A non-intrusive parameter identification method for single-phase PWM inverter circuit

Abstract

A single-phase PWM inverter circuit non-intrusive parameter identification method, in which a PWM inverter circuit model is established, a space state equation set of the PWM inverter circuit model is written, and a 4th order Runge-Kutta method is used to calculate the value of the space state equation set; an evaluation function is constructed, an optimization particle swarm optimization algorithm is used for optimization, the coordinate dimension of a particle is 4, the minimum value of the evaluation function is found, the particle coordinate corresponding to the minimum value of the evaluation function is found, and the coordinate is updated as the identified circuit parameter under the state. f rmse A single-phase PWM inverter circuit non-intrusive parameter identification method, in which a PWM inverter circuit model is established, a space state equation set of the PWM inverter circuit model is written, and a 4th order Runge-Kutta method is used to calculate the value of the space state equation set; an evaluation function is constructed, an optimization particle swarm optimization algorithm is used for optimization, the coordinate dimension of a particle is 4, the minimum value of the evaluation function is found, the particle coordinate corresponding to the minimum value of the evaluation function is found, and the coordinate is updated as the identified circuit parameter under the state.

Description

Technical Field

[0001] This invention relates to the field of single-phase PWM inverter circuit technology, and in particular to a non-intrusive parameter identification method for single-phase PWM inverter circuits. Background Technology

[0002] Inverters play a crucial role in modern power systems and power electronics. Studies have shown that power semiconductor devices and capacitors are the most vulnerable components in inverters. These components are subjected to complex and frequent electrothermal stresses during long-term operation, leading to reduced reliability and directly impacting the safe and stable operation of the system. Summary of the Invention

[0003] This invention provides a non-intrusive parameter identification method for single-phase PWM inverter circuits, offering a non-intrusive, low-cost, and accurate solution for predictive maintenance and life assessment of single-phase PWM inverter systems.

[0004] A non-intrusive parameter identification method for single-phase PWM inverter circuits, wherein,

[0005] The single-phase PWM inverter circuit is a single-phase two-level PWM inverter circuit. The circuit includes a DC power supply side, power switching devices T1 and T2, and an AC output side.

[0006] The DC power supply side includes capacitors C1 and C2 connected in series for voltage division.

[0007] Power switching device T1 and power switching device T2 are connected in parallel with capacitor C1 and capacitor C2 respectively, and power switching device T1 and power switching device T2 are connected in series.

[0008] The power switching device T1 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on1 ;

[0009] The power switching device T2 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on2 ;

[0010] A DC voltage source with voltage V and resistance R s The line resistance, after being connected in series, is also connected in parallel with capacitors C1 and C2;

[0011] The AC output side is drawn from the series connection of capacitors C1 and C2, and from the series connection of power switching device T1 and power switching device T2.

[0012] The AC output side uses an inductor L for filtering, and the load is a resistor R. LWherein, the inductance L and the resistance R L Forming a series relationship;

[0013] The method includes the following steps:

[0014] S1: Collect the voltage u of capacitor C1 respectively. C1 The voltage u of capacitor C2 C2 and the output voltage u on the load of the AC output side L Output current i L =u L / R L ,

[0015] Voltage u C1 Voltage u C2 Output current i L As an observation of the physical circuit to be identified, it is denoted as u. C1_real u C2_real i L_real, ,

[0016] The control signals are acquired by the controller. The switching signals of power switching device T1 and power switching device T2 are denoted as S1 and S2, respectively, where S... i =1 or 0, 1 represents on, 2 represents off, i=1,2, use signal S to represent the switch signal, let switch signal S1 be S, then switch signal S2=1-S, where S=0 or 1;

[0017] S2: Determine the spatial state equations of the single-phase PWM inverter circuit model;

[0018] S3: Calculate the values ​​of the space state equations using the fourth-order Runge-Kutta method;

[0019] S4: Construct an evaluation function by comparing the real system model and the digital twin model;

[0020] S5: Optimization is performed using a particle swarm optimization algorithm. The coordinate dimension of the particles is 4, and each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, R. on1 R on2 To find the evaluation function f rmse Find the minimum value of the evaluation function, locate the particle coordinates corresponding to the minimum value, and update these coordinates with the identified circuit parameters C1, C2, and R in this state. on1 R on2 This enables health monitoring of voltage-dividing capacitors and power switching devices;

[0021] in,

[0022] The two Rs at power switching device T1 on1 In the resistor, the first Ron1 A resistor is connected in series between the power switching device T1 and the load resistor R. L It is connected in series between power switching device T1 and power switching device T2; the second R on1 After the resistor is connected in series with the first diode, it is further connected in series with the first R. on1 The resistor and the power switching device T1 are connected in parallel;

[0023] The two Rs at power switching device T2 on2 In the resistor, the first R on2 A resistor is connected in series between the power switching device T2 and the capacitor C2; the second R on2 After the resistor is connected in series with the second diode, it is further connected in series with the first R. on2 The resistor and the power switching device T2 are connected in parallel.

