Three-phase three-level ANPC inverter parameter monitoring method and platform based on digital twin
By establishing a mathematical model and digital twin of a three-phase three-level ANPC inverter based on digital twins, designing a multi-parameter monitoring sequence and optimization algorithm, solving the problems of high state monitoring costs and equipment damage in the existing technology, and achieving high-precision and low-cost parameter monitoring and key component health management.
Patent Information
- Application Number
- CN202410933788.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2025-02-28
- Estimated Expiration
- 2044-07-12
AI Technical Summary
The existing three-level ANPC inverter status monitoring strategy is mainly achieved by adding hardware and using intelligent algorithms, which leads to equipment damage, high monitoring costs and high data processing burden, and fails to effectively solve the problem of failure caused by component aging in the inverter.
The three-phase and three-level ANPC inverter parameter monitoring method based on digital twin is adopted, and the internal parameters of the digital twin are updated by establishing an inverter mathematical model, discrete processing, building a digital twin, designing a multi-parameter monitoring sequence and optimizing algorithms, so as to simulate the inverter input and output characteristics and monitor the parameters of key components.
The non-invasive parameter monitoring of three-level ANPC inverters is realized, which reduces the risk of equipment damage and monitoring costs, improves monitoring accuracy and reliability, and can effectively manage the aging of key components of the inverter and extends the equipment life.
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of DC-AC converters in the field of power electronics and power transmission, and in particular to a three-phase three-level ANPC inverter parameter monitoring method and platform based on digital twins. Background Art
[0002] Compared with the traditional two-level inverter, the voltage that the power semiconductor devices in the three-level inverter bear is half of the DC voltage source, which can improve the inverter output power level and the service life of the power semiconductor devices. In addition, the three-level inverter also has a lower voltage instantaneous change rate, lower output voltage, and current total harmonic distortion (THD). Therefore, the three-level inverter has been widely studied and applied in the fields of high-speed trains, photovoltaic grid connection, wind power generation, electric vehicles, etc. The three-level neutral point clamped (NPC) inverter and the three-level active neutral point clamped (ANPC) inverter are two typical three-level inverter topologies, and their research and application are relatively large. However, the single zero-level state of the three-level NPC inverter causes some power semiconductor devices to withstand higher junction temperatures due to frequent operation, affecting the output power of the inverter and the service life of the power semiconductor devices. The three-level ANPC inverter uses active power semiconductor devices to replace the clamping diodes in the NPC inverter, increases the number of zero-level states to four, and combines special strategies to achieve loss balancing, solving the problem of uneven losses between devices and helping to improve the service life of power semiconductors. Therefore, the three-level ANPC inverter has become a new research hotspot.
[0003] The high reliability of ANPC inverters is an important topic in the field of power electronics, and is also an important guarantee for ensuring the stable operation of the conversion system and adapting to complex working conditions. The investigation found that the DC side support capacitor and the main circuit power semiconductor devices are the most important and most vulnerable components in the inverter. When the inverter is subjected to greater stress, it may cause sudden failures of key components due to overstress; in addition, long-term and frequent operation of the inverter may cause the aging of key components and cause system crashes. Therefore, it is necessary to monitor the key components of the inverter to ensure the reliability of the system. At present, a variety of efficient fault diagnosis strategies have been proposed, which can effectively solve sudden failures, and the research is relatively complete; and in response to failures caused by component aging, condition monitoring is considered to be one of the important means. By monitoring the health status of key components, health assessments are performed to timely update aging equipment, avoid failures caused by aging, and save maintenance costs. The current condition monitoring strategy is mainly achieved by adding hardware and using intelligent algorithms, which inevitably causes damage to the equipment and increases monitoring costs and data processing burdens.
