Multi-beat control algorithm for SiC inverter

Abstract

The invention relates to a power electronic technology, and discloses a multi-beat control algorithm for a SiC inverter, which is characterized in that a sensor or an instrument is used for collecting signal sample data needing to be analyzed, so that the sampling frequency is high enough, the high-frequency component of a signal is captured, and after the signal is obtained, the collected signal is preprocessed, such as noise removal and filtering, so that the multi-beat control algorithm for the SiC inverter is obtained. Fourier transform is carried out on the preprocessed signals, the signals are converted from a time domain to a frequency domain, the Fourier transform can decompose the signals into a series of frequency components including fundamental frequency and harmonic waves, main frequency components and corresponding amplitudes are searched in a spectrogram obtained after Fourier transform is analyzed, and the harmonic waves are generally integer multiples of the fundamental frequency. According to the multi-beat control algorithm for the SiC inverter, the sampling calculation frequency of the control algorithm can be greatly improved under the condition that a more expensive DSP (Digital Signal Processor) is not needed, so that the switching frequency of SiC MOSFET (Metal-Oxide-Semiconductor Field Effect Transistor) equipment is improved, and the performance of the inverter is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power electronics, and in particular to a multi-beat control algorithm for a SiC inverter. Background Art

[0002] With the development of science and technology, the new generation of SiC MOSFET power semiconductor devices has been widely used in new energy converters, automotive electric drives, switching power supplies, etc. due to its high power density, small size, low loss, high switching frequency and other advantages. With the increase of switching frequency, the sampling frequency of the inverter control algorithm should also be increased, so as to maximize the advantages of SiC MOSFET's high switching frequency and reduce the price of equipment inductance. However, the sampling calculation of the traditional single-beat control algorithm requires a certain amount of time, so it is necessary to purchase a DSP processing chip with a higher calculation frequency and higher cost. Therefore, the calculation time limits the switching frequency of the SiC MOSFET inverter, thereby reducing the control performance.

[0003] To address the above issues, it is necessary to design a 1+stepN multi-beat control mode. It is hoped that this mode can greatly improve the sampling calculation frequency of the control algorithm, increase the switching frequency of SiC MOSFET devices and improve inverter performance without the need for a more expensive DSP. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a multi-beat control algorithm for a SiC inverter, which has the advantages of improving inverter performance and solving the problem of insufficient inverter performance.

[0006] (II) Technical solution

[0007] In order to achieve the above-mentioned purpose of improving the inverter performance, the present invention provides the following technical solution: a multi-beat control algorithm for SiC inverter, comprising the following steps:

[0008] S1, Harmonic extraction of multi-beat 1+StepN control algorithm

[0009] Use sensors or instruments to collect signal sample data that needs to be analyzed, so that the sampling frequency is high enough and the high-frequency components of the signal are captured. After obtaining the signal, preprocess the collected signal, such as removing noise, filtering, etc., and then perform Fourier transform on the preprocessed signal to convert the signal from the time domain to the frequency domain. The Fourier transform can decompose the signal into a series of frequency components, including the fundamental frequency and harmonics. In the spectrum diagram after Fourier transform analysis, find the main frequency components and the corresponding amplitudes. Harmonics are usually integer multiples of the fundamental frequency. The existence of harmonics can be determined by finding the peaks in the spectrum diagram, and the frequency and amplitude information of the harmonic components can be extracted based on the results of the spectrum analysis. You can choose to extract specific harmonics or all harmonics as needed;

[0010] S2, Reactive power extraction of multi-beat 1+StepN control algorithm

[0011] The acquired signal is converted into digital form for processing. Similarly, the acquired signal may also contain noise or interference. Then, preprocessing is also required, such as filtering, denoising, etc., and mathematical methods such as Fourier transform and instantaneous reactive power theory are used to calculate the reactive component in the signal. This usually involves decomposing the signal into active and reactive components, and extracting the reactive component separately. The extracted reactive component is used as the input of the multi-beat 1+StepN control algorithm. The algorithm adjusts the control parameters or output of the system according to the size and direction of the reactive component to achieve the goal of reactive compensation or optimize system performance.

