Voltage-sharing control method and system for modular multilevel converter, medium and product

Abstract

The application discloses a voltage equalization control method and system of a modular multilevel converter, a medium and a product, and belongs to the field of AC / DC converter control. The method is as follows: real-time electrical quantities and switching states of each sub-module of the converter are acquired; real-time junction temperature of a power device is calculated based on a thermal network model to determine a control mode; if in a high-temperature protection mode, only target sub-modules meeting voltage conditions are selected from bypassed or switched-on sub-modules according to modulation requirements to generate switching state data of a current period; if not in the high-temperature protection mode, all sub-module capacitor voltages are sorted, and a corresponding number of sub-modules are selected according to modulation requirements to generate switching state data of the current period; and drive pulse signals are generated according to the switching state data to control the operation of the converter. Through implementation of the application, the problem that the overload capacity is limited due to rapid rise of the junction temperature of the power device caused by rapid increase of switching loss of the modular multilevel converter under short-time overload working conditions can be solved.

Description

Technical Field

[0001] This invention belongs to the field of AC / DC converter control, and relates to a voltage equalization control method, system, medium and product for a modular multilevel converter. Background Technology

[0002] Modular multilevel converters, with their excellent voltage scalability and low harmonic characteristics, have become the core power conversion devices in the fields of flexible DC transmission and new energy grid connection. In actual operation, they need to be able to withstand short-term overloads of 1.2 to 1.5 times the rated current for several seconds to tens of seconds to cope with transient conditions such as grid voltage drops. Ensuring that the converter does not suffer hardware damage during overload and maintaining grid stability is a key requirement for the safe operation of new power systems.

[0003] Existing technologies typically address overload thermal stress by increasing the number of parallel devices or enhancing the hardware redundancy of the heat dissipation system. However, this significantly increases the size, weight, and construction cost of the converter valve. Alternatively, existing thermal equilibrium control strategies can be employed, which merely redistribute thermal stress among various sub-modules. These strategies cannot fundamentally reduce the overall heat generation of the bridge arm and have very limited ability to suppress the overall junction temperature peak under severe overload conditions. Furthermore, transient junction temperature exceeding limits has become a core physical bottleneck restricting the improvement of converter overload capacity. Summary of the Invention

[0004] This application provides a voltage equalization control method, system, medium, and product for a modular multilevel converter, which can solve the problem that the junction temperature of power devices rises rapidly due to the surge in switching losses under short-term overload conditions, thus limiting the overload capacity of the modular multilevel converter.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a voltage equalization control method for a modular multilevel converter, comprising: Acquire real-time operating electrical quantities and switching status signals of each submodule in the modular multilevel converter; Based on the real-time operating electrical quantities and switch status signals, the real-time junction temperature data of the power devices in each sub-module is calculated using a preset thermal network model, and the corresponding control mode is determined based on the real-time junction temperature data. When the control mode is high temperature protection mode, according to the current modulation requirements, only the target sub-module that meets the voltage conditions is selected from the sub-modules that are in either the bypass state or the on state to be put into or cut out, so as to obtain the switching status data of each sub-module in the current control cycle; otherwise, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected to be put into operation according to the modulation requirements, so as to obtain the switching status data of each sub-module in the current control cycle. Based on the switching status data of each submodule in the current control cycle, drive pulse signals are generated to control the operation of the converter.

[0006] Compared to existing technologies, the embodiments of this application have the following advantages: They acquire real-time operating electrical quantities and switching status signals of each submodule in the converter, providing accurate underlying data support for subsequent thermal state assessment and control decisions; based on the real-time operating electrical quantities and switching status signals, they calculate the real-time junction temperature data of the power devices in each submodule, enabling online soft measurement of the internal thermal state of the power devices, avoiding the increased hardware costs and response lag issues caused by directly installing temperature sensors, and accurately capturing transient temperature rise trends under overload conditions; based on the real-time junction temperature data, they determine the corresponding control mode. When the control mode is the high-temperature protection mode, according to the current modulation requirements, they select target submodules that meet the voltage conditions from the submodules in either the bypass or active state for activation or deactivation, obtaining the switching status data of each submodule in the current control cycle. This feature significantly reduces the number of submodules involved in the sequencing and action by limiting the set of submodules participating in the process. Unnecessary switching operations are minimized, significantly reducing switching losses and effectively curbing further increases in junction temperature during overload. This solves the problem of thermal runaway caused by frequent switching in traditional voltage equalization strategies under heavy loads. Alternatively, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected for operation based on modulation requirements. This yields the switching status data of each sub-module in the current control cycle, ensuring that the system maintains high-precision capacitor voltage balance under non-overheating conditions, guaranteeing power quality on the grid side and DC voltage stability. Finally, based on the switching status data of each sub-module in the current control cycle, a drive pulse signal is generated to control the converter operation, achieving a closed loop from control algorithm to physical execution. The synergistic effect of these features, through the combination of "real-time junction temperature monitoring" and "dual-mode strategy switching," significantly improves the thermal survivability and operational reliability of the modular multilevel converter under short-term overload conditions without increasing hardware heat dissipation redundancy.