[0004] In the context of "Industry 4.0" with digitalization, networking and intelligence as the core, digital twins have attracted widespread attention due to their deep integration of the concept of new generation information technology and digital models. They provide an implementation path for the information-physical integration of complex dynamic systems throughout their life cycle, and are expected to bring the innovative design, manufacturing efficiency and operation and maintenance level of complex products to a new level. At present, digital twin technology has been successfully applied to the fields of DC-DC, AC-DC and two-level DC-AC, realizing parameter monitoring of key components and further evaluating the health status of key equipment. However, the more power devices in the three-level inverter make the inverter failure probability higher, so it is necessary to adopt a new strategy to realize non-invasive parameter monitoring of the three-level inverter. At present, no relevant three-level inverter parameter monitoring method based on digital twins has been reported. Summary of the invention
[0005] In order to solve the problems existing in the above-mentioned prior art, the present invention provides a three-phase three-level ANPC inverter parameter monitoring method and platform based on digital twins, which solves the technical problem that the existing ANPC inverter status monitoring strategy is mainly implemented by adding hardware and using intelligent algorithms, which inevitably causes damage to the equipment and increases the monitoring cost and data processing burden.
[0006] A three-phase three-level ANPC inverter parameter monitoring method based on digital twins includes: establishing an inverter mathematical model and performing discrete processing, the inverter mathematical model includes an inverter system mathematical model and an inverter main circuit power semiconductor device mathematical model, and the inverter main circuit power semiconductor device mathematical model uses the on-state voltage drop as a monitoring indicator to characterize the health status of the power semiconductor device; building an inverter digital twin based on the inverter mathematical model after discrete processing; designing an inverter multi-parameter monitoring sequence; using an optimization algorithm based on actual parameters of the inverter to update the internal key parameters of the inverter digital twin to obtain an accurate inverter digital twin; using the accurate inverter digital twin to realize input and output characteristic simulation and key component parameter monitoring of the physical inverter based on the inverter multi-parameter monitoring sequence, wherein the inverter is a three-phase three-level ANPC inverter.
[0007] Preferably, the inverter system mathematical model is as follows:
[0008]
[0009] Among them, u an 、u bn 、u cn are the three-phase output voltages of the three-level ANPC inverter, i a 、i b 、i c is the three-phase output phase current, p represents the differential coefficient, L a, L b , L c With R a , R b , R c They are the three-phase load inductance and resistance, U c1 , U c2 represents the voltage across the capacitor collected by the voltage sensor, C1 and C2 represent the capacitance of the two supporting capacitors on the DC side, and R ESR1 , R ESR2 Represents the parasitic resistance of the capacitor, u c1 、u c2 Represents the voltage supported by the capacitor.
[0010] Preferably, the indicators characterizing the health status of power semiconductor devices include: electrical parameters, thermal parameters, and sound, light and other parameters. Among them, the on-state voltage drop is sensitive to the degradation of power semiconductor devices and is easy to monitor, so the on-state voltage drop is selected as the monitoring indicator of the health status of power semiconductor devices. The mathematical model of the inverter main circuit designed based on the power semiconductor device model is as follows:
[0011]
[0012] Wherein, u represents the modulation voltage signal, u>0, represents the power semiconductor device of the upper half bridge arm of the inverter is working, and similarly, u<0, represents the power semiconductor device of the lower half bridge arm of the inverter is working. ao 、u on represents the terminal voltage, where u ao Represents the voltage between the inverter midpoint 'o' and the A-phase bridge arm output point 'a'. on Represents the voltage between the midpoint 'o' of the inverter and the midpoint 'n' of the inverter load. p is the switching state of the power semiconductor device (P state corresponds to A p The value is 1, and the O state corresponds to A p The value is 0, and the N state corresponds to A p value is -1), A absp A p The absolute value of i d Represents the direction of current flow (current flowing to the load is 1, otherwise it is -1), V CE 、V F Represents the conduction voltage drop of the power IGBT and the anti-parallel diode, V CE 、V F It can be equivalent to a voltage source series resistor. The mathematical model of the power semiconductor device in the inverter main circuit is expressed as:
[0013] V CE =V o +R o i c (5)
[0014] V F =V Do +R Do i c (6)
[0015] Among them, V o , R o is the on-state voltage and on-state resistance of the power IGBT, i c is the current flowing through the chip, V Do , R Do are the on-state voltage and on-state resistance of the anti-parallel diode.