[0024] Preferably, in step S2,

[0025] The spatial state equations of the single-phase PWM inverter circuit model are as follows:

[0026]

[0027] Right now:

[0028]

[0029] Where L is the inductance value, S is the switching signal, and R is the inductance value. L Let f1 be the load resistance value, and f3 be the functional relationship between the variables.

[0030] Preferably, in step S3,

[0031] The initial conditions and solution interval are determined based on the spatial state equations.

[0032] Divide the solution interval into n equal subintervals, with a time step of h for each subinterval. Within each subinterval, use the 4th-order Runge-Kutta iteration formula to calculate approximate solutions for several points in the subinterval:

[0033]

[0034] in, h is the interval step size, x i For the sampling time point,

[0035] The approximate solutions at each step size are weighted and averaged to obtain the approximate solution for the entire solution interval, which is then used as the value of the spatial state equation system.

[0036] Preferably, in step S4,

[0037] The digital twin model is the model solved using the Runge-Kutta method, while the real system model is a physical inverter model built using measurements and sampling to obtain the load current i. L_real and capacitor voltage u C1_real ,u C2_real The data is filtered using an adaptive filtering algorithm to remove noise, resulting in the filtered measurement value i. L_m , u C1_m , u C2_m The i-th circuit under a certain parameter state is calculated using a digital twin model. L ,u C1 ,u C2 Data, and construct an evaluation function f rmse .

[0038] Preferably, in step S5,

[0039] The optimization method using the improved particle swarm optimization algorithm includes:

[0040] Initialization: Randomly generate several groups of particles with a dimension of 4. Each particle contains four circuit parameters to be optimized, namely C1, C2, R... on1 R on2 ,

[0041] Calculate particle fitness: that is, calculate the evaluation function f using the parameters of the current particle. rmse ,

[0042] Update individual and global optima: Find the historical optimal parameters for each particle and the parameters of the current globally optimal particle.

[0043] Velocity and position updates: The particle velocity is updated based on the historical best position of the individual and the group, and the particle position (i.e., circuit parameters) is adjusted according to the updated velocity. The individual learning factor and the group learning factor change dynamically to avoid getting trapped in local optima and accelerate convergence.

[0044] Results Evaluation: For each particle representing the circuit parameters, calculate the evaluation function f between the model predictions and actual measurements. rmse value,

[0045] When the evaluation function is minimized, the corresponding C1, C2, and R are obtained. on1 R on2 Furthermore, it is believed that the closest parameters to the obtained circuit parameters are used to monitor the health status of the voltage divider capacitors and power switching devices.

[0046] Furthermore, this invention also discloses a non-intrusive parameter identification system for a single-phase PWM inverter circuit, wherein,

[0047] The single-phase PWM inverter circuit is a single-phase two-level PWM inverter circuit. The circuit includes a DC power supply side, power switching devices T1 and T2, and an AC output side.

[0048] The DC power supply side includes capacitors C1 and C2 connected in series for voltage division.

[0049] Power switching device T1 and power switching device T2 are connected in parallel with capacitor C1 and capacitor C2 respectively, and power switching device T1 and power switching device T2 are connected in series.

[0050] The power switching device T1 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on1 ;

[0051] The power switching device T2 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on2 ;

[0052] A DC voltage source with voltage V and resistance R s The line resistance, after being connected in series, is also connected in parallel with capacitors C1 and C2;

[0053] The AC output side is drawn from the series connection of capacitors C1 and C2, and from the series connection of power switching device T1 and power switching device T2.