[0007] In some embodiments of the first aspect of this application, the step of calculating the real-time junction temperature data of the power devices of each submodule using a preset thermal network model based on the real-time operating electrical quantities and switch status signals includes: Extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities; The type of power device in the conducting state in each submodule is determined based on the switch status signal at the current moment, and the transient conduction loss value of each power device is calculated by combining the bridge arm current data and the preset device conduction characteristic parameters. The power device that is switching at the current moment is identified based on the switching status signal, and the transient switching loss value of each power device is calculated by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. Add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device; The total instantaneous power loss data is input as a heat source into a preset thermal resistance-capacitance network model, and the transient junction temperature estimate of each power device is obtained through iterative calculation. For each submodule, the maximum value among the estimated transient junction temperatures of all power devices within that submodule is taken as the real-time junction temperature data for that submodule.

[0008] Compared to existing technologies, the above embodiments have the following advantages: They extract the current arm current data and capacitor voltage data of each submodule from real-time operating electrical quantities, clarifying the physical input variables for heat loss calculation. Furthermore, based on the switching status signal, they determine the type of power device in the conducting state of each submodule at the current moment. Combining the arm current data and preset device conduction characteristic parameters, they calculate the transient conduction loss value of each power device, achieving refined quantification of steady-state heat sources and reflecting the cumulative ohmic loss effect of the device during continuous conduction. Simultaneously, based on the switching status signal, they identify the power device that has switched at the current moment. Combining the capacitor voltage data, arm current data, and preset device switching characteristic parameters, they calculate the transient switching loss value of each power device, accurately capturing the dynamic heating impact caused by state transitions, and compensating for the temperature rise estimation deviation caused by only considering conduction losses. By adding the transient conduction loss value and transient switching loss value of each power device, the total transient power loss data of each power device is obtained, and a complete transient heat source model is constructed, providing accurate power excitation for thermal network iteration. Then, the total transient power loss data is used as a heat source input into a preset thermal resistance-capacitance network model, and the transient junction temperature estimate of each power device is obtained through iterative calculation. The physical process of heat transfer and accumulation inside the device is simulated by utilizing the inertial characteristics of the thermal resistance-capacitance model, realizing high-precision reconstruction of the chip junction temperature that cannot be directly measured. Finally, for each submodule, the maximum value of the transient junction temperature estimate of all power devices in the submodule is taken as the real-time junction temperature data of the submodule. Following the "weakest link" principle, the worst thermal state is used as the control basis, ensuring that the protection mechanism can be triggered in time even under the most dangerous operating conditions, thus improving the thermal safety margin of the system.

[0009] In some embodiments of the first aspect of this application, determining the corresponding control mode based on the real-time junction temperature data includes: Determine whether the real-time junction temperature data is greater than or equal to the preset upper temperature threshold. If the determination result is yes, then it is determined to be in high temperature protection mode; otherwise, determine whether the real-time junction temperature data is less than the preset lower temperature threshold. If the duration for which the real-time junction temperature data is less than the preset lower temperature threshold is greater than the preset delay time, then it is determined to be in normal pressure equalization mode; otherwise, the control mode of the previous control cycle remains unchanged.

[0010] Compared to existing technologies, the above embodiments have the following advantages: By determining whether the real-time junction temperature data is greater than or equal to a preset upper temperature threshold, if the determination result is yes, it is determined to be in high-temperature protection mode, realizing an immediate response to overheating risks and ensuring that the device quickly switches to a low-power control strategy to prevent thermal breakdown when it approaches the safety limit; otherwise, it is determined whether the real-time junction temperature data is less than a preset lower temperature threshold. If the duration of the real-time junction temperature data being less than the preset lower temperature threshold is greater than a preset delay time, it is determined to be in conventional voltage equalization mode, introducing a time-dimensional stability criterion to prevent frequent oscillations in the control mode caused by temperature measurement noise or short-term fluctuations, ensuring the smoothness of system state switching; if none of the above conditions are met, the control mode of the previous control cycle remains unchanged, constructing hysteresis control logic, maintaining the continuity of the control strategy in the transition zone between the upper and lower temperature limits, avoiding the ping-pong effect near the critical point, thereby improving the control stability and robustness of the converter during dynamic load changes.

[0011] In some embodiments of the first aspect of this application, when the control mode is a high-temperature protection mode, according to the current modulation requirements, a target submodule that meets the voltage conditions is selected from the submodules in either a bypass state or an active state for activation or deactivation, to obtain the switching state data of each submodule in the current control cycle, including: Based on the nearest level approximation modulation algorithm, calculate the total number of sub-modules that need to be put into operation in the bridge arm of the converter at the current moment; The difference in quantity is obtained by subtracting the total number of submodules that need to be deployed in the bridge arm at the current moment from the total number of submodules actually deployed in the previous control cycle. If the difference in quantity is positive, then from the sub-modules currently in the bypass state, sort them in order of capacitor voltage from low to high, select the sub-modules with the difference in quantity mentioned above, change their state to the engaged state, and keep the states of the remaining sub-modules in the bypass state and all sub-modules in the engaged state unchanged. If the quantity difference is negative, then from the sub-modules currently in the input state, sort them in descending order of capacitor voltage, select the sub-modules with the absolute value of the quantity difference mentioned above, change their state to bypass state, and keep the states of the remaining sub-modules in the input state and all sub-modules in the bypass state unchanged. If the difference in quantity is zero, then the switching state of all sub-modules remains consistent with the previous control cycle.