[0016] After obtaining the mathematical model of the ANPC inverter, the fourth-order Runge-Kutta method is used to discretize the mathematical model to solve the model. In addition, the SVPWM strategy is combined with the PWMI wave mode to control the on and off of the inverter power semiconductor devices. However, the unbalanced midpoint potential is an inherent problem of the inverter. In order to avoid the situation where the unbalanced midpoint potential causes the inverter output waveform to be distorted or even causes the inverter to fail to work properly, the charge conservation method is used to balance the midpoint potential.
[0017] Preferably, the inverter digital twin constructed based on the inverter mathematical model after discrete processing includes: writing the inverter mathematical model after discrete processing into MATLAB to obtain an inverter digital twin that can be independently operated offline, and the control part is mainly completed in a digital microcontroller; the input of the digital twin includes two supporting capacitor voltages, three-phase output currents, line voltages and three-phase modulation signals generated in the digital microcontroller obtained by the sampling circuit, and the capacitor voltage and three-phase output current at the next moment are output after being solved by the digital twin, simulating the input and output characteristics of the actual physical inverter system. The input of the control part is the two supporting capacitor voltages and three-phase output currents on the DC side obtained by sampling the existing sensors of the system, and the output is the drive signal of the inverter power semiconductor device.
[0018] Preferably, a multi-parameter monitoring sequence for a three-level ANPC inverter is designed to reduce the number of parameters monitored simultaneously with the power semiconductor devices. The load inductance and resistance degrade relatively slowly, so they can be monitored first based on the output characteristics of the inverter. The DC side support capacitor and its parasitic resistance are related to the input voltage of the DC side voltage source, so parameter monitoring is performed based on the mathematical model of the DC side of the inverter. More power semiconductor devices inside the inverter are a difficult point in parameter monitoring. Based on the working principle of the inverter, the power semiconductor devices of the main circuit are monitored first, and then the power semiconductor devices used for clamping are monitored. It is worth noting that the designed parameter monitoring sequence does not require additional capacitor construction subsystems, and does not require complex data processing.
[0019] Preferably, the improved particle swarm algorithm is used as a bridge between the digital twin and the physical ANPC inverter system. Based on the parameter monitoring sequence and the dedicated target evaluation function, the key parameters inside the twin (for example: the capacitance and parasitic resistance of the capacitor, the inductance value, the on-state voltage drop of each power semiconductor device, etc.) are iterated and optimized to make the input and output characteristics of the constructed twin consistent with the input and output characteristics of the physical ANPC inverter used, and obtain the optimized inverter digital twin model. Among them, the target evaluation function used by the optimization algorithm to update the parameters is:
[0020]
[0021] Among them, f obj,1 For parameter monitoring of capacitance and parasitic resistance, f obj,2 Used for parameter monitoring of load inductance and resistance,
[0022] f obj,3 Used for parameter monitoring of power semiconductor devices, N is the amount of sampled data, U is c1,j ,U c2,j is the voltage sampling value of the two supporting capacitors on the DC side of the inverter, U cm1,j ,U cm2,j The voltage values of the two supporting capacitors calculated for the digital twin model, i a,j 、i b,j 、i c,j is the inverter three-phase output current sampling value, i am,j 、i bm,j 、i cm,j The three-phase output current values calculated for the digital twin model. ab,j is the inverter output line voltage sampling value, u abm,j Output line voltage values calculated for the digital twin model.
[0023] A three-phase three-level ANPC inverter parameter monitoring platform based on digital twins includes a data acquisition module, a parameter monitoring module and a data display module. The data acquisition module is used to obtain experimental data of the three-phase three-level ANPC inverter in steady state. The parameter monitoring module includes an accurate three-phase three-level ANPC inverter digital twin obtained in a three-phase three-level ANPC inverter parameter monitoring method based on digital twins, and is used to simulate actual input and output characteristics and monitor parameters of key inverter components according to the experimental data of the three-phase three-level ANPC inverter in steady state; the data display module is used to average n monitoring values obtained by the parameter monitoring module as the final monitoring result and display them.
[0024] The beneficial effects of the present invention include:
[0025] 1) Establish a detailed three-level ANPC inverter digital twin model to simulate the input and output characteristics of the physical three-level ANPC inverter.