[0012] S3, unbalance extraction of multi-beat 1+StepN control algorithm

[0013] After the processed signal is processed and preprocessed accordingly, it is necessary to extract features related to the imbalance state. These features may be statistical quantities of the data, such as mean, variance, spectral features, time domain features, etc. The purpose of feature extraction is to convert the original data into a low-dimensional representation that can reflect the imbalance state. Then, the extracted features can be used to build an imbalance detection model. The model may be based on machine learning algorithms, statistical methods or expert systems to determine whether the system is in an imbalanced state. The imbalance detection model needs to be trained and learned to accurately identify imbalance features. Once the imbalance state is detected, the algorithm needs to extract specific imbalance information, such as the type, degree, and location of the imbalance. This information can be used for subsequent fault diagnosis, optimal control or maintenance decisions. At the same time, the results of the imbalance extraction can also be fed back to the system to adjust the control parameters or optimize the algorithm performance.

[0014] S4, voltage balancing and stabilization algorithm of multi-beat 1+StepN control algorithm

[0015] In the multi-beat mechanism, the voltage of each unit that needs voltage balancing and stabilization will be sampled in real time according to the set time beat or event beat. In an energy storage system composed of multiple series battery packs, the current voltage value of each battery pack is accurately obtained through the voltage sensor at every fixed time interval or beat. These sampled data will be used as the basis for subsequent judgment and control adjustment, and the voltage difference between each unit voltage obtained by sampling will be calculated, and these differences will be compared with the pre-set voltage balancing threshold. Once the difference exceeds the threshold range, it means that voltage imbalance has occurred and adjustment is required, and the start The trigger judgment basis of voltage equalization adjustment is that when it is determined that voltage equalization is required, the process enters Step N, i.e., the voltage equalization and voltage regulation process. For units with higher voltage, appropriate switch control can be used to discharge through resistors to reduce the voltage. For units with lower voltage, energy can be supplemented from other power supplies or energy storage units through corresponding circuits. In addition, the voltage feedback at the output end of the entire system is combined with a PID (proportional-integral-differential) control strategy to dynamically adjust the output power link of the power supply (e.g., adjust the duty cycle of the DC-DC converter, etc.) so that the final output voltage is stabilized near the set target value.

[0016] S5, Current loop algorithm of multi-beat 1+StepN control algorithm

[0017] After the execution of the above method and the reverse current loop algorithm, the target current value must first be set according to the system's operating requirements. This value may be determined based on load characteristics, such as the rated current of the motor at different speeds and torque requirements, the overall power planning of the system, and other factors. Then, the actual current value obtained by sampling is compared with the target current value to calculate the current deviation, that is, deviation = actual current - target current. When the deviation exceeds the pre-set allowable range, this is a key step in determining whether the current deviates from the normal control requirements. For the "1" part, it is necessary to start the subsequent adjustment steps to correct the current deviation so that the actual current approaches the target current.

[0018] Preferably, in step S1, the harmonic extraction process may also need to consider the control logic and dynamic characteristics of the algorithm. For example, if the algorithm involves precise control of motor current or voltage, the harmonic extraction process may need to be closely integrated with the motor control strategy to ensure that the normal operation of the system is not affected or interfered with while the harmonics are extracted.

[0019] Preferably, in step S1, different control algorithms and application scenarios may require different harmonic extraction methods. For example, in power systems, more complex signal processing techniques may be used to deal with complex harmonic situations, while in motor control, a harmonic detection method based on instantaneous reactive power theory may be used to measure harmonics in real time and accurately.

[0020] Preferably, in step S2, a feedback mechanism may be introduced in the process of adjusting the control parameters or output of the system to dynamically adjust the control parameters according to the difference between the actual response of the system and the expected target, so as to ensure the accuracy and effectiveness of reactive power extraction and compensation.

[0021] Preferably, in step S3, during the imbalance extraction process, the accuracy and reliability of the data are crucial. Therefore, it is necessary to ensure that the collected data is true and valid and is not affected by noise, interference or other factors. This may require the use of data processing techniques such as filtering and denoising to improve the quality of the data.

[0022] Preferably, in step S3, the unbalanced extraction process needs to consider the dynamic characteristics of the system. For example, factors such as the response speed, stability, and nonlinearity of the system may affect the unbalanced extraction. During the extraction process, these factors need to be fully considered and corresponding measures need to be taken to ensure the accuracy of the extraction results.