[0012] Compared to existing technologies, the above embodiments have the following advantages: The total number of submodules that need to be engaged in the converter arm at the current moment is calculated based on the nearest-nearest-level approximation modulation algorithm, establishing the basic constraints for meeting the grid voltage following requirements; further, the total number of submodules that need to be engaged in the bridge arm at the current moment is subtracted from the total number of submodules actually engaged in the previous control cycle to obtain the quantity difference, quantifying the state change requirements of the current control cycle relative to the previous cycle, providing a calculation benchmark for local optimization; if the quantity difference is positive, the submodules currently in the bypass state are sorted in ascending order of capacitor voltage, and the first few submodules with the quantity difference are selected and their states are changed to the engaged state, while keeping the states of the remaining bypassed submodules and all engaged submodules unchanged. This incremental engagement strategy ensures charging balance (prioritizing low voltage) while maximizing the preservation of the original... The switching state of the input module is adjusted to reduce switching losses caused by the turn-on action. If the quantity difference is negative, the sub-modules currently in the input state are sorted in descending order of capacitor voltage, and the sub-modules with the absolute value of the aforementioned quantity difference are selected and their states are changed to bypass state. At the same time, the states of the remaining input sub-modules and all sub-modules in the bypass state remain unchanged. This reduction-type cut-off strategy ensures discharge balance (prioritizing high voltage cut-off) while avoiding large-scale state rearrangement and further suppressing the heat generated by the turn-off action. If the quantity difference is zero, the switching state of all sub-modules is kept consistent with the previous control cycle, achieving the ideal low-power state of zero switching action. Overall, through the mechanism of "state maintenance + local screening", the switching frequency is reduced to the theoretical minimum in high-temperature protection mode, directly serving the core technical objective of suppressing the rapid rise of junction temperature.

[0013] In some embodiments of the first aspect of this application, the step of sorting the capacitor voltages of all sub-modules and selecting a corresponding number of sub-modules for activation according to modulation requirements to obtain the switching state data of each sub-module in the current control cycle includes: Based on the modulation requirements and the rated voltage of the submodules, calculate the number of submodules required to be put into operation in the bridge arm of the converter during the current control cycle. The real-time capacitor voltage data of all sub-modules within the bridge arm are fully sorted to obtain a global voltage sorting sequence. The charging and discharging requirements are determined based on the current direction of the current in the current arm. If charging is required, several sub-modules with the lowest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. If discharging is required, several sub-modules with the highest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. The unselected submodules are marked as bypassed, thereby obtaining the switching status data of each submodule in the current control cycle.

[0014] Compared to existing technologies, the above embodiments have the following advantages: Based on modulation requirements and the rated voltage of the submodules, the required number of submodules to be deployed in the converter arm during the current control cycle is calculated, transforming abstract control commands into a specific number of discrete execution units; further, the real-time capacitor voltage data of all submodules within the arm are fully sorted to obtain a global voltage ranking sequence, providing a data foundation for globally optimal voltage equalization and ensuring the transparency and controllability of voltage distribution; then, the charging and discharging requirements are determined based on the current direction of the current arm. If charging is required, the required number of submodules with the lowest capacitor voltage values ​​are selected from the global voltage ranking sequence. The submodules are marked as "on" and their voltage is increased by absorbing energy from the low-voltage module, which conforms to the principle of capacitor charging equalization. If discharge is required, several submodules with the highest capacitor voltage values ​​are selected from the global voltage sorting sequence and marked as "on". The high-voltage module releases energy to reduce their voltage, which conforms to the principle of capacitor discharge equalization. Finally, the unselected submodules are marked as "bypass". This obtains the switching status data of each submodule in the current control cycle. Through optimal matching in the global range, the rapid convergence and high balance of capacitor voltages of all submodules in the bridge arm are achieved, ensuring the minimization of DC side voltage ripple and the high quality of AC side output waveform.

[0015] In some embodiments of the first aspect of this application, generating drive pulse signals to control the operation of the converter based on the switching state data of each submodule in the current control cycle includes: Based on the switching status data of each submodule in the current control cycle, generate the corresponding pulse width modulation (PWM) drive signal; The PWM drive signal is sent to the gate drive unit of each submodule to control the on and off of the insulated gate bipolar transistor (IGBT) in order to control the operation of the converter.

[0016] Compared with the prior art, the above embodiments have the following beneficial effects: by generating corresponding pulse width modulation (PWM) drive signals based on the switching state data of each submodule in the current control cycle, the conversion from digital logic level to analog drive waveform is completed, ensuring the physical executability of control commands; furthermore, the PWM drive signals are sent to the gate drive units of each submodule to control the conduction and turn-off of the insulated gate bipolar transistor (IGBT) to control the operation of the converter, realizing precise regulation of power flow, ensuring that the converter can stably output power according to the preset voltage equalization strategy and thermal protection logic, and completing the final execution link of the entire control closed loop.

[0017] Secondly, the present invention also provides a voltage equalization control system for a modular multilevel converter, comprising: a data acquisition module, a calculation module, an adjustment module, and a control module; The data acquisition module is used to acquire the real-time operating electrical quantities and switching status signals of each sub-module in the modular multilevel converter. The calculation module is used to calculate the real-time junction temperature data of the power devices of each sub-module based on the real-time operating electrical quantities and switch status signals using a preset thermal network model, and to determine the corresponding control mode based on the real-time junction temperature data. The adjustment module is used to select, when the control mode is high temperature protection mode, target sub-modules that meet the voltage conditions from the sub-modules in either the bypass state or the active state according to the current modulation requirements, and switch them on or off to obtain the switching status data of each sub-module in the current control cycle; otherwise, sort the capacitor voltages of all sub-modules and select the corresponding number of sub-modules to be activated according to the modulation requirements to obtain the switching status data of each sub-module in the current control cycle. The control module is used to generate drive pulse signals to control the operation of the converter based on the switching status data of each sub-module in the current control cycle.