[0026] 2) A multi-parameter monitoring sequence and corresponding target evaluation function are designed without adding any additional hardware equipment or complex data processing. Accurate monitoring of multiple parameters of the three-level ANPC inverter (DC side capacitance and parasitic resistance, forward voltage drop of power IGBT and anti-parallel diode, load inductance) is achieved, and the parameter monitoring error is less than 5%.
[0027] 3) The present invention realizes parameter monitoring of each power device in the main circuit, which can be further used for health management of key components such as inverter capacitors and power devices. And this strategy is also applicable under multiple working conditions such as different DC voltages, different loads, different collector currents, and different parameter initial values. In addition, non-invasive parameter monitoring is realized, which reduces damage to devices and reduces monitoring and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 This is the main circuit topology diagram of the three-phase three-level ANPC inverter.
[0029] Figure 2 It is a multi-parameter monitoring process based on improved PSO.
[0030] Figure 3 The experimental waveforms of the midpoint potential and three-phase output current of the three-phase three-level ANPC inverter.
[0031] Figure 4 Comparison between the digital twin model of the three-phase three-level ANPC inverter and the inverter midpoint potential waveform.
[0032] Figure 5 Comparison between the digital twin model of the three-phase three-level ANPC inverter and the inverter output A-phase current waveform.
[0033] Figure 6 Comparison between the three-phase three-level ANPC inverter digital twin model and the inverter output B-phase current waveform.
[0034] Figure 7 These are the monitoring results of DC side capacitance and parasitic resistance, load inductance and resistance parameters.
[0035] Figure 8 The monitoring results of the on-voltage parameters of power switches S1, S2, and S5.
[0036] Fig. 9 The monitoring results of the on-resistance parameters of power switches S1, S2, and S5.
[0037] Fig.10It is the monitoring result of the on-state voltage and on-state resistance parameters of the anti-parallel diode D5.
[0038] Fig.11 It is the monitoring result of the conduction voltage drop parameters of power switch tubes S1 and S2.
[0039] Fig.12 It is the monitoring result of the conduction voltage drop parameter of the power switch tube S5 and the anti-parallel diode D5.
[0040] Fig.13 The figure is a flow chart of the parameter monitoring method of three-phase three-level ANPC inverter based on digital twin. DETAILED DESCRIPTION
[0041] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.
[0042] The following is combined with Figure 1-13 The specific embodiments of the present invention are described in detail;
[0043] A three-phase three-level ANPC inverter parameter monitoring method based on digital twin, such as Fig.13 As shown, the following steps are included: including: establishing a three-phase three-level ANPC inverter mathematical model and performing discrete processing, the three-phase three-level ANPC inverter mathematical model includes an inverter system mathematical model and an inverter main circuit power semiconductor device mathematical model, and the inverter main circuit power semiconductor device mathematical model uses the on-state voltage drop as a monitoring indicator to characterize the health status of the power semiconductor device; building a three-phase three-level ANPC inverter digital twin based on the three-phase three-level ANPC inverter mathematical model after discrete processing; designing a three-phase three-level ANPC inverter multi-parameter monitoring sequence; based on the actual parameters of the three-phase three-level ANPC inverter, using an optimization algorithm to update the internal key parameters of the three-phase three-level ANPC inverter digital twin to obtain an accurate three-phase three-level ANPC inverter digital twin model; using the accurate three-phase three-level ANPC inverter digital twin to realize input and output characteristic simulation and multi-parameter monitoring of the physical three-phase three-level ANPC inverter.
[0044] In another embodiment, the specific process of establishing the three-phase three-level ANPC inverter mathematical model is as follows:
[0045] The working principle, different switching states and switching sequences of the three-phase three-level ANPC inverter are analyzed as shown in Table 1. In the ANPC inverter system, there are two typical commutation modes, namely PWMⅠ mode and PWMⅡ mode. In PWMⅠ mode, the loss imbalance between devices is lower than that in PWMⅡ mode, so PWMⅠ mode is used to control the switching state of the inverter power semiconductor devices. In addition, the on-state voltage drop is considered to be the most effective health status monitoring parameter for power I GBTs and anti-parallel diodes. The three-phase three-level ANPC inverter in this embodiment adopts PWMⅠ wave generation mode to generate drive signals for power semiconductor devices, that is,
[0046] Table 1 Switching states and switching sequences of ANPC inverter
[0047]
[0048] First, based on Kirchhoff's voltage law, the mathematical model of the three-phase output voltage of the three-phase three-level ANPC inverter is:
[0049]
[0050] Among them, u an 、u bn 、u cn are the three-phase voltages, i a 、i b 、i c is the three-phase output current, L a , L b , L c With R a , R b , R c They are the three-phase load inductance and resistance respectively.
