the inverter-side inductance and grid-side inductance of a three-phase LCL inverter, respectively.

[0141] For the filter capacitors of a three-phase LCL inverter;

[0142] and These are the equivalent load of the line and the passive damping resistance, respectively.

[0143] These are the grid voltages for phases a, b, and c on the grid side. These are the switching modulation signals for phases a, b, and c of the inverter.

[0144] Switching modulation signal This indicates the three-phase bridge arm switch of a three-phase LCL inverter. The conduction status is expressed mathematically as follows:

[0145] ,

[0146] ,

[0147] ,

[0148] express The upper switch transistor of the phase bridge arm is turned on and the lower switch transistor is turned off. express The upper switch on the phase bridge arm is turned off, and the lower switch is turned on. Similarly.

[0149] Based on the state-space equations of the grid-connected three-phase LCL inverter described above, the system matrix A and control matrix B can be obtained:

[0150] ,

[0151] ,

[0152] The state-space equations are discretized using the bilinear transformation method, resulting in a discretized mathematical equation model. The expression of the discretized mathematical equation model is as follows:

[0153] ,

[0154] In the formula, This represents the calculated output current of phase a inverter at time n+1;

[0155] This represents the calculated output current of the b-phase inverter at time n+1.

[0156] This represents the calculated output current of the c-phase inverter at time n+1.

[0157] This represents the calculated value of the output current on the grid side of phase a at time n+1;

[0158] This represents the calculated value of the output current on the grid side of phase b at time n+1;

[0159] This represents the calculated value of the c-phase grid-side output current at time n+1;

[0160] This represents the calculated value of the phase a filter capacitor voltage at time n+1;

[0161] This represents the calculated value of the phase b filter capacitor voltage at time n+1;

[0162] This represents the calculated value of the c-phase filter capacitor voltage at time n+1;

[0163] This represents the calculated output current of phase a inverter at time n.

[0164] This represents the calculated output current of phase b inverter at time n.

[0165] This represents the calculated output current of the c-phase inverter at time n.

[0166] This represents the calculated value of the output current on the grid side of phase a at time n;

[0167] This represents the calculated value of the output current on the grid side of phase b at time n;

[0168] This represents the calculated value of the c-phase grid-side output current at time n;

[0169] This represents the calculated value of the phase a filter capacitor voltage at time n;

[0170] This represents the calculated value of the phase b filter capacitor voltage at time n;

[0171] This represents the calculated value of the c-phase filter capacitor voltage at time n;

[0172] This represents the phase a switching modulation signal at time n+1;

[0173] This represents the phase b switching modulation signal at time n+1;

[0174] This represents the c-phase switching modulation signal at time n+1;

[0175] This represents the measured grid voltage signal of phase a at time n+1;

[0176] This represents the measured voltage signal of phase b of the power grid at time n+1;

[0177] This represents the measured voltage signal of phase c of the power grid at time n+1;

[0178] This represents the phase a switching modulation signal at time n;

[0179] This represents the phase b switching modulation signal at time n;

[0180] This represents the c-phase switching modulation signal at time n;

[0181] This represents the measured voltage signal of phase a of the power grid at time n;

[0182] This represents the measured voltage signal of phase b of the power grid at time n;

[0183] This represents the measured voltage signal of phase c of the power grid at time n;

[0184] Represents the first coefficient matrix; This represents the second coefficient matrix.

[0185] Among them, the first coefficient matrix Second coefficient matrix It can be calculated using the formula of the bilinear transformation method:

[0186] ,

[0187] ,

[0188] In the formula, A represents the system matrix; B represents the control matrix; Indicates the solution step size of the digital twin model; Represents the identity matrix;

[0189] 2.1.2 Closed-loop virtual controller.

[0190] Based on the discrete equation model of the inverter described above, the calculated value of the grid-side output current at time n can be obtained. Then, it is transformed to the dq synchronous rotating coordinate system to obtain the grid-side output current in the dq coordinate system at time n. Similarly, the grid voltage measurement signal at time n... We also transform to the dq synchronous rotating coordinate system to obtain the grid voltage in the dq coordinate system at time n. Furthermore, phase-locked loop (PLL) technology is used to obtain the phase and angular frequency of the grid voltage. .

[0191] A passive damping control method is adopted, which involves connecting a damping resistor in series in the capacitor branch of the filter. This effectively suppresses the self-resonance of the LCL filter. Based on this, the grid-connected current can be directly controlled by feedback PI in the dq synchronous rotating coordinate system, and by combining voltage feedforward decoupling and cross-coupling compensation techniques, independent decoupling control of the d-axis and q-axis can be achieved.