[0023] Preferably, in step S4, due to the use of multi-beat periodic sampling and detailed difference comparison and judgment, voltage imbalance and output voltage deviation can be discovered in a timely and relatively accurate manner, so that a quick response can be made for adjustment. In addition, through the combination of multiple adjustment strategies in Step N, it can adapt to different power system architectures, different load changes and various initial voltage distribution states, and can be well applied to both battery energy storage systems and industrial control scenarios powered by multiple power sources.

[0024] Preferably, in step S5, with the help of frequent sampling of multiple beats and a flexible StepN adjustment strategy, the current can be tracked and controlled with high precision, so that the error between the actual current and the target current is kept within a very small range, meeting the application scenarios with high requirements for current accuracy.

[0025] (III) Beneficial effects

[0026] Compared with the prior art, the present invention provides a multi-beat control algorithm for SiC inverter, which has the following beneficial effects:

[0027] 1. The multi-beat control algorithm for SiC inverters can more accurately control the output voltage, current and other parameters of the power electronic converter by adjusting the control amount multiple times within a switching cycle, so that the error between the output voltage and current and the reference value is smaller, which effectively improves the control accuracy. It reduces the sampling period and increases the calculation frequency during operation. It can fully improve the switching frequency of SiC MOSFET in SiC inverters. With the increase of the switching frequency of power switching devices, the inductance value used by the inverter can also be greatly reduced, so that a smaller inductor can meet the product design requirements. Compared with IGBT products of the same power, it has a higher power density and less calculation time than a single beat. In the field of software applications, it can not only improve the response speed of the software, but also the software function is not lost. Without the need for a more expensive DSP, the sampling and calculation frequency of the control algorithm can be greatly improved, thereby increasing the switching frequency of the SiC MOSFET device and greatly improving the inverter performance.

[0028] 2. The multi-beat control algorithm for SiC inverters can effectively reduce the harmonic content of output voltage and current by optimizing the control strategy and finely controlling the on and off of power switching devices in multiple beats, making the output waveform closer to the ideal waveform, thereby improving the output power quality and reducing electromagnetic interference to the power grid and other electrical equipment. When faced with dynamic situations such as load changes and input voltage fluctuations, it can quickly respond and adjust the control strategy. Since the system status can be monitored and adjusted in each beat, compared with the traditional single-beat control algorithm, the algorithm of the present invention can restore the system to stability in a shorter time, greatly enhancing the dynamic response capability of the power electronic conversion system. In addition, with the help of frequent sampling of multiple beats and flexible StepN adjustment strategy, high-precision tracking and control of current can be achieved, so that the error between the actual current and the target current is kept within a very small range, meeting the application scenarios with high requirements for current accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the preparation method of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] Embodiment 1:

[0032] A multi-beat control algorithm for a SiC inverter comprises the following steps:

[0033] S1, Harmonic extraction of multi-beat 1+StepN control algorithm

[0034] Use sensors or instruments to collect signal sample data that needs to be analyzed, so that the sampling frequency is high enough and the high-frequency components of the signal are captured. After obtaining the signal, preprocess the collected signal, such as removing noise, filtering, etc., and then perform Fourier transform on the preprocessed signal to convert the signal from the time domain to the frequency domain. The Fourier transform can decompose the signal into a series of frequency components, including the fundamental frequency and harmonics. In the spectrum diagram after Fourier transform analysis, find the main frequency components and the corresponding amplitudes. Harmonics are usually integer multiples of the fundamental frequency. The existence of harmonics can be determined by finding the peaks in the spectrum diagram, and the frequency and amplitude information of the harmonic components can be extracted based on the results of the spectrum analysis. You can choose to extract specific harmonics or all harmonics as needed;

[0035] S2, Reactive power extraction of multi-beat 1+StepN control algorithm

[0036] The acquired signal is converted into digital form for processing. Similarly, the acquired signal may also contain noise or interference. Then, preprocessing is also required, such as filtering, denoising, etc., and mathematical methods such as Fourier transform and instantaneous reactive power theory are used to calculate the reactive component in the signal. This usually involves decomposing the signal into active and reactive components, and extracting the reactive component separately. The extracted reactive component is used as the input of the multi-beat 1+StepN control algorithm. The algorithm adjusts the control parameters or output of the system according to the size and direction of the reactive component to achieve the goal of reactive compensation or optimize system performance.