[0051] The mathematical model of the DC side input voltage of the three-phase three-level ANPC inverter is:
[0052]
[0053] Among them, U c1 , U c2 is the voltage across the DC side support capacitor collected by the sensor, C1 and C2 are the capacitance values of the capacitors, and u c1 、u c2 is the voltage on the capacitor, R ESR1 , R ESR2 is the parasitic resistance of the capacitor.
[0054] A mathematical model containing the key parameters of each power device of the three-phase three-level ANPC inverter is established. The indicators that characterize the health status of power semiconductor devices include: electrical parameters, thermal parameters, and acoustic and optical parameters. Among them, the on-state voltage drop is sensitive to the degradation of power semiconductor devices and is easy to monitor. Therefore, the on-state voltage drop is selected as the monitoring indicator to characterize the health status of power semiconductor devices. an Taking the inverter as an example, a mathematical model including various power semiconductor devices in the inverter main circuit and considering different commutation modes is constructed:
[0055]
[0056] Among them, u represents the modulation voltage signal, u>0, represents the power semiconductor device of the upper half bridge arm of the inverter is working, and similarly, u<0, represents the power semiconductor device of the lower half bridge arm of the inverter is working. In addition, the zero-crossing point of u=0 has been established in the model. ao 、u on represents the terminal voltage, where u ao Represents the voltage between the inverter midpoint 'o' and the A-phase bridge arm output point 'a'. on Represents the voltage between the midpoint 'o' of the inverter and the midpoint 'n' of the inverter load. p is the switching state of the power semiconductor device (P state corresponds to A p The value is 1, and the O state corresponds to A p The value is 0, and the N state corresponds to A p value is -1), A absp A p The absolute value of i d Represents the direction of current flow (current flowing to the load is 1, otherwise it is -1), V CE 、V F Represents the conduction voltage drop of the power IGBT and the anti-parallel diode, V CE 、V F It can be equivalent to a voltage source in series with a resistor. The mathematical model of the power semiconductor device is expressed as:
[0057] V CE =V o +R o i c (5)
[0058] V F =V Do +R Do i c (6)
[0059] Among them, V o , R o is the on-state voltage and on-state resistance of the power IGBT, i cis the current flowing through the chip, V Do , R Do are the on-state voltage and on-state resistance of the anti-parallel diode.
[0060] The fourth-order Runge-Kutta method is used to discretize the mathematical model of the ANPC inverter. At this time, the discretized mathematical model of the ANPC inverter can be described as:
[0061]
[0062] In another embodiment, the construction of a three-phase three-level ANPC inverter digital twin based on the three-phase three-level ANPC inverter mathematical model after discrete processing includes: writing the three-level ANPC inverter mathematical model after discrete processing into MATLAB to obtain a three-phase three-level ANPC inverter digital twin that can be independently operated offline, and the control part is mainly completed in a digital microcontroller. The input of the control part is the two supporting capacitor voltages and three-phase output currents on the DC side obtained by sampling the existing sensors of the system, and the output is a driving signal of the inverter power semiconductor device; the input of the digital twin includes the two supporting capacitor voltages, three-phase output currents, and line voltages obtained by the sampling circuit, and the three-phase modulation signal generated in the digital microcontroller. After being solved by the digital twin, the capacitor voltage and three-phase output current at the next moment are output to simulate the input and output characteristics of the actual physical ANPC inverter system.