[0054] The AC output side uses an inductor L for filtering, and the load is a resistor R. L Wherein, the inductance L and the resistance R L Forming a series relationship;

[0055] The system includes:

[0056] The acquisition unit is used to: acquire the voltage u of capacitor C1 respectively. C1 The voltage u of capacitor C2 C2 and the output voltage u on the load of the AC output side L Output current i L =u L / R L ,

[0057] Voltage u C1 Voltage u C2 Output current i L As an observation of the physical circuit to be identified, it is denoted as u. C1_real u C2_real i L_real, ,

[0058] The control signals are acquired by the controller. The switching signals of power switching device T1 and power switching device T2 are denoted as S1 and S2, respectively, where S... i =1 or 0, 1 represents on, 2 represents off, i=1,2, use signal S to represent the switch signal, let switch signal S1 be S, then switch signal S2=1-S, where S=0 or 1;

[0059] A determination unit is used to determine the spatial state equation set of the single-phase PWM inverter circuit model;

[0060] A computational unit used to calculate the values ​​of the space state equations using the fourth-order Runge-Kutta method;

[0061] Evaluation function construction unit, which is used to construct evaluation functions by comparing real system models and digital twin models;

[0062] The optimization unit is used to perform optimization using the particle swarm optimization algorithm. The coordinate dimension of the particles is 4, and each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, R. on1 R on2 To find the evaluation function f rmse Find the minimum value of the evaluation function, locate the particle coordinates corresponding to the minimum value, and update these coordinates with the identified circuit parameters C1, C2, and R in this state. on1 R on2 This enables health monitoring of voltage-dividing capacitors and power switching devices;

[0063] in,

[0064] The two Rs at power switching device T1 on1 In the resistor, the first R on1 A resistor is connected in series between the power switching device T1 and the load resistor R. L It is connected in series between power switching device T1 and power switching device T2; the second R on1 After the resistor is connected in series with the first diode, it is further connected in series with the first R. on1 The resistor and the power switching device T1 are connected in parallel;

[0065] The two Rs at power switching device T2 on2 In the resistor, the first R on2 A resistor is connected in series between the power switching device T2 and the capacitor C2; the second R on2 After the resistor is connected in series with the second diode, it is further connected in series with the first R. on2 The resistor and the power switching device T2 are connected in parallel.

[0066] Preferably, for a given unit, wherein,

[0067] The spatial state equations of the single-phase PWM inverter circuit model are as follows:

[0068]

[0069] Right now:

[0070]

[0071] Where L is the inductance value, S is the switching signal, and R is the inductance value. L Let f1 be the load resistance value, and f3 be the functional relationship between the variables.

[0072] Preferably, for a computing unit, wherein,

[0073] The initial conditions and solution interval are determined based on the spatial state equations.

[0074] Divide the solution interval into n equal subintervals, with a time step of h for each subinterval. Within each subinterval, use the 4th-order Runge-Kutta iteration formula to calculate approximate solutions for several points in the subinterval:

[0075]

[0076] in, h is the interval step size, x i For the sampling time point,

[0077] The approximate solutions at each step size are weighted and averaged to obtain the approximate solution for the entire solution interval, which is then used as the value of the spatial state equation system.

[0078] Preferably, for the evaluation function construction unit, where,

[0079] The digital twin model is the model solved using the Runge-Kutta method, while the real system model is a physical inverter model built using measurements and sampling to obtain the load current i. L_real and capacitor voltage u C1_real ,u C2_real The data is filtered using an adaptive filtering algorithm to remove noise, resulting in the filtered measurement value i. L_m , u C1_m , u C2_m The i-th circuit under a certain parameter state is calculated using a digital twin model. L ,u C1 ,u C2 Data, and construct an evaluation function f rmse .

[0080] Preferably, for the optimization unit, where,

[0081] The optimization method using the improved particle swarm optimization algorithm includes:

[0082] Initialization: Randomly generate several groups of particles with a dimension of 4. Each particle contains four circuit parameters to be optimized, namely C1, C2, R... on1 R on2 ,

[0083] Calculate particle fitness: that is, calculate the evaluation function f using the parameters of the current particle. rmse ,

[0084] Update individual and global optima: Find the historical optimal parameters for each particle and the parameters of the current globally optimal particle.

[0085] Velocity and position updates: The particle velocity is updated based on the historical best position of the individual and the group, and the particle position (i.e., circuit parameters) is adjusted according to the updated velocity. The individual learning factor and the group learning factor change dynamically to avoid getting trapped in local optima and accelerate convergence.

[0086] Results Evaluation: For each particle representing the circuit parameters, calculate the evaluation function f between the model predictions and actual measurements. rmse value,

[0087] When the evaluation function is minimized, the corresponding C1, C2, and R are obtained. on1 R on2 Furthermore, it is believed that the closest parameters to the obtained circuit parameters are used to monitor the health status of the voltage divider capacitors and power switching devices.

[0088] Furthermore, the present invention discloses a computer storage medium comprising computer instructions that, when executed on a computer, cause the computer to perform the method described in any of the preceding claims.

[0089] Furthermore, the present invention also discloses an electronic device, wherein the electronic device comprises:

[0090] Memory, processor, and computer programs stored in memory and executable on the processor, wherein,

[0091] When the processor executes the program, it implements the method as described in any of the preceding descriptions.