[0192] Therefore, the discrete equations of the closed-loop virtual controller are as follows:

[0193] ,

[0194] ,

[0195] ,

[0196] ,

[0197] The d-axis output signal represents the discrete equation of the closed-loop virtual controller in the (k+1)th control cycle.

[0198] This represents the q-axis output signal of the discrete equations of the closed-loop virtual controller in the (k+1)th control cycle;

[0199] This represents the proportional control parameters of the PI controller;

[0200] This represents the reference value of the d-axis current in the dq coordinate system.

[0201] This represents the reference value of the q-axis current in the dq coordinate system;

[0202] The d-axis component of the grid-side output current in the dq coordinate system at time n is represented.

[0203] The q-axis component of the grid-side output current in the dq coordinate system at time n is represented.

[0204] This represents the d-axis output signal of the integral control section in the (k+1)th control cycle;

[0205] This represents the q-axis output signal of the integral control section in the (k+1)th control cycle;

[0206] The angular frequency of the mains voltage;

[0207] This indicates the inverter-side inductance of a three-phase LCL inverter;

[0208] This represents the grid-side inductance of a three-phase LCL inverter;

[0209] Represents the d-axis component of the grid voltage in the dq coordinate system at time n;

[0210] Represents the q-axis component of the grid voltage in the dq coordinate system at time n;

[0211] This represents the d-axis output signal of the integral control section in the k-th control cycle;

[0212] This represents the q-axis output signal of the integral control section in the k-th control cycle;

[0213] This represents the integral control parameters of the PI controller;

[0214] This indicates the control cycle of the closed-loop virtual controller.

[0215] Ultimately, the output signal of the closed-loop virtual controller is modulated into a modulation signal, namely a switching modulation signal, through space vector pulse width modulation. The value of is used to control the conduction state of the switching transistors of the three-phase bridge arm of the inverter, thereby achieving precise control of the three-phase inverter.

[0216] 2.2 The adaptive particle swarm optimization algorithm includes the following steps:

[0217] 2.2.1 Initialize multiple sets of monitoring parameters.

[0218] The monitoring parameters in the monitoring parameter set include the inverter-side filter inductor, grid-side filter inductor, filter capacitor, and line equivalent load. The expression for each set of monitoring parameters is as follows:

[0219] ,

[0220] In the formula, The set of monitoring parameters representing the digital twin model;

[0221] Indicates the inverter-side filter inductance;

[0222] Indicates the grid-side filter inductance;

[0223] Indicates the filter capacitor;

[0224] This indicates the equivalent load of the line.

[0225] 2.2.2 Using the working status data as the input signal of the digital twin model, the digital twin model is run under each set of monitoring parameters to obtain the corresponding multiple sets of model output data.

[0226] Run the digital twin model under each set of monitoring parameters, including:

[0227] Obtain the true sampling period for current measurement data;

[0228] Configure the solution step size of the digital twin model according to the actual sampling period, so that the actual sampling period is an integer multiple of the solution step size;

[0229] For any set of monitoring parameters, the working status data is used as the input signal of the digital twin model, and the corresponding digital twin model is run to solve for the step size to collect the output data of the original model.

[0230] The original model output data is preprocessed synchronously using filter middleware to obtain synchronized model output data.

[0231] Specifically, a true three-phase LCL inverter uses the true sampling period collect The three-phase current measurement data is denoted as Digital twin models are used to solve for step size. collect The calculated data for three-phase current is denoted as .

[0232] Set the solution step size for the digital twin model. The value of makes the actual sampling period It is an integer multiple of that. A filter middleware is introduced to process the solution data of the digital twin model. Perform synchronization preprocessing to obtain the synchronized model output data, as shown in the following expression:

[0233] ,

[0234] In the formula, This represents the output data of the j-th synchronized model; This represents the original model output data; This indicates the actual sampling period of the current measurement data; Indicates the step size for solving; This represents the actual number of samples for the current measurement data.

[0235] 2.2.3 Calculate the objective function of the error between the output data of each model and the current measurement data, and take the monitoring parameter set corresponding to the minimum objective function value as the optimal monitoring parameter set.

[0236] The model output data includes the calculated output current on the inverter side and the calculated output current on the grid side. The difference between the calculated output current on the inverter side and the calculated output current on the grid side is the calculated output current of the filter capacitor.