[0063] In another embodiment, considering that there are many key components of the three-level ANPC inverter, all parameters are identified simultaneously based on the particle swarm optimization algorithm (PSO), which will fall into the local optimum and make it difficult to find the most effective value. To avoid this situation, a parameter monitoring sequence is designed based on the aging characteristics of the components and the working principle of the inverter to reduce the number of parameters monitored simultaneously and improve the accuracy of parameter monitoring. The specific implementation is as follows: First, the load inductance and resistance degrade slowly and have little impact on other parameters, so they are monitored first, and the monitoring results are used in other parameter monitoring processes. Secondly, the voltage of the DC side support capacitor of the inverter is related to the input of the DC voltage source, and accurate parameter monitoring can be achieved based on the DC side mathematical model. Finally, the main circuit has more power semiconductor devices. According to the long conduction time of the power semiconductor devices in the main circuit in the PWMⅠ mode, the conduction time of the power semiconductor devices used for clamping is short, which has little impact on the parameter monitoring of the main circuit. Therefore, the parameters of the power semiconductor devices of the main circuit are monitored first, and then the parameters of the power semiconductor devices used for clamping are monitored. In addition, based on the characteristics of the PSO algorithm that the larger the inertia factor, the stronger the global search ability, and the smaller the inertia factor, the stronger the local search ability, the method of linearly decreasing the inertia factor is adopted to improve the PSO algorithm, ensuring that the particle finds the global optimal value in the early stage of optimization and approaches the global optimal value at a lower speed in the later stage. This method avoids the probability of falling into the local optimal value as much as possible.
[0064] The multi-parameter monitoring sequence and corresponding target evaluation function of the ANPC inverter are designed to monitor the load side inductance and resistance first, without adding additional hardware devices to build subsystems and filters to process data. Secondly, the DC side capacitance and parasitic resistance are only based on the DC side digital twin model to realize parameter monitoring. The power semiconductor devices in the main circuit have a long conduction time, and the power semiconductor devices used for clamping have little effect on the parameter monitoring of the main circuit. Therefore, the power semiconductor devices in the main circuit are monitored first, and the power semiconductor devices in the clamping circuit are monitored last. In addition, the parameter monitoring uses an improved particle swarm optimization algorithm as a bridge between the digital twin model and the physical inverter to update and iterate the key parameters of the digital twin model (for example: DC side support capacitance and parasitic resistance, each power IGBT and anti-parallel diode conduction voltage drop of the main circuit, load inductance, etc.), and optimize the model to reduce the difference between the twin model and the physical inverter, so that the constructed digital twin model can simulate the actual input and output characteristics of the inverter, and finally obtain an accurate digital twin model of the three-phase three-level ANPC inverter. Among them, the objective function for updating parameters is:
[0065]
[0066] Among them, N is the amount of sampled data, U c1,j ,U c2,j is the voltage sampling value of the two supporting capacitors on the DC side of the inverter, U cm1,j ,U cm2,j The voltage values of the two supporting capacitors calculated for the digital twin model, i x,j is the inverter three-phase output current sampling value, i xm,j The three-phase output current values calculated for the digital twin model. ab,j is the inverter output line voltage sampling value, u abm,j The output line voltage value calculated by the digital twin model. Formula (9) is used for parameter monitoring of capacitors and parasitic resistors, formula (10) is used for parameter monitoring of load inductance and resistance, and formula (11) is used for parameter monitoring of power semiconductor devices.
[0067] The above three-level ANPC inverter digital twin model is programmed and written into MATLAB. The input and output data of the inverter during operation (the voltage of the two supporting capacitors on the DC side, the three-phase output current of the inverter, and the line voltage) are collected based on the existing sensors of the system, and the three-phase modulation wave of the modulation link in the digital microprocessor is used as the input signal of the digital twin model. After the offline digital twin model is solved, the DC side capacitor voltage and three-phase output current at the next moment are output to simulate the input and output characteristics of the actual physical inverter system. Figure 3-6 As shown, Figure 4 Comparison between the digital twin model of the three-phase three-level ANPC inverter and the inverter midpoint potential waveform; Figure 5 Comparison between the three-phase three-level ANPC inverter digital twin model and the inverter output A-phase current waveform; Figure 6 Comparison between the three-phase three-level ANPC inverter digital twin model and the inverter output B-phase current waveform.