[0092] Compared with the prior art, the present invention has the following advantages:

[0093] Because the solution of this invention is achieved non-invasively, based on the collected electrical parameters and control signals and ultimately through signal processing, without incurring any expensive hardware costs, this invention is non-invasive and low-cost. Furthermore, this invention also features high identification accuracy. Attached Figure Description

[0094] Various other advantages and benefits of the present invention will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. It is obvious that the drawings described below are merely some embodiments of the invention, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0095] In the attached diagram:

[0096] Figure 1 This is a circuit diagram of a non-intrusive parameter identification method for a single-phase PWM inverter circuit according to one embodiment of the present invention;

[0097] Figure 2 This is a schematic diagram of the current reference positive direction of a circuit for a non-intrusive parameter identification method for a single-phase PWM inverter circuit according to an embodiment of the present invention.

[0098] Figure 3 This is a flowchart illustrating a non-intrusive parameter identification method for a single-phase PWM inverter circuit according to one embodiment of the present invention.

[0099] Figure 4 This is a schematic diagram illustrating the objective function iterative convergence speed of a non-intrusive parameter identification method for a single-phase PWM inverter circuit according to one embodiment of the present invention.

[0100] The present invention will be further explained below with reference to the accompanying drawings and embodiments. Detailed Implementation

[0101] Specific embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While specific embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0102] It should be noted that certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that different terms may be used to refer to the same component. This specification and claims do not distinguish components based on differences in terminology, but rather on differences in function. The terms "comprising" or "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising but not limited to." The following descriptions are preferred embodiments for carrying out the invention; however, these descriptions are for the purpose of understanding the general principles of the specification and are not intended to limit the scope of the invention. The scope of protection of this invention is determined by the appended claims.

[0103] To facilitate understanding of the embodiments of the present invention, further explanations and descriptions will be provided below with reference to the accompanying drawings and specific embodiments. The accompanying drawings do not constitute a limitation on the embodiments of the present invention.

[0104] like Figures 1 to 4 As shown, the non-intrusive parameter identification method for a single-phase PWM inverter circuit includes the following steps:

[0105] S1: Establish a PWM inverter circuit model, where the DC power supply is connected in series with resistor Rs, capacitor C1, and capacitor C2. Power switch T1 is connected in series with inductor L and load resistor RL, and the two capacitors on both sides of C1 are connected in series. Power switch T2 is connected in series with inductor L and load resistor RL, and the two capacitors on both sides of C2 are connected in series. The AC output terminal is formed by inductor L and resistor RL connected in series, and the output voltage is provided by load resistor RL. The switching signals of power switch T1 and power switch T2 are generated by PWM and are denoted as S1 and S2 respectively, represented by signal S. Their relationship is:

[0106] (1)

[0107] Where Si = 1 or 0 (1 represents on, 2 represents off, i = 1, 2);

[0108] S2: Write the space state equations for the PWM inverter circuit model, which are:

[0109] ;

[0110] S3: The fourth-order Runge-Kutta method for calculating the values ​​of the space state equations;

[0111] S4: Construct an evaluation function. A real system model and a digital twin model are built for comparison. The real system model is a physical inverter entity model. Load current iL_rs and capacitor voltage uc1_rs, uc2_rs data are obtained through measurement sampling. Noise is filtered out using an adaptive filtering algorithm to obtain the filtered measured values ​​iL_m, uc1_m, uc2_m. The iL, uC1, uC2 data under a certain circuit parameter state are calculated using the DT model. Based on the above data, the evaluation function frmse is constructed.

[0112] (2)

[0113] S5: Use the particle swarm optimization algorithm for optimization. The coordinate dimension of the particles is 4. Each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, Ron1, Ron2. Find the minimum value of the evaluation function frmse, find the particle coordinates corresponding to the minimum value of the evaluation function, and update this coordinate to the circuit parameters identified in this state.

[0114] In one embodiment, Figure 1 This is a single-phase two-level PWM inverter, where T1 and T2 are power switching devices. The DC power supply side is divided by two capacitors, C1 and C2, and Rs represents the line resistance. The AC output is formed by an inductor L and a resistor RL connected in series, with the output voltage provided by the load resistor RL. The switching device here is an IGBT. Because it is necessary to monitor the aging of the IGBT, considering that the IGBT's health state parameter Vce is approximately linearly related to the current, the ideal IGBT model is replaced with a model that includes the equivalent on-resistance Ron, as follows: Figure 2 As shown. In Figure 2 The reference positive direction of the current in each branch is specified.