[0237] The current measurement data includes the inverter-side output current measurement value and the grid-side output current measurement value. The difference between the inverter-side output current measurement value and the grid-side output current measurement value is the filter capacitor output current measurement value.

[0238] Specifically, obtain the current measurement data of the actual inverter system, including the inverter-side output current measurement value and the grid-side output current measurement value, denoted as follows: and By keeping the digital twin model and the real system in the same operating state, the calculated values ​​of the inverter-side output current and the grid-side output current of the digital twin model are solved, and denoted as follows: and Based on the above data, an error objective function is constructed, and its expression is as follows:

[0239] ,

[0240] In the formula, Represent the objective function of the error;

[0241] This represents the calculated output current of the j-th phase a-filter capacitor in the digital twin model.

[0242] This represents the actual measured output current value of the j-th phase a filter capacitor;

[0243] This represents the calculated output current of the j-th phase b-filter capacitor in the digital twin model.

[0244] This represents the actual measured output current value of the j-th phase b filter capacitor;

[0245] This represents the calculated output current of the j-th c-phase filter capacitor from the digital twin model.

[0246] This represents the actual measured output current value of the j-th c-phase filter capacitor;

[0247] This represents the actual number of samples for the current measurement data.

[0248] 2.2.4 If the error objective function value of the optimal monitoring parameter set is less than the preset function threshold, then the optimal monitoring parameter set is taken as the final optimal monitoring parameter set; otherwise, the weighting factor is adjusted and multiple sets of monitoring parameter sets are updated to continue iterative optimization.

[0249] Using the adaptive particle swarm optimization algorithm as a bridge, the digital twin is connected to the real three-phase LCL inverter. Based on the constructed objective function and the monitoring parameter sequence, the key parameters inside the digital twin are iteratively optimized.

[0250] When the optimal objective function value is less than a set threshold or the number of iterations reaches a preset maximum value, the iterative optimization result becomes the monitoring result of each key parameter in the real system, thereby realizing the degradation monitoring function of key components and enabling the digital twin to have the same input and output characteristics as the real three-phase LCL inverter. Specifically, the parameter monitoring method for grid-connected three-phase LCL inverters oriented towards digital twin modeling has the following structure and flow: Figure 5 As shown.

[0251] Step 3: Based on the digital twin model constructed from the final optimal monitoring parameter set, perform parameter monitoring of the grid-connected three-phase LCL inverter.

[0252] In this embodiment, an experimental verification platform was built, and the digital twin model of a grid-connected three-phase LCL inverter was programmed and implemented in MATLAB. Operating status data of a real grid-connected three-phase LCL inverter was collected as the external input signal for the digital twin model. Both the digital twin model and the real inverter were independently configured with closed-loop controllers of the same structure. Simultaneously, current measurement data from the real inverter was collected as input data for the adaptive particle swarm optimization algorithm.

[0253] The effectiveness of the proposed parameter monitoring method was verified by measuring the output data of a real grid-connected three-phase LCL inverter under steady-state conditions, thereby monitoring the key component parameters of the inverter. The average of 12 monitoring results was taken as the final monitored value and compared with the actual parameter values. The results are as follows. Figure 6 As shown in Table 1, the monitoring results, average values, actual values, and monitoring errors for all parameters are summarized. The results demonstrate that the proposed parameter monitoring strategy can effectively monitor the status of key components such as filter inductors, filter capacitors, and resistors, verifying the practicality and accuracy of the method.

[0254] To verify whether the digital twin model can accurately reproduce the equivalent response of the real inverter, under the same steady-state operating conditions, the key component parameters of the digital twin model were set to the monitoring results, and the output characteristics of the digital twin model and the real system were compared. The results are as follows: Figure 7 and Figure 8 As shown in the figure, the output characteristics of the digital twin model are basically consistent with those of the real system, verifying the effectiveness of the digital twin model.