[0037] S3, unbalance extraction of multi-beat 1+StepN control algorithm

[0038] After the processed signal is processed and preprocessed accordingly, it is necessary to extract features related to the imbalance state. These features may be statistical quantities of the data, such as mean, variance, spectral features, time domain features, etc. The purpose of feature extraction is to convert the original data into a low-dimensional representation that can reflect the imbalance state. Then, the extracted features can be used to build an imbalance detection model. The model may be based on machine learning algorithms, statistical methods or expert systems to determine whether the system is in an imbalanced state. The imbalance detection model needs to be trained and learned to accurately identify imbalance features. Once the imbalance state is detected, the algorithm needs to extract specific imbalance information, such as the type, degree, and location of the imbalance. This information can be used for subsequent fault diagnosis, optimal control or maintenance decisions. At the same time, the results of the imbalance extraction can also be fed back to the system to adjust the control parameters or optimize the algorithm performance.

[0039] S4, voltage balancing and stabilization algorithm of multi-beat 1+StepN control algorithm

[0040] In the multi-beat mechanism, the voltage of each unit that needs voltage balancing and stabilization will be sampled in real time according to the set time beat or event beat. In an energy storage system composed of multiple series battery packs, the current voltage value of each battery pack is accurately obtained through the voltage sensor at every fixed time interval or beat. These sampled data will be used as the basis for subsequent judgment and control adjustment, and the voltage difference between each unit voltage obtained by sampling will be calculated, and these differences will be compared with the pre-set voltage balancing threshold. Once the difference exceeds the threshold range, it means that voltage imbalance has occurred and adjustment is required, and the start The trigger judgment basis of voltage equalization adjustment is that when it is determined that voltage equalization is required, the process enters Step N, i.e., the voltage equalization and voltage regulation process. For units with higher voltage, appropriate switch control can be used to discharge through resistors to reduce the voltage. For units with lower voltage, energy can be supplemented from other power supplies or energy storage units through corresponding circuits. In addition, the voltage feedback at the output end of the entire system is combined with a PID (proportional-integral-differential) control strategy to dynamically adjust the output power link of the power supply (e.g., adjust the duty cycle of the DC-DC converter, etc.) so that the final output voltage is stabilized near the set target value.

[0041] S5, Current loop algorithm of multi-beat 1+StepN control algorithm

[0042] After the execution of the above method and the reverse current loop algorithm, the target current value must first be set according to the operating requirements of the system. This value may be determined based on load characteristics, such as the rated current of the motor at different speeds and torque requirements, the overall power planning of the system, and other factors. The actual current value obtained by sampling is then compared with the target current value to calculate the current deviation, that is, deviation = actual current - target current. When the deviation exceeds the pre-set allowable range, this is a key step in determining whether the current deviates from the normal control requirements. Corresponding to the "1" part, it is necessary to start subsequent adjustment steps to correct the current deviation so that the actual current approaches the target current.

[0043] Embodiment 2:

[0044] A multi-beat control algorithm for SiC inverter, characterized by comprising the following steps:

[0045] S1, Harmonic extraction of multi-beat 1+StepN control algorithm

[0046] Use sensors or instruments to collect signal sample data that needs to be analyzed, so that the sampling frequency is high enough and the high-frequency components of the signal are captured. After obtaining the signal, preprocess the collected signal, such as removing noise, filtering, etc., and then perform Fourier transform on the preprocessed signal to convert the signal from the time domain to the frequency domain. The Fourier transform can decompose the signal into a series of frequency components, including the fundamental frequency and harmonics. In the spectrum diagram after Fourier transform analysis, find the main frequency components and the corresponding amplitudes. Harmonics are usually integer multiples of the fundamental frequency. The existence of harmonics can be determined by finding the peaks in the spectrum diagram, and the frequency and amplitude information of the harmonic components can be extracted based on the results of the spectrum analysis. You can choose to extract specific harmonics or all harmonics as needed;

[0047] S2, Reactive power extraction of multi-beat 1+StepN control algorithm

[0048] The acquired signal is converted into digital form for processing. Similarly, the acquired signal may also contain noise or interference. Then, preprocessing is also required, such as filtering, denoising, etc., and mathematical methods such as Fourier transform and instantaneous reactive power theory are used to calculate the reactive component in the signal. This usually involves decomposing the signal into active and reactive components, and extracting the reactive component separately. The extracted reactive component is used as the input of the multi-beat 1+StepN control algorithm. The algorithm adjusts the control parameters or output of the system according to the size and direction of the reactive component to achieve the goal of reactive compensation or optimize system performance.