[0115] The switching signals for power devices T1 and T2 are generated by PWM and denoted as S1 and S2, respectively. Represented by signal S, the relationship between them is:

[0116] (3)

[0117] Where Si = 1 or 0 (1 represents on, 2 represents off, i = 1, 2)

[0118] Based on the switch signal S, draw the circuit operation diagrams for the two states, such as... Figure 2 As shown. For example Figure 2 Given two circuit states, use capacitor voltages uc1 and uc2 and load current iL to write circuit equations. Unify the equations using the S signal and reorganize them into a system of differential equations:

[0119]

[0120] Right now:

[0121] 2. Calculate the values ​​of the differential equations. Since the sampled signal is discrete and the switching signal is not differentiable, a system of differential equations needs to be solved. The fourth-order Runge-Kutta method can be used to discretize and model the circuit. The Runge-Kutta method offers high accuracy and speed.

[0122] 3. Constructing the Evaluation Function. For the target circuit to be identified, we constructed a real system model and a digital twin model for comparison. The real system model is a physical inverter entity model. We obtained the load current iL_rs and capacitor voltage uc1_rs, uc2_rs data through measurement sampling, and used an adaptive filtering algorithm to filter out noise, obtaining the filtered measurement values ​​iL_m, uc1_m, uc2_m. We calculated the iL, uC1, uC2 data under a certain circuit parameter state using the DT model, and constructed the evaluation function frmse based on the above data:

[0123] (4)

[0124] 4. Optimization is performed using the Particle Swarm Optimization (PSO) algorithm. The particle coordinates are 4-dimensional, with each coordinate representing a monitored parameter in the circuit (C1, C2, Ron1, Ron2). Under the given algorithm strategy, the minimum value of the evaluation function frmse is found, and the particle coordinates corresponding to this minimum value are updated to reflect the identified circuit parameters in this state. The specific flowchart is shown below. Figure 3 As shown.

[0125] Furthermore, this invention also discloses a non-intrusive parameter identification system for a single-phase PWM inverter circuit, wherein,

[0126] The single-phase PWM inverter circuit is a single-phase two-level PWM inverter circuit. The circuit includes a DC power supply side, power switching devices T1 and T2, and an AC output side.

[0127] The DC power supply side includes capacitors C1 and C2 connected in series for voltage division.

[0128] Power switching device T1 and power switching device T2 are connected in parallel with capacitor C1 and capacitor C2 respectively, and power switching device T1 and power switching device T2 are connected in series.

[0129] The power switching device T1 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on1 ;

[0130] The power switching device T2 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on2 ;

[0131] A DC voltage source with voltage V and resistance R s The line resistance, after being connected in series, is also connected in parallel with capacitors C1 and C2;

[0132] The AC output side is drawn from the series connection of capacitors C1 and C2, and from the series connection of power switching device T1 and power switching device T2.

[0133] The AC output side uses an inductor L for filtering, and the load is a resistor R. L Wherein, the inductance L and the resistance R L Forming a series relationship;

[0134] The system includes:

[0135] The acquisition unit is used to: acquire the voltage u of capacitor C1 respectively. C1 The voltage u of capacitor C2 C2 and the output voltage u on the load of the AC output side L Output current i L =u L / R L ,

[0136] Voltage u C1 Voltage u C2 Output current i L As an observation of the physical circuit to be identified, it is denoted as u. C1_real u C2_real i L_real, ,

[0137] The control signals are acquired by the controller. The switching signals of power switching device T1 and power switching device T2 are denoted as S1 and S2, respectively, where S... i =1 or 0, 1 represents on, 2 represents off, i=1,2, use signal S to represent the switch signal, let switch signal S1 be S, then switch signal S2=1-S, where S=0 or 1;

[0138] A determination unit is used to determine the spatial state equation set of the single-phase PWM inverter circuit model;

[0139] A computational unit used to calculate the values ​​of the space state equations using the fourth-order Runge-Kutta method;

[0140] Evaluation function construction unit, which is used to construct evaluation functions by comparing real system models and digital twin models;

[0141] The optimization unit is used to perform optimization using the particle swarm optimization algorithm. The coordinate dimension of the particles is 4, and each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, R. on1 R on2 To find the evaluation function f rmse Find the minimum value of the evaluation function, locate the particle coordinates corresponding to the minimum value, and update these coordinates with the identified circuit parameters C1, C2, and R in this state. on1 R on2 This enables health monitoring of voltage-dividing capacitors and power switching devices;

[0142] in,

[0143] The two Rs at power switching device T1 on1 In the resistor, the first R on1 A resistor is connected in series between the power switching device T1 and the load resistor R. L It is connected in series between power switching device T1 and power switching device T2; the second R on1 After the resistor is connected in series with the first diode, it is further connected in series with the first R. on1 The resistor and the power switching device T1 are connected in parallel;

[0144] The two Rs at power switching device T2 on2 In the resistor, the first R on2 A resistor is connected in series between the power switching device T2 and the capacitor C2; the second R on2 After the resistor is connected in series with the second diode, it is further connected in series with the first R. on2 The resistor and the power switching device T2 are connected in parallel.