[0068] When the parameter iteration meets the set number of iterations, the monitoring results of each parameter are output. At this point, the input and output characteristics of the twin model are basically consistent with those of the physical inverter, and the internal parameters of the twin model are basically similar to the parameters of the key components of the inverter system. The digital twin model of the three-level ANPC can be obtained. The specific multi-parameter monitoring process of the three-level ANPC inverter based on digital twin is as follows: Figure 2 shown.
[0069] In order to verify the effectiveness of the parameter monitoring of the established digital twin model, the experimental data of the three-level ANPC inverter in steady state are obtained, the parameters of the key components of the inverter are monitored, and the average of 10 monitoring values is obtained as the final monitoring result. The parameter monitoring value is compared with the actual measurement value, such as Figure 7-12 As shown, Figure 7 It is the monitoring result of DC side capacitance and parasitic resistance, load inductance and resistance parameters; Figure 8 The monitoring results of the on-state voltage parameters of the power switch tubes S1, S2, and S5; Fig. 9 The monitoring results of the on-resistance parameters of power switch tubes S1, S2, and S5; Fig.10 It is the monitoring result of the on-state voltage and on-state resistance parameters of the anti-parallel diode D5; Fig.11 The monitoring results of the conduction voltage drop parameters of the power switch tubes S1 and S2; Fig.12 The monitoring results of the conduction voltage drop parameters of the power switch tube S5 and the anti-parallel diode D5 are summarized in Table 2 and Table 3. The strategy realizes the monitoring of various power devices and parameters such as capacitance, parasitic resistance, and inductance.
[0070] Table 2 Monitoring results of the on-state voltage drop of power devices in three-level ANPC inverter
[0071]
[0072]
[0073] Table 3 Multi-parameter monitoring results of key components of three-level ANPC inverter
[0074]
[0075] In another embodiment, a three-phase three-level ANPC inverter parameter monitoring platform based on digital twin is provided, including a data acquisition module, a parameter monitoring module and a data display module, wherein the data acquisition module is used to obtain experimental data of the three-phase three-level ANPC inverter in steady state, and the parameter monitoring module includes an accurate digital twin model of the three-phase three-level ANPC inverter obtained in the three-phase three-level ANPC inverter multi-parameter monitoring method based on digital twin, which is used to simulate actual input and output characteristics and monitor parameters of key components of the inverter according to the experimental data of the three-phase three-level ANPC inverter in steady state; the data display module is used to average n monitoring values obtained by the parameter monitoring module as the final monitoring result and display them.
[0076] The above-mentioned embodiments only express the specific implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the protection scope of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the technical solution concept of the present application, and these all belong to the protection scope of the present application.
Claims
1. A three-phase three-level ANPC inverter parameter monitoring method based on digital twin, characterized in that: include: Establish an inverter mathematical model and perform discrete processing, the inverter mathematical model includes an inverter system mathematical model and an inverter main circuit power semiconductor device mathematical model, in which the on-state voltage drop is used as a monitoring indicator to characterize the health status of the power semiconductor device; build an inverter digital twin based on the inverter mathematical model after discrete processing; design an inverter multi-parameter monitoring sequence; based on the actual parameters of the inverter, use an optimization algorithm to update the internal key parameters of the inverter digital twin to obtain an accurate inverter digital twin; use the accurate inverter digital twin to simulate the input and output characteristics of the physical inverter and monitor the parameters of key components based on the inverter multi-parameter monitoring sequence, wherein the inverter is a three-phase three-level ANPC inverter; The inverter system mathematical model is as follows: Among them, u an 、u bn 、u cn are the three-phase output voltages of the three-level ANPC inverter, i a 、i b 、i c is the three-phase output phase current, p represents the differential coefficient, L a , L b , L c With R a , R b , R c They are the three-phase load inductance and resistance, U c1 , U c2 represents the voltage across the capacitor collected by the voltage sensor, C1 and C2 represent the capacitance of the two supporting capacitors on the DC side, and R ESR1 , R ESR2 Represents the parasitic resistance of the capacitor, u c1 、u c2 Represents the voltage supported by the capacitor.