[0018] Compared with the prior art, the above embodiments of this application have the following beneficial effects: They acquire real-time operating electrical quantities and switching status signals of each submodule in the converter, providing accurate underlying data support for subsequent thermal state assessment and control decisions; based on the real-time operating electrical quantities and switching status signals, they calculate the real-time junction temperature data of the power devices in each submodule, realizing online soft measurement of the internal thermal state of the power devices, avoiding the increased hardware cost and response lag problems caused by directly installing temperature sensors, and accurately capturing the transient temperature rise trend under overload conditions; based on the real-time junction temperature data, they determine the corresponding control mode. When the control mode is the high-temperature protection mode, according to the current modulation requirements, they select target submodules that meet the voltage conditions from the submodules in either the bypass state or the active state for activation or deactivation, obtaining the switching status data of each submodule in the current control cycle. This feature significantly reduces the time required by limiting the set of submodules participating in the sorting and action. By reducing unnecessary switching operations, switching losses are significantly reduced, effectively curbing further increases in junction temperature during overload periods and solving the thermal runaway problem caused by frequent switching in traditional voltage equalization strategies under heavy loads. Otherwise, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected for operation according to modulation requirements to obtain the switching status data of each sub-module in the current control cycle. This ensures that the system can maintain high-precision capacitor voltage balance under non-overheating conditions, guaranteeing the power quality on the grid side and the stability of the DC side voltage. Finally, based on the switching status data of each sub-module in the current control cycle, a drive pulse signal is generated to control the converter operation, realizing a closed loop from control algorithm to physical execution. The synergistic effect of these features, through the combination of "real-time junction temperature monitoring" and "dual-mode strategy switching," significantly improves the thermal survivability and operational reliability of the modular multilevel converter under short-term overload conditions without increasing hardware heat dissipation redundancy.

[0019] In some embodiments of the second aspect of this application, the calculation module includes: an extraction unit, a first calculation unit, a second calculation unit, an addition unit, a model calculation unit, and a value retrieval unit; The extraction unit is used to extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities. The first calculation unit is used to determine the type of power device in the conducting state in each submodule at the current moment based on the switch state signal, and calculate the transient conduction loss value of each power device by combining the bridge arm current data and preset device conduction characteristic parameters. The second calculation unit is used to identify the power device that is switching at the current moment based on the switching status signal, and calculate the transient switching loss value of each power device by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. The summing unit is used to add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device. The model calculation unit is used to input the total instantaneous power loss data as a heat source into a preset thermal resistance-capacitance network model, and obtain the estimated transient junction temperature of each power device through iterative calculation. The value-taking unit is used to take the maximum value among the estimated transient junction temperatures of all power devices in each submodule as the real-time junction temperature data of that submodule.

[0020] Compared to existing technologies, the above embodiments have the following advantages: They extract the current arm current data and capacitor voltage data of each submodule from real-time operating electrical quantities, clarifying the physical input variables for heat loss calculation. Furthermore, based on the switching status signal, they determine the type of power device in the conducting state of each submodule at the current moment. Combining the arm current data and preset device conduction characteristic parameters, they calculate the transient conduction loss value of each power device, achieving refined quantification of steady-state heat sources and reflecting the cumulative ohmic loss effect of the device during continuous conduction. Simultaneously, based on the switching status signal, they identify the power device that has switched at the current moment. Combining the capacitor voltage data, arm current data, and preset device switching characteristic parameters, they calculate the transient switching loss value of each power device, accurately capturing the dynamic heating impact caused by state transitions, and compensating for the temperature rise estimation deviation caused by only considering conduction losses. By adding the transient conduction loss value and transient switching loss value of each power device, the total transient power loss data of each power device is obtained, and a complete transient heat source model is constructed, providing accurate power excitation for thermal network iteration. Then, the total transient power loss data is used as a heat source input into a preset thermal resistance-capacitance network model, and the transient junction temperature estimate of each power device is obtained through iterative calculation. The physical process of heat transfer and accumulation inside the device is simulated by utilizing the inertial characteristics of the thermal resistance-capacitance model, realizing high-precision reconstruction of the chip junction temperature that cannot be directly measured. Finally, for each submodule, the maximum value of the transient junction temperature estimate of all power devices in the submodule is taken as the real-time junction temperature data of the submodule. Following the "weakest link" principle, the worst thermal state is used as the control basis, ensuring that the protection mechanism can be triggered in time even under the most dangerous operating conditions, thus improving the thermal safety margin of the system.

[0021] Thirdly, the present invention also provides a computer program product, including a computer program or instructions, which, when executed, implement the voltage equalization control method of any modular multilevel converter of the present invention.

[0022] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the voltage equalization control method for any modular multilevel converter of the present invention. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a voltage equalization control method for a modular multilevel converter provided in some embodiments of the present invention.

[0024] Figure 2 This is a schematic diagram of the voltage equalization control system of a modular multilevel converter provided in some embodiments of the present invention.