[0049] S3, unbalance extraction of multi-beat 1+StepN control algorithm

[0050] After the processed signal is processed and preprocessed accordingly, it is necessary to extract features related to the imbalance state. These features may be statistical quantities of the data, such as mean, variance, spectral features, time domain features, etc. The purpose of feature extraction is to convert the original data into a low-dimensional representation that can reflect the imbalance state. Then, the extracted features can be used to build an imbalance detection model. The model may be based on machine learning algorithms, statistical methods or expert systems to determine whether the system is in an imbalanced state. The imbalance detection model needs to be trained and learned to accurately identify imbalance features. Once the imbalance state is detected, the algorithm needs to extract specific imbalance information, such as the type, degree, and location of the imbalance. This information can be used for subsequent fault diagnosis, optimal control or maintenance decisions. At the same time, the results of the imbalance extraction can also be fed back to the system to adjust the control parameters or optimize the algorithm performance.

[0051] S4, voltage balancing and stabilization algorithm of multi-beat 1+StepN control algorithm

[0052] In the multi-beat mechanism, the voltage of each unit that needs voltage balancing and stabilization will be sampled in real time according to the set time beat or event beat. In an energy storage system composed of multiple series battery packs, the current voltage value of each battery pack is accurately obtained through the voltage sensor at every fixed time interval or beat. These sampled data will be used as the basis for subsequent judgment and control adjustment, and the voltage difference between each unit voltage obtained by sampling will be calculated, and these differences will be compared with the pre-set voltage balancing threshold. Once the difference exceeds the threshold range, it means that voltage imbalance has occurred and adjustment is required, and the start The trigger judgment basis of voltage equalization adjustment is that when it is determined that voltage equalization is required, the process enters Step N, i.e., the voltage equalization and voltage regulation process. For units with higher voltage, appropriate switch control can be used to discharge through resistors to reduce the voltage. For units with lower voltage, energy can be supplemented from other power supplies or energy storage units through corresponding circuits. In addition, the voltage feedback at the output end of the entire system is combined with a PID (proportional-integral-differential) control strategy to dynamically adjust the output power link of the power supply (e.g., adjust the duty cycle of the DC-DC converter, etc.) so that the final output voltage is stabilized near the set target value.

[0053] S5, Current loop algorithm of multi-beat 1+StepN control algorithm

[0054] After the execution of the above method and the reverse current loop algorithm, the target current value must first be set according to the system's operating requirements. This value may be determined based on load characteristics, such as the rated current of the motor at different speeds and torque requirements, the overall power planning of the system, and other factors. Then, the actual current value obtained by sampling is compared with the target current value to calculate the current deviation, that is, deviation = actual current - target current. When the deviation exceeds the preset allowable range, this is a key step in determining whether the current deviates from the normal control requirements. Corresponding to the "1" part, it is necessary to start the subsequent adjustment steps to correct the current deviation so that the actual current approaches the target current.

[0055] Specifically, in step S1, the harmonic extraction process may also need to consider the control logic and dynamic characteristics of the algorithm. For example, if the algorithm involves precise control of motor current or voltage, the harmonic extraction process may need to be closely integrated with the motor control strategy to ensure that the normal operation of the system is not affected or disturbed while the harmonics are extracted;

[0056] Specifically, in step S1, different control algorithms and application scenarios may require different harmonic extraction methods. In power systems, more complex signal processing techniques may be used to deal with complex harmonic situations, while in motor control, a harmonic detection method based on instantaneous reactive power theory may be used to measure harmonics in real time and accurately.