[0145] Preferably, for a given unit, wherein,

[0146] The spatial state equations of the single-phase PWM inverter circuit model are as follows:

[0147]

[0148] Right now:

[0149]

[0150] Where L is the inductance value, S is the switching signal, and R is the inductance value. L Let f1 be the load resistance value, and f3 be the functional relationship between the variables.

[0151] Preferably, for a computing unit, wherein,

[0152] The initial conditions and solution interval are determined based on the spatial state equations.

[0153] Divide the solution interval into n equal subintervals, with a time step of h for each subinterval. Within each subinterval, use the 4th-order Runge-Kutta iteration formula to calculate approximate solutions for several points in the subinterval:

[0154]

[0155] in, h is the interval step size, x i For the sampling time point,

[0156] The approximate solutions at each step size are weighted and averaged to obtain the approximate solution for the entire solution interval, which is then used as the value of the spatial state equation system.

[0157] Preferably, for the evaluation function construction unit, where,

[0158] The digital twin model is the model solved using the Runge-Kutta method, while the real system model is a physical inverter model built using measurements and sampling to obtain the load current i. L_real and capacitor voltage u C1_real ,u C2_real The data is filtered using an adaptive filtering algorithm to remove noise, resulting in the filtered measurement value i. L_m , u C1_m , u C2_m The i-th circuit under a certain parameter state is calculated using a digital twin model. L ,u C1 ,u C2 Data, and construct an evaluation function f rmse .

[0159] Preferably, for the optimization unit, where,

[0160] The optimization method using the improved particle swarm optimization algorithm includes:

[0161] Initialization: Randomly generate several groups of particles with a dimension of 4. Each particle contains four circuit parameters to be optimized, namely C1, C2, R... on1 R on2 ,

[0162] Calculate particle fitness: that is, calculate the evaluation function f using the parameters of the current particle. rmse ,

[0163] Update individual and global optima: Find the historical optimal parameters for each particle and the parameters of the current globally optimal particle.

[0164] Velocity and position updates: The particle velocity is updated based on the historical best position of the individual and the group, and the particle position (i.e., circuit parameters) is adjusted according to the updated velocity. The individual learning factor and the group learning factor change dynamically to avoid getting trapped in local optima and accelerate convergence.

[0165] Results Evaluation: For each particle representing the circuit parameters, calculate the evaluation function f between the model predictions and actual measurements. rmse value,

[0166] When the evaluation function is minimized, the corresponding C1, C2, and R are obtained. on1 R on2 Furthermore, it is believed that the closest parameters to the obtained circuit parameters are used to monitor the health status of the voltage divider capacitors and power switching devices.

[0167] Furthermore, the present invention discloses a computer storage medium comprising computer instructions that, when executed on a computer, cause the computer to perform the method described in any of the preceding claims.

[0168] Furthermore, the present invention also discloses an electronic device, wherein the electronic device comprises:

[0169] Memory, processor, and computer programs stored in memory and executable on the processor, wherein,

[0170] When the processor executes the program, it implements the method as described in any of the preceding descriptions.

[0171] In another, more specific embodiment,

[0172] A simulation platform for a two-level single-phase inverter was built. The inverter adopts a sinusoidal pulse width modulation control algorithm. The DC voltage V is 200V, the control switching frequency fsw is 4kHz, the carrier frequency f0 is 50Hz, the load inductance L is 5mH, the load resistance RL is 1Ω, the DC side capacitors C1 and C2 are both 76mF, and the line resistance Rs is approximately 0.1Ω.

[0173] Table 1 presents five sets of inverter parameters at different aging levels. The proposed method was used to identify these parameters under these conditions. These parameter combinations cover cases where the IGBT on-resistance increases by 0-5% and the capacitance decreases by 0-5%, as shown in Table 2. The results are shown in Table 2, and the convergence speed is as follows: Figure 4 As shown.