[0025] Figure 3 This is a schematic diagram of the main circuit topology and half-bridge submodule structure of a modular multilevel converter provided in some embodiments of the present invention.

[0026] Figure 4 This is a schematic diagram of a hysteresis switching and delay recovery mechanism provided in some embodiments of the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Example 1: refer to Figure 3 The diagram illustrates the main circuit topology and half-bridge submodule structure of a modular multilevel converter (MMC). The standard MMC's main circuit topology is shown in the figure. This MMC consists of three phases (or single phases), each phase composed of upper and lower bridge arms. Each bridge arm contains several series-connected half-bridge (or full-bridge) submodules and bridge arm reactors. The internal circuitry of each submodule consists of fully controlled power switching devices and energy storage capacitors. Under overload conditions, the current stress flowing through the MMC bridge arms increases significantly, leading to a sharp rise in the conduction and switching losses of the power semiconductor switching devices (such as IGBTs and their anti-parallel diodes) within the submodules. Limited by the physical characteristics of the devices themselves and the thermal inertia of the heat dissipation system, the surge in overload losses causes a rapid increase in the device junction temperature. When the junction temperature approaches or exceeds the device's allowable safety limit (e.g., 150°C), thermal breakdown is highly likely to occur, leading to complete damage to the converter valve. Therefore, transient junction temperature exceeding the limit has become a core physical bottleneck restricting the improvement of MMC overload capacity.

[0029] Please refer to Figure 1 To address the problem of rapid temperature rise in power device junctions under short-term overload conditions due to a surge in switching losses, thus limiting the overload capacity of modular multilevel converters, an embodiment of the present invention provides a voltage equalization control method for modular multilevel converters, comprising steps S1 to S4: Step S1: Obtain the real-time operating electrical quantities and switching status signals of each submodule in the modular multilevel converter; In the actual operation of modular multilevel converters, the power semiconductor devices (such as IGBTs and their anti-parallel diodes) inside the submodules are the main bearers of thermal stress. Traditional control strategies often only focus on capacitor voltage balance or grid-side current tracking, neglecting real-time sensing of the junction temperature inside the devices, or relying on temperature sensors mounted on the heat sink surface for feedback. However, due to the significant thermal resistance and thermal capacitance between the semiconductor chip and the heat sink, changes in surface temperature have a large lag, failing to reflect the drastic changes in transient temperature rise inside the chip under overload conditions in a timely manner, which can easily lead to thermal breakdown of the device before the controller can react.

[0030] Therefore, this application aims to provide fundamental data support for the subsequent construction of a high-precision "soft measurement" thermal model by acquiring real-time operating electrical quantities and switch status signals. The real-time operating electrical quantities mainly include physical quantities that directly reflect the power flow state, such as bridge arm current and submodule port voltage, while the switch status signals record which specific switch in each submodule was in the on or off state at the previous moment. This data acquisition method eliminates the need for additional hardware temperature sensors, reducing system cost and failure rate. Furthermore, it enables online, high-frequency monitoring of device thermal status through digital means, laying a data foundation for subsequent refined control based on junction temperature feedback.

[0031] Step S2: Based on the real-time operating electrical quantities and switch status signals, calculate the real-time junction temperature data of the power devices in each sub-module using a preset thermal network model, and determine the corresponding control mode based on the real-time junction temperature data. Furthermore, step S2 can be implemented through the following preferred embodiments, including steps S21-S27: S21: Extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities; S22: Determine the type of power device in the conducting state in each submodule at the current moment based on the switch status signal, and calculate the transient conduction loss value of each power device by combining the bridge arm current data and the preset device conduction characteristic parameters. S23: Identify the power device that is switching at the current moment based on the switching status signal, and calculate the transient switching loss value of each power device by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. S24: Add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device; S25: Input the total instantaneous power loss data as a heat source into the preset thermal resistance-capacitance network model, and obtain the estimated value of the transient junction temperature of each power device through iterative calculation; S26: For each submodule, take the maximum value among the estimated transient junction temperatures of all power devices in that submodule as the real-time junction temperature data of that submodule.

[0032] Steps S21 to S26 above detail how to reconstruct the chip junction temperature, which is difficult to measure directly, using easily measurable electrical quantities through the "electrothermal analogy" method. The core of this process lies in establishing an accurate thermal network model. In actual engineering, the losses of power devices mainly consist of two parts: conduction losses and switching losses. Conduction losses depend on the magnitude of the current flowing through the device and the device's on-resistance, and are continuous; while switching losses occur at the instant of device state switching, are closely related to the switching frequency, DC bus voltage, and load current, and have pulse characteristics. By calculating these two parts of the loss separately and superimposing them, the total heat source power generated by the device at the current moment can be obtained. Subsequently, this heat source power is input into a preset thermal resistance-capacitance network model (usually a Foster or Cauer equivalent circuit), which can simulate the physical process of heat transfer from the chip junction region through the packaging material and insulating layer to the heat sink. Due to the existence of thermal capacitance, temperature changes have inertia, so the current transient junction temperature can be accurately estimated through iterative calculation.

[0033] Specifically, a half-bridge submodule contains multiple switching devices (such as T1, T2, D1, D2), and the heat generation of each device varies under different current directions and switching states. To ensure system safety, this embodiment adopts the "worst-case principle," that is, taking the maximum value among the estimated junction temperatures of all devices within the submodule as the representative junction temperature of that submodule. Although this approach is conservative, it effectively avoids overall module failure due to overheating of individual devices, providing a reliable safety criterion for subsequent mode switching.