[0057] Specifically, in step S2, in the process of adjusting the control parameters or output of the system, a feedback mechanism may be introduced to dynamically adjust the control parameters according to the difference between the actual response of the system and the expected target, so as to ensure the accuracy and effectiveness of reactive power extraction and compensation;

[0058] Specifically, in step S3, in the imbalance extraction process, the accuracy and reliability of the data are crucial. Therefore, it is necessary to ensure that the collected data is true and valid, and is not affected by noise, interference or other factors. This may require the use of data processing techniques such as filtering and denoising to improve the quality of the data;

[0059] Specifically, in step S3, the unbalanced extraction process needs to consider the dynamic characteristics of the system. For example, factors such as the response speed, stability, and nonlinearity of the system may affect the unbalanced extraction. In the extraction process, these factors need to be fully considered and corresponding measures need to be taken to ensure the accuracy of the extraction results.

[0060] Specifically, in step S4, due to the use of multi-beat periodic sampling and detailed difference comparison and judgment, voltage imbalance and output voltage deviation can be discovered in a timely and relatively accurate manner, so that adjustments can be made in a quick response. In addition, through the combination of multiple adjustment strategies in Step N, different power system architectures, different load changes, and various initial voltage distribution states can be adapted, and both battery energy storage systems and industrial control scenarios powered by multiple power sources can be well applied;

[0061] Specifically, in step S5, with the help of frequent sampling of multiple beats and a flexible StepN adjustment strategy, the current can be tracked and controlled with high precision, so that the error between the actual current and the target current is kept within a very small range, meeting the application scenarios with high requirements for current accuracy.

[0062] In actual operation, some of the algorithms need to be calculated in each sampling period, such as the current loop control algorithm of the SIC inverter, and the other part is the step-by-step calculation algorithm, such as the harmonic extraction, reactive power extraction, unbalance extraction, and voltage stabilization algorithm in the SIC inverter. This can not only reduce the calculation time of the MCU in each step, but also achieve the same algorithm purpose after multiple steps, so that low-performance MCU chips can be selected without affecting the function and performance of the device.

[0063] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-beat control algorithm for SiC inverter, characterized in that: The following steps are involved: S1, Harmonic extraction of multi-beat 1+StepN control algorithm Use sensors or instruments to collect signal sample data that needs to be analyzed, so that the sampling frequency is high enough and the high-frequency components of the signal are captured. After obtaining the signal, preprocess the collected signal, such as removing noise, filtering, etc., and then perform Fourier transform on the preprocessed signal to convert the signal from the time domain to the frequency domain. The Fourier transform can decompose the signal into a series of frequency components, including the fundamental frequency and harmonics. In the spectrum diagram after Fourier transform analysis, find the main frequency components and the corresponding amplitudes. Harmonics are usually integer multiples of the fundamental frequency. The existence of harmonics can be determined by finding the peaks in the spectrum diagram, and the frequency and amplitude information of the harmonic components can be extracted based on the results of the spectrum analysis. You can choose to extract specific harmonics or all harmonics as needed; S2, Reactive power extraction of multi-beat 1+StepN control algorithm The acquired signal is converted into digital form for processing. Similarly, the acquired signal may also contain noise or interference. Then, preprocessing is also required, such as filtering, denoising, etc., and mathematical methods such as Fourier transform and instantaneous reactive power theory are used to calculate the reactive component in the signal. This usually involves decomposing the signal into active and reactive components, and extracting the reactive component separately. The extracted reactive component is used as the input of the multi-beat 1+StepN control algorithm. The algorithm adjusts the control parameters or output of the system according to the size and direction of the reactive component to achieve the goal of reactive compensation or optimize system performance. S3, unbalance extraction of multi-beat 1+StepN control algorithm After the processed signal is processed and preprocessed accordingly, it is necessary to extract features related to the imbalance state. These features may be statistical quantities of the data, such as mean, variance, spectral features, time domain features, etc. The purpose of feature extraction is to convert the original data into a low-dimensional representation that can reflect the imbalance state. Then, the extracted features can be used to build an imbalance detection model. The model may be based on machine learning algorithms, statistical methods or expert systems to determine whether the system is in an imbalanced state. The imbalance detection model needs to be trained and learned to accurately identify imbalance features. Once the imbalance state is detected, the algorithm needs to extract specific imbalance information, such as the type, degree, and location of the imbalance. This information can be used for subsequent fault diagnosis, optimal control or maintenance decisions. At the same time, the results of the imbalance extraction can also be fed back to the system to adjust the control parameters or optimize the algorithm performance. S4, voltage balancing and stabilization algorithm of multi-beat 1+StepN control algorithm In the multi-beat mechanism, the voltage of each unit that needs voltage balancing and stabilization will be sampled in real time according to the set time beat or event beat. In an energy storage system composed of multiple series battery packs, the current voltage value of each battery pack is accurately obtained through the voltage sensor at every fixed time interval or beat. These sampled data will be used as the basis for subsequent judgment and control adjustment, and the voltage difference between each unit voltage obtained by sampling will be calculated, and these differences will be compared with the pre-set voltage balancing threshold. Once the difference exceeds the threshold range, it means that voltage imbalance has occurred and adjustment is required, and the start The trigger judgment basis of voltage equalization adjustment is that when it is determined that voltage equalization is required, the process enters Step N, i.e., the voltage equalization and voltage regulation process. For units with higher voltage, appropriate switch control can be used to discharge through resistors to reduce the voltage. For units with lower voltage, energy can be supplemented from other power supplies or energy storage units through corresponding circuits. In addition, the voltage feedback at the output end of the entire system is combined with a PID (proportional-integral-differential) control strategy to dynamically adjust the output power link of the power supply (e.g., adjust the duty cycle of the DC-DC converter, etc.) so that the final output voltage is stabilized near the set target value. S5, Current loop algorithm of multi-beat 1+StepN control algorithm After the execution of the above method and the reverse current loop algorithm, the target current value must first be set according to the system's operating requirements. This value may be determined based on load characteristics, such as the rated current of the motor at different speeds and torque requirements, the overall power planning of the system, and other factors. Then, the actual current value obtained by sampling is compared with the target current value to calculate the current deviation, that is, deviation = actual current - target current. When the deviation exceeds the pre-set allowable range, this is a key step in determining whether the current deviates from the normal control requirements. Corresponding to the "1" part, it is necessary to start the subsequent adjustment steps to correct the current deviation so that the actual current approaches the target current.