2. The three-phase three-level ANPC inverter parameter monitoring method based on digital twin according to claim 1 is characterized in that: The mathematical model of the inverter main circuit designed based on the power semiconductor device model is as follows: Wherein, u represents the modulation voltage signal, u>0, represents the power semiconductor device of the upper half bridge arm of the inverter is working, similarly, u<0, represents the power semiconductor device of the lower half bridge arm of the inverter is working; u ao 、u on represents the terminal voltage, where u ao Represents the voltage between the inverter midpoint 'o' and the A-phase bridge arm output point 'a'; u on Represents the voltage between the midpoint 'o' of the inverter and the midpoint 'n' of the inverter load end; A p is the switching state of the power semiconductor device, and the P state corresponds to the A p The value is 1, and the O state corresponds to A p The value is 0, and the N state corresponds to A p The value is -1, A absp A p The absolute value of i d Represents the direction of current flow. If the current flows to the load, it is 1, otherwise it is -1. V CE 、V F Represents the conduction voltage drop of the power IGBT and the anti-parallel diode, V CE 、V F It can be equivalent to a voltage source series resistor. The mathematical model of the power semiconductor device in the inverter main circuit is expressed as: V CE =V o +R o i c (5) V F =V Do +R Do i c (6) Among them, V o , R o is the on-state voltage and on-state resistance of the power IGBT, i c is the current flowing through the chip, V Do , R Do are the on-state voltage and on-state resistance of the anti-parallel diode.
3. The three-phase three-level ANPC inverter parameter monitoring method based on digital twin according to claim 1 is characterized in that: The method of building an inverter digital twin based on the inverter mathematical model after discrete processing includes: writing the inverter mathematical model after discrete processing into MATLAB to obtain an inverter digital twin that can be independently operated offline, and the control part is mainly completed in a digital microcontroller; the input of the digital twin includes two supporting capacitor voltages, three-phase output current, line voltage and three-phase modulation signal generated in the digital microcontroller obtained by the sampling circuit, and the capacitor voltage and three-phase output current at the next moment are output after being solved by the digital twin, so as to simulate the input and output characteristics of the actual physical ANPC inverter system.
4. The three-phase three-level ANPC inverter parameter monitoring method based on digital twin according to claim 1 is characterized in that: The design of the inverter multi-parameter monitoring sequence includes: monitoring the parameters of the support capacitor and its parasitic resistance based on the mathematical model of the inverter DC side; based on the working principle of the inverter, first monitoring the power semiconductor device parameters of the main circuit, and then monitoring the power semiconductor device parameters used for clamping.
5. The three-phase three-level ANPC inverter parameter monitoring method based on digital twin according to claim 1 is characterized in that: The objective evaluation function used by the optimization algorithm to update parameters is: Among them, f obj,1 For parameter monitoring of capacitance and parasitic resistance, f obj,2 Used for parameter monitoring of load inductance and resistance, f obj,3 Used for parameter monitoring of power semiconductor devices, N is the amount of sampled data, U is c1,j ,U c2,j is the voltage sampling value of the two supporting capacitors on the DC side of the inverter, U cm1,j ,U cm2,j The voltage values of the two supporting capacitors calculated for the digital twin model, i a,j 、i b,j 、i c,j is the inverter three-phase output current sampling value, i am,j 、i bm,j 、i cm,j The three-phase output current value calculated by the digital twin model, u ab,j is the inverter output line voltage sampling value, u abm,j Output line voltage values calculated for the digital twin model.
6. The three-phase three-level ANPC inverter parameter monitoring platform based on digital twin is characterized by: It includes a data acquisition module, a parameter monitoring module and a data display module. The data acquisition module is used to obtain experimental data of the three-phase three-level ANPC inverter in steady state. The parameter monitoring module includes the accurate three-phase three-level ANPC inverter digital twin obtained in the three-phase three-level ANPC inverter parameter monitoring method based on digital twin as described in any one of claims 1-5, and is used to simulate actual input and output characteristics and monitor parameters of key components of the inverter according to the experimental data of the three-phase three-level ANPC inverter in steady state; the data display module is used to average the n monitoring values obtained by the parameter monitoring module as the final monitoring result and display them.