[0174] Table 1. Parameters for different aging degrees

[0175]

[0176] Table 2 Identification results of the present invention

[0177]

[0178] Although embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the specific embodiments and application fields described above. The specific embodiments described above are merely illustrative and instructive, and not restrictive. Those skilled in the art can make many other forms based on the guidance of this specification and without departing from the scope of protection of the claims of the present invention, and all of these are within the scope of protection of the present invention.

Claims

1. A non-intrusive parameter identification method for a single-phase PWM inverter circuit, characterized in that, The single-phase PWM inverter circuit is a single-phase two-level PWM inverter circuit. The circuit includes a DC power supply side, power switching devices T1 and T2, and an AC output side. The DC power supply side includes capacitors C1 and C2 connected in series for voltage division. Power switching device T1 and power switching device T2 are connected in parallel with capacitor C1 and capacitor C2 respectively, and power switching device T1 and power switching device T2 are connected in series. The power switching device T1 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on1 ; The power switching device T2 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on2 ; A DC voltage source with voltage V and resistance R s The line resistance, after being connected in series, is also connected in parallel with capacitors C1 and C2; The AC output side is drawn from the series connection of capacitors C1 and C2, and from the series connection of power switching device T1 and power switching device T2. The AC output side uses an inductor L for filtering, and the load is a resistor R. L Wherein, the inductance L and the resistance R L Forming a series relationship; The method includes the following steps: S1: Collect the voltage u of capacitor C1 respectively. C1 The voltage u of capacitor C2 C2 and the output voltage u on the load of the AC output side L Output current i L =u L / R L , Voltage u C1 Voltage u C2 Output current i L As an observation of the physical circuit to be identified, it is denoted as u. C1_real u C2_real i L_real , The control signals are acquired by the controller. The switching signals of power switching device T1 and power switching device T2 are denoted as S1 and S2, respectively, where S... i =1 or 0, 1 represents on, 0 represents off, i=1,2, use signal S to represent the switch signal, let switch signal S1 be S, then switch signal S2=1-S, where S=0 or 1; S2: Determine the spatial state equations of the single-phase PWM inverter circuit model; S3: Calculate the values ​​of the space state equations using the fourth-order Runge-Kutta method; S4: An evaluation function is constructed by comparing the real system model and the digital twin model. The digital twin model is the model solved using the Runge-Kutta method, and the real system model is the physical inverter entity model. The load current i is obtained through measurement sampling. L_real and capacitor voltage u C1_real ,u C2_real The data is filtered using an adaptive filtering algorithm to remove noise, resulting in the filtered measurement value i. L_m , u C1_m , u C2_m The i-th circuit under a certain parameter state is calculated using a digital twin model. L ,u C1 ,u C2 Data, and construct an evaluation function f rmse ; S5: Optimization is performed using a particle swarm optimization algorithm. The coordinate dimension of the particles is 4, and each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, R. on1 R on2 In order to find the evaluation function f rmse Find the minimum value of the evaluation function, locate the particle coordinates corresponding to the minimum value, and update these coordinates with the identified circuit parameters C1, C2, and R in this state. on1 R on2 This enables health monitoring of voltage-dividing capacitors and power switching devices; in, The two Rs at power switching device T1 on1 In the resistor, the first R on1 A resistor is connected in series between the power switching device T1 and the load resistor R. L It is connected in series between power switching device T1 and power switching device T2; the second R on1 After the resistor is connected in series with the first diode, it is further connected in series with the first R. on1 The resistor and the power switching device T1 are connected in parallel; The two Rs at power switching device T2 on2 In the resistor, the first R on2 A resistor is connected in series between the power switching device T2 and the capacitor C2; the second R on2 After the resistor is connected in series with the second diode, it is further connected in series with the first R. on2 The resistor and the power switching device T2 are connected in parallel.

2. The method according to claim 1, wherein, In step S2, The spatial state equations of the single-phase PWM inverter circuit model are as follows: ， Right now: ， Where L is the inductance value, S is the switching signal, and R... L f1 represents the load resistance value, and f3 represents the functional relationship between the variables.

3. The method according to claim 1, wherein, In step S3, The initial conditions and solution interval are determined based on the spatial state equations. Divide the solution interval into n equal subintervals, with a time step of h for each subinterval. Within each subinterval, use the 4th-order Runge-Kutta iteration formula to calculate approximate solutions for several points in the subinterval: ， in, h is the interval step size, x i For the sampling time point, The approximate solutions at each step size are weighted and averaged to obtain the approximate solution for the entire solution interval, which is then used as the value of the spatial state equation system.