[0034] In this preferred embodiment, the current arm current data and capacitor voltage data of each submodule are extracted from real-time operating electrical quantities to clarify the physical input variables for heat loss calculation. Furthermore, based on the switch status signal, the type of power device in the conducting state of each submodule at the current moment is determined. Combined with the arm current data and preset device conduction characteristic parameters, the transient conduction loss value of each power device is calculated, achieving refined quantification of the steady-state heat source and reflecting the cumulative ohmic loss effect of the device during continuous conduction. Simultaneously, based on the switch status signal, the power device that has switched at the current moment is identified. Combined with capacitor voltage data, arm current data, and preset device switching characteristic parameters, the transient switching loss value of each power device is calculated, accurately capturing the dynamic heating impact caused by state transitions and compensating for the temperature rise estimation deviation caused by only considering conduction losses. By combining the power... The transient conduction loss and transient switching loss of each power device are added together to obtain the total transient power loss data of each power device, thus constructing a complete transient heat source model and providing accurate power excitation for thermal network iteration. This total transient power loss data is then input as a heat source into a preset thermal resistance-capacitor (RTC) network model. Through iterative calculation, the transient junction temperature estimate of each power device is obtained. The inertial characteristics of the RTC model are used to simulate the physical process of heat transfer and accumulation within the device, achieving high-precision reconstruction of the chip junction temperature, which cannot be directly measured. Finally, for each submodule, the maximum value among all transient junction temperature estimates of all power devices within that submodule is taken as the real-time junction temperature data of that submodule. Following the "weakest link" principle, the worst thermal state is used as the control basis, ensuring that the protection mechanism can be triggered in a timely manner even under the most dangerous operating conditions, thus improving the system's thermal safety margin.

[0035] S27: Determine whether the real-time junction temperature data is greater than or equal to the preset upper temperature threshold. If the determination result is yes, then determine the high temperature protection mode; otherwise, determine whether the real-time junction temperature data is less than the preset lower temperature threshold. If the duration for which the real-time junction temperature data is less than the preset lower temperature threshold is greater than the preset delay time, then determine the normal pressure equalization mode; otherwise, keep the control mode of the previous control cycle unchanged.

[0036] In existing technologies, thermal management of MMCs typically employs a fixed voltage equalization strategy, such as Nearest Level Modulation (NLM) combined with a global sorting algorithm. Under this conventional strategy, to maintain a high degree of capacitor voltage balance, the controller reorders the voltages of all submodules in each control cycle and frequently switches between connected and bypassed submodule combinations. While this method achieves excellent voltage balance, frequent switching operations under short-term overload conditions lead to a sharp increase in switching losses. Because the current amplitude flowing through the devices is large during overload, and switching losses are positively correlated with current, this causes the device junction temperature to rise rapidly. If a high-frequency switching strategy based on global sorting is still used under these conditions, the device temperature will quickly exceed the safety limit (e.g., 150°C), forcing the converter to operate at a derating rate or even shut down for protection, severely limiting the overload capacity of the MMC.

[0037] Furthermore, while some existing improvement schemes have introduced temperature weighting factors in an attempt to balance temperature during sorting, their essence remains global optimization across all sub-modules. This does not fundamentally reduce the number of switching actions, and under normal operating conditions, excessive focus on temperature balance may sacrifice the dynamic response speed of voltage balance.

[0038] Therefore, this application proposes a dual-mode switching control approach, which dynamically adjusts the control strategy based on the junction temperature data obtained in real time: when the temperature is low, priority is given to ensuring voltage quality, and when the temperature is too high, priority is given to reducing switching losses to protect the device, thereby solving the technical contradiction that a single strategy cannot simultaneously achieve "high-precision voltage equalization" and "high thermal survivability".

[0039] In addition, refer to Figure 4The diagram illustrates a hysteresis switching and delay recovery mechanism. This embodiment also introduces a dual-threshold comparison mechanism with hysteresis characteristics to address the "ping-pong effect" problem easily caused by single-threshold control. If only a single upper temperature threshold is set, the control system may frequently switch between "high-temperature protection mode" and "normal voltage equalization mode" when the junction temperature fluctuates slightly around the threshold. This frequent strategy switching not only leads to disordered oscillations in the switching states of submodules, increasing additional switching losses, but may also cause distortion of the bridge arm output voltage, affecting the power quality of the grid. By setting an upper temperature threshold (e.g., 110℃) and a lower temperature threshold (e.g., 95℃), combined with delay judgment logic, the system forms a stable hysteresis range. When the temperature exceeds the upper limit, the system decisively enters the high-temperature protection mode, sacrificing some voltage equalization accuracy for a lower switching frequency; only when the temperature drops sufficiently and falls below the lower limit, and remains below it for a certain period of time (ensuring that the internal heat of the device has been fully dissipated), does it switch back to the normal voltage equalization mode. This design ensures smooth and stable control mode switching, avoids malfunctions caused by noise interference or short-term load fluctuations, improves the robustness of the system under dynamic operating conditions, and enables the converter to maintain high-performance operation for as long as possible while ensuring safety.