2. A multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S1, the harmonic extraction process may also need to consider the control logic and dynamic characteristics of the algorithm. For example, if the algorithm involves precise control of motor current or voltage, the harmonic extraction process may need to be closely integrated with the motor control strategy to ensure that the normal operation of the system is not affected or interfered with while the harmonics are extracted.

3. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S1, different control algorithms and application scenarios may require different harmonic extraction methods. For example, in power systems, more complex signal processing techniques may be used to deal with complex harmonic situations, while in motor control, a harmonic detection method based on instantaneous reactive power theory may be used to measure harmonics in real time and accurately.

4. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S2, a feedback mechanism may be introduced in the process of adjusting the control parameters or output of the system to dynamically adjust the control parameters according to the difference between the actual response of the system and the expected target, so as to ensure the accuracy and effectiveness of reactive power extraction and compensation.

5. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S3, during the imbalance extraction process, the accuracy and reliability of the data are crucial. Therefore, it is necessary to ensure that the collected data is true and valid and is not affected by noise, interference or other factors. This may require the use of data processing techniques such as filtering and denoising to improve the quality of the data.

6. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S3, the unbalanced extraction process needs to consider the dynamic characteristics of the system. For example, factors such as the response speed, stability, and nonlinearity of the system may affect the unbalanced extraction. During the extraction process, these factors need to be fully considered and corresponding measures need to be taken to ensure the accuracy of the extraction results.

7. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S4, due to the use of multi-beat periodic sampling and detailed difference comparison and judgment, voltage imbalance and output voltage deviation can be discovered in a timely and relatively accurate manner, so that a quick response can be made for adjustment. In addition, through the combination of multiple adjustment strategies in Step N, it can adapt to different power system architectures, different load changes, and various initial voltage distribution states, and can be well applied to both battery energy storage systems and industrial control scenarios powered by multiple power sources.

8. The multi-beat control algorithm for SiC inverter according to claim 1, characterized in that: In step S5, with the help of frequent sampling of multiple beats and a flexible StepN adjustment strategy, the current can be tracked and controlled with high precision, so that the error between the actual current and the target current is kept within a very small range, meeting the application scenarios with high requirements for current accuracy.

Images