4. A non-intrusive parameter identification system for a single-phase PWM inverter circuit, characterized in that, The single-phase PWM inverter circuit is a single-phase two-level PWM inverter circuit. The circuit includes a DC power supply side, power switching devices T1 and T2, and an AC output side. The DC power supply side includes capacitors C1 and C2 connected in series for voltage division. Power switching device T1 and power switching device T2 are connected in parallel with capacitor C1 and capacitor C2 respectively, and power switching device T1 and power switching device T2 are connected in series. The power switching device T1 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on1 ; The power switching device T2 has an equivalent on-state resistance R connected in series at both the forward conduction point and the reverse diode point. on2 ; A DC voltage source with voltage V and resistance R s The line resistance, after being connected in series, is also connected in parallel with capacitors C1 and C2; The AC output side is drawn from the series connection of capacitors C1 and C2, and from the series connection of power switching device T1 and power switching device T2. The AC output side uses an inductor L for filtering, and the load is a resistor R. L Wherein, the inductance L and the resistance R L Forming a series relationship; The system includes: The acquisition unit is used to: acquire the voltage u of capacitor C1 respectively. C1 The voltage u of capacitor C2 C2 and the output voltage u on the load of the AC output side L Output current i L =u L / R L , Voltage u C1 Voltage u C2 Output current i L As an observation of the physical circuit to be identified, it is denoted as u. C1_real u C2_real i L_real , The control signals are acquired by the controller. The switching signals of power switching device T1 and power switching device T2 are denoted as S1 and S2, respectively, where S... i =1 or 0, 1 represents on, 0 represents off, i=1,2, use signal S to represent the switch signal, let switch signal S1 be S, then switch signal S2=1-S, where S=0 or 1; A determination unit is used to determine the spatial state equation set of the single-phase PWM inverter circuit model; A computational unit used to calculate the values ​​of the space state equations using the fourth-order Runge-Kutta method; The evaluation function construction unit is used to construct evaluation functions by comparing a real system model and a digital twin model. The digital twin model is the model solved using the Runge-Kutta method, and the real system model is a physical inverter entity model built by measurement sampling to obtain the load current i. L_real and capacitor voltage u C1_real ,u C2_real The data is filtered using an adaptive filtering algorithm to remove noise, resulting in the filtered measurement value i. L_m , u C1_m , u C2_m The i-th circuit under a certain parameter state is calculated using a digital twin model. L ,u C1 ,u C2 Data, and construct an evaluation function f rmse ; The optimization unit is used to perform optimization using the particle swarm optimization algorithm. The coordinate dimension of the particles is 4, and each coordinate represents a parameter to be monitored in the circuit, namely C1, C2, R. on1 R on2 In order to find the evaluation function f rmse Find the minimum value of the evaluation function, locate the particle coordinates corresponding to the minimum value, and update these coordinates with the identified circuit parameters C1, C2, and R in this state. on1 R on2 This enables health monitoring of voltage-dividing capacitors and power switching devices; in, The two Rs at power switching device T1 on1 In the resistor, the first R on1 A resistor is connected in series between the power switching device T1 and the load resistor R. L It is connected in series between power switching device T1 and power switching device T2; the second R on1 After the resistor is connected in series with the first diode, it is further connected in series with the first R. on1 The resistor and the power switching device T1 are connected in parallel; The two Rs at power switching device T2 on2 In the resistor, the first R on2 A resistor is connected in series between the power switching device T2 and the capacitor C2; the second R on2 After the resistor is connected in series with the second diode, it is further connected in series with the first R. on2 The resistor and the power switching device T2 are connected in parallel.

5. The system according to claim 4, wherein, For a given unit, where, The spatial state equations of the single-phase PWM inverter circuit model are as follows: ， Right now: ， Where L is the inductance value, S is the switching signal, and R... L f1 represents the load resistance value, and f3 represents the functional relationship between the variables.

6. The system according to claim 4, wherein, For the computational unit, where, The initial conditions and solution interval are determined based on the spatial state equations. Divide the solution interval into n equal subintervals, with a time step of h for each subinterval. Within each subinterval, use the 4th-order Runge-Kutta iteration formula to calculate approximate solutions for several points in the subinterval: ， in, h is the interval step size, x i For the sampling time point, The approximate solutions at each step size are weighted and averaged to obtain the approximate solution for the entire solution interval, which is then used as the value of the spatial state equation system.

7. A computer storage medium, wherein, The storage medium includes computer instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 3.

8. An electronic device, wherein, The electronic device includes: Memory, processor, and computer programs stored in memory and executable on the processor, wherein, When the processor executes the program, it implements the method as described in any one of claims 1 to 3.

Images