[0040] In this preferred embodiment, by determining whether the real-time junction temperature data is greater than or equal to the upper temperature threshold, if the determination result is yes, the high-temperature protection mode is determined, realizing an immediate response to overheating risks and ensuring that the device quickly switches to a low-power control strategy to prevent thermal breakdown when it approaches the safety limit; otherwise, it is determined whether the real-time junction temperature data is less than the lower temperature threshold. If the duration of the real-time junction temperature data being less than the lower temperature threshold is greater than the preset delay time, the conventional voltage equalization mode is determined, introducing a time-dimensional stability criterion to prevent frequent oscillations of the control mode caused by temperature measurement noise or short-term fluctuations, ensuring the smoothness of system state switching; if none of the above conditions are met, the control mode of the previous control cycle remains unchanged, constructing a hysteresis control logic, maintaining the continuity of the control strategy in the transition zone between the upper and lower temperature limits, avoiding the ping-pong effect near the critical point, thereby improving the control stability and robustness of the converter during dynamic load changes.

[0041] Step S3: When the control mode is high temperature protection mode, according to the current modulation requirements, select the target sub-module that meets the voltage conditions from the sub-modules that are in either the bypass state or the active state, and put it into or cut it out to obtain the switching status data of each sub-module in the current control cycle; otherwise, sort the capacitor voltages of all sub-modules, and select the corresponding number of sub-modules to be put into operation according to the modulation requirements to obtain the switching status data of each sub-module in the current control cycle. Furthermore, the high-temperature protection mode portion of step S3 can be implemented through the following preferred embodiments, including steps S31-S33: S31: Calculate the total number of sub-modules that need to be put into operation in the bridge arm of the converter at the current moment, based on the nearest level approximation modulation algorithm; S32: Subtract the total number of sub-modules that need to be deployed in the bridge arm at the current moment from the total number of sub-modules actually deployed in the previous control cycle to obtain the quantity difference; S33: If the quantity difference is positive, select the sub-modules currently in bypass state according to the order of capacitor voltage from low to high, and change their state to active state, while keeping the states of the remaining sub-modules in bypass state and all sub-modules in active state unchanged; if the quantity difference is negative, select the sub-modules currently in active state according to the order of capacitor voltage from high to low, and change their state to bypass state, while keeping the states of the remaining sub-modules in active state and all sub-modules in bypass state unchanged; if the quantity difference is zero, keep the switching state of all sub-modules consistent with the previous control cycle.

[0042] In the high-temperature protection mode, the "incremental switching" or "state maintenance" strategy adopted in this embodiment is based on the core principle of maximizing the reuse of the switching states of the previous cycle. Unlike the conventional mode, which reorders all quantum modules every time, this strategy first calculates the difference between the total number of sub-modules that need to be put into operation according to modulation requirements at the current moment and the actual total number put into operation at the previous moment (i.e., the quantity difference). If the difference is zero, it means that the grid voltage reference value has not changed significantly. At this time, all sub-module states are kept unchanged, the number of switching actions is zero, and the switching loss is theoretically reduced to the minimum. If the difference is positive, it means that more modules need to be put into operation. The system only performs local screening in the set of modules in the "bypass" state, selecting the module with the lowest voltage to be put into operation, while the modules that were originally in the "put into operation" state continue to be conducted without going through the "turn-off-turn-on" process. Similarly, if the difference is negative, only the module with the highest voltage in the "put into operation" state is selected to be cut off. This local optimization method reduces the number of modules involved in sorting and action from "all" to "partial", which greatly reduces the switching frequency. Although this may result in a slightly larger voltage deviation between submodules than in the global sorting mode, the primary goal is to suppress temperature rise and prevent device burnout under emergency conditions of overload and overheating. Slight voltage fluctuations are well within the tolerance range of the MMC system, thus achieving a reasonable balance between thermal safety and electrical performance. This effectively overcomes the bottleneck of thermal runaway caused by frequent switching under heavy loads in traditional voltage equalization strategies.

[0043] In this preferred embodiment, the total number of submodules that need to be engaged in the bridge arm of the converter at the current moment is calculated according to the nearest-level approximation modulation algorithm, establishing the basic constraint condition to meet the grid voltage following requirement. Furthermore, the total number of submodules that need to be engaged in the bridge arm at the current moment is subtracted from the total number of submodules actually engaged in the previous control cycle to obtain the quantity difference, quantifying the state change requirements of the current control cycle relative to the previous cycle, providing a calculation benchmark for local optimization. If the quantity difference is positive, the submodules currently in the bypass state are sorted in ascending order of capacitor voltage, and the top few submodules with the largest quantity difference are selected and their states are changed to the engaged state. Simultaneously, the states of the remaining bypass submodules and all engaged submodules remain unchanged. This incremental engagement strategy ensures charging balance (prioritizing low voltage) while maximizing the retention of the conduction of the original engaged modules. The system reduces switching losses caused by the turn-on action. If the quantity difference is negative, the sub-modules currently in the active state are sorted in descending order of capacitor voltage. The sub-modules with the absolute value of the aforementioned quantity difference are selected and their states are changed to bypass state. At the same time, the states of the remaining active sub-modules and all sub-modules in the bypass state remain unchanged. This reduction-type cut-off strategy ensures discharge balance (prioritizing high voltage cut-off) while avoiding large-scale state rearrangement, further suppressing the heat generated by the turn-off action. If the quantity difference is zero, the switching state of all sub-modules is kept consistent with the previous control cycle, achieving an ideal low-power state with zero switching action. Overall, through the "state maintenance + local screening" mechanism, the switching frequency is reduced to the theoretical minimum in high-temperature protection mode, directly serving the core technical objective of suppressing rapid junction temperature rise.