[0255] Table 1. Monitoring Results of Parameters of Grid-Connected Three-Phase LCL Inverters

[0256] Serial Number L_1(mH) C(μF) L_2(mH) R(Ω) 1 5.494 2.530 0.887 0.565 2 5.229 2.570 0.920 0.529 3 5.299 2.406 0.890 0.536 4 5.191 2.571 0.860 0.528 5 5.354 2.395 0.911 0.503 6 5.210 2.418 0.947 0.497 7 5.298 2.299 0.949 0.559 8 5.413 2.319 0.868 0.647 9 5.297 2.279 0.925 0.533 10 5.324 2.536 0.921 0.505 11 5.266 2.310 0.854 0.648 12 5.173 2.479 0.929 0.522 average value 5.296 2.426 0.905 0.548 actual value 5.000 2.500 1.000 0.500 error 5.92% 2.96% 9.50% 9.60%

[0257] In summary, this invention uses the bilinear transformation method to establish a discrete mathematical equation model of a grid-connected three-phase LCL inverter, and on this basis, constructs a closed-loop virtual controller to form a detailed digital twin model of the grid-connected three-phase LCL inverter, thereby simulating the input and output characteristics of the real inverter.

[0258] This invention designs a set of monitoring parameters and corresponding error objective functions, and uses an adaptive particle swarm optimization algorithm to monitor key parameters of a grid-connected three-phase LCL inverter. The monitoring error of key parameters is less than 10%, and the monitoring results of all parameters can be obtained simultaneously, providing an efficient means for fault monitoring of real systems and accurate simulation of physical system characteristics by digital twin models.

[0259] This invention employs a non-invasive design, eliminating the need for complex calibration and additional circuitry, to monitor the status of multiple key components in a grid-connected three-phase LCL inverter. This reduces damage to the actual system and lowers parameter monitoring costs. Furthermore, this invention is not only applicable to parameter monitoring of grid-connected three-phase LCL inverters but can also be extended to other types of power electronic converters to achieve status monitoring of key components.

[0260] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0261] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0262] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0263] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0264] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A parameter monitoring method for grid-connected three-phase LCL inverters based on digital twin modeling, characterized in that, include: Obtain actual measurement data of the grid-connected three-phase LCL inverter; The actual measurement data includes operating status data and current measurement data; Based on the actual measurement data, the monitoring parameter set of the pre-constructed digital twin model of the grid-connected three-phase LCL inverter is iteratively optimized using the adaptive particle swarm optimization algorithm, and the final optimal monitoring parameter set is output. A digital twin model is constructed based on the final optimal monitoring parameter set to monitor the parameters of the grid-connected three-phase LCL inverter. The adaptive particle swarm optimization algorithm includes the following steps: Initialize multiple sets of monitoring parameters; The working status data is used as the input signal of the digital twin model. The digital twin model is run under each set of monitoring parameters to obtain multiple sets of corresponding model output data. Calculate the objective function of the error between each set of model output data and the current measurement data, and take the monitoring parameter set corresponding to the minimum objective function value as the optimal monitoring parameter set; If the error objective function value of the optimal monitoring parameter set is less than a preset function threshold, then the optimal monitoring parameter set is taken as the final optimal monitoring parameter set; otherwise, the weighting factor is adjusted and multiple sets of monitoring parameter sets are updated to continue iterative optimization. The digital twin model includes a discretized mathematical equation model and a closed-loop virtual controller; The discretized mathematical equation model is used to output the calculated values ​​of the inverter-side output current, the grid-side output current, and the filter capacitor voltage at the next moment, based on the grid voltage measurement signals at the current moment and the switching modulation signals output by the closed-loop virtual controller. The closed-loop virtual controller is used to perform coordinate system transformation on the current grid-side output current calculation value and grid voltage measurement signal to obtain the grid-side output current and grid voltage in the dq coordinate system; and then obtain the switching modulation signal for the next moment through discrete equations and space vector pulse width modulation technology. Running the digital twin model under each set of monitoring parameters includes: Obtain the true sampling period for current measurement data; Based on the actual sampling period, configure the solution step size of the digital twin model to ensure that the actual sampling period is an integer multiple of the solution step size; For any set of monitoring parameters, the working status data is used as the input signal of the digital twin model, the corresponding digital twin model is run, and the original model output data is collected with the solution step size. The original model output data is preprocessed synchronously using filter middleware to obtain synchronized model output data.

2. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The monitoring parameters in the monitoring parameter set include inverter-side filter inductor, grid-side filter inductor, filter capacitor, and line equivalent load.

3. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The method for constructing the discretized mathematical equation model is as follows: Based on Kirchhoff's laws, a set of state-space equations for a grid-connected three-phase LCL inverter is established. The state-space equations are discretized using the bilinear transformation method to obtain a discretized mathematical equation model.

4. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The expression for the discretized mathematical equation model is as follows: ； In the formula, This represents the calculated output current of phase a inverter at time n+1; This represents the calculated output current of the b-phase inverter at time n+1. This represents the calculated output current of the c-phase inverter at time n+1. This represents the calculated value of the output current on the grid side of phase a at time n+1; This represents the calculated value of the output current on the grid side of phase b at time n+1; This represents the calculated value of the c-phase grid-side output current at time n+1; This represents the calculated value of the phase a filter capacitor voltage at time n+1; This represents the calculated value of the phase b filter capacitor voltage at time n+1; This represents the calculated value of the c-phase filter capacitor voltage at time n+1; This represents the calculated output current of phase a inverter at time n. This represents the calculated output current of phase b inverter at time n. This represents the calculated output current of the c-phase inverter at time n. This represents the calculated value of the output current on the grid side of phase a at time n; This represents the calculated value of the output current on the grid side of phase b at time n; This represents the calculated value of the c-phase grid-side output current at time n; This represents the calculated value of the phase a filter capacitor voltage at time n; This represents the calculated value of the phase b filter capacitor voltage at time n; This represents the calculated value of the c-phase filter capacitor voltage at time n; This represents the phase a switching modulation signal at time n+1; This represents the phase b switching modulation signal at time n+1; This represents the c-phase switching modulation signal at time n+1; This represents the measured grid voltage signal of phase a at time n+1; This represents the measured voltage signal of phase b of the power grid at time n+1; This represents the measured voltage signal of phase c of the power grid at time n+1; This represents the phase a switching modulation signal at time n; This represents the phase b switching modulation signal at time n; This represents the c-phase switching modulation signal at time n; This represents the measured voltage signal of phase a of the power grid at time n; This represents the measured voltage signal of phase b of the power grid at time n; This represents the measured voltage signal of phase c of the power grid at time n; Represents the first coefficient matrix; This represents the second coefficient matrix.

5. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The discrete equations of the closed-loop virtual controller are as follows: ； ； ； ； This represents the d-axis output signal of the discrete equations of the closed-loop virtual controller in the (k+1)th control cycle; This represents the q-axis output signal of the discrete equations of the closed-loop virtual controller in the (k+1)th control cycle; This represents the proportional control parameters of the PI controller. This represents the reference value of the d-axis current in the dq coordinate system. This represents the reference value of the q-axis current in the dq coordinate system; The d-axis component of the grid-side output current in the dq coordinate system at time n is represented. The q-axis component of the grid-side output current in the dq coordinate system at time n is represented. This represents the d-axis output signal of the integral control section in the (k+1)th control cycle; This represents the q-axis output signal of the integral control section in the (k+1)th control cycle; The angular frequency of the mains voltage; This indicates the inverter-side inductance of a three-phase LCL inverter; This represents the grid-side inductance of a three-phase LCL inverter; Represents the d-axis component of the grid voltage in the dq coordinate system at time n; Represents the q-axis component of the grid voltage in the dq coordinate system at time n; This represents the d-axis output signal of the integral control section in the k-th control cycle; This represents the q-axis output signal of the integral control section in the k-th control cycle; This represents the integral control parameters of the PI controller; This indicates the control cycle of the closed-loop virtual controller.

6. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The expression for the model's output data is as follows: ； In the formula, This represents the output data of the j-th synchronized model; This represents the original model output data; This indicates the actual sampling period of the current measurement data; Indicates the step size for solving; This represents the actual number of samples for the current measurement data.

7. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 1, characterized in that, The model output data includes the calculated value of the inverter-side output current and the calculated value of the grid-side output current. The difference between the calculated value of the inverter-side output current and the calculated value of the grid-side output current is the calculated value of the filter capacitor output current. The current measurement data includes the inverter-side output current measurement value and the grid-side output current measurement value. The difference between the inverter-side output current measurement value and the grid-side output current measurement value is the filter capacitor output current measurement value.

8. The method for monitoring parameters of a grid-connected three-phase LCL inverter based on digital twin modeling as described in claim 7, characterized in that, The expression for the objective function of the error is as follows: ； In the formula, Represent the objective function of the error; This represents the calculated output current of the j-th phase a-filter capacitor in the digital twin model. This represents the actual measured output current value of the j-th phase a filter capacitor; This represents the calculated output current of the j-th phase b-filter capacitor in the digital twin model. This represents the actual measured output current value of the j-th phase b filter capacitor; This represents the calculated output current of the j-th c-phase filter capacitor from the digital twin model. This represents the actual measured output current value of the j-th c-phase filter capacitor; This represents the actual number of samples for the current measurement data.

Images