[0044] Furthermore, the non-high-temperature protection mode portion of step S3 can be implemented through the following preferred embodiments, including steps S34-S37: S34: Based on the modulation requirements and the rated voltage of the sub-modules, calculate the number of sub-modules required to be put into operation in the bridge arm of the converter during the current control cycle; S35: Perform a full sorting of the real-time capacitor voltage data of all sub-modules within the bridge arm to obtain a global voltage sorting sequence; S36: Determine the charging and discharging requirements based on the current direction of the current in the current arm. If charging is required, select several sub-modules with the lowest capacitor voltage value from the global voltage sorting sequence and mark them as in operation. If discharging is required, select several sub-modules with the highest capacitor voltage value from the global voltage sorting sequence and mark them as in operation. S37: Mark the unselected submodules as bypass state to obtain the switching status data of each submodule in the current control cycle.

[0045] When the system is in non-high-temperature protection mode (i.e., conventional voltage equalization mode), the control objective returns to the ultimate pursuit of grid power quality. At this time, the junction temperature of the devices is within a safe range, and switching losses are no longer a bottleneck restricting system operation. Therefore, a traditional global sorting algorithm can be used to achieve optimal capacitor voltage equalization. This strategy performs a full sorting of the capacitor voltages of all sub-modules within the bridge arm, regardless of their previous state, and reselects the optimal module combination based on the current charging and discharging requirements (determined by the direction of the bridge arm current). For example, during the charging phase, the module with the lowest voltage is prioritized for operation, allowing it to quickly absorb energy; during the discharging phase, the module with the highest voltage is prioritized for operation, allowing it to quickly release energy. This globally optimal matching ensures that the voltages of all sub-modules converge rapidly to near their rated values ​​within a short time, suppressing DC-side voltage ripple and AC-side output harmonics, and guaranteeing the efficient and stable operation of the converter under normal operating conditions. By flexibly switching between these two modes, this solution retains the advantages of traditional MMC high-precision pressure equalization while also giving it thermal survivability under extreme conditions. This can significantly improve the overall operational reliability of the equipment and resolve the technical contradiction in existing technologies where a single strategy cannot simultaneously achieve both "high-precision pressure equalization" and "high thermal survivability".

[0046] In this preferred embodiment, based on modulation requirements and the rated voltage of the submodules, the number of submodules required to be engaged in the converter arm during the current control cycle is calculated, transforming abstract control commands into a specific number of discrete execution units. Furthermore, the real-time capacitor voltage data of all submodules within the arm are fully sorted to obtain a global voltage ranking sequence, providing a data foundation for globally optimal voltage equalization and ensuring the transparency and controllability of voltage distribution. Next, the charging and discharging requirements are determined based on the current direction of the current arm. If charging is required, the submodules with the lowest capacitor voltage values ​​are selected from the global voltage ranking sequence and marked as engaged. In the current control cycle, the submodules with the highest capacitor voltage values ​​are selected from the global voltage ranking sequence and marked as in operation. The high-voltage modules then release energy to lower their voltage, conforming to the capacitor discharge balancing principle. Finally, the unselected submodules are marked as bypassed, thus obtaining the switching status data of each submodule in the current control cycle. Through optimal matching within the global range, rapid convergence and high balance of capacitor voltages in all submodules within the bridge arm are achieved, ensuring minimal DC-side voltage ripple and high-quality AC-side output waveform.

[0047] Step S4: Generate a drive pulse signal to control the operation of the converter based on the switching status data of each sub-module in the current control cycle.

[0048] Furthermore, step S4 can be implemented through the following preferred embodiments, including steps S41-S42: S41: Generate the corresponding pulse width modulation (PWM) drive signal based on the switching status data of each submodule in the current control cycle; S42: Send the PWM drive signal to the gate drive unit of each submodule to control the on and off of the insulated gate bipolar transistor IGBT, so as to control the operation of the converter.

[0049] In this preferred embodiment, by generating corresponding pulse width modulation (PWM) drive signals based on the switching state data of each submodule in the current control cycle, the conversion from digital logic level to analog drive waveform is completed, ensuring the physical executability of the control commands. Furthermore, the PWM drive signals are sent to the gate drive units of each submodule to control the on and off of the insulated gate bipolar transistor (IGBT) to control the operation of the converter, realizing precise regulation of power flow and ensuring that the converter can stably output power according to the preset voltage equalization strategy and thermal protection logic, thus completing the final execution stage of the entire control closed loop.

[0050] In summary, compared with the prior art, the above embodiments of this application have the following beneficial effects: They acquire real-time operating electrical quantities and switching status signals of each submodule in the converter, providing accurate underlying data support for subsequent thermal state assessment and control decisions; based on the real-time operating electrical quantities and switching status signals, they calculate the real-time junction temperature data of the power devices in each submodule, realizing online soft measurement of the internal thermal state of the power devices, avoiding the increased hardware cost and response lag problems caused by directly installing temperature sensors, and accurately capturing the transient temperature rise trend under overload conditions; based on the real-time junction temperature data, they determine the corresponding control mode. When the control mode is the high-temperature protection mode, according to the current modulation requirements, they select target submodules that meet the voltage conditions from the submodules in either the bypass state or the active state for activation or deactivation, obtaining the switching status data of each submodule in the current control cycle. This feature, by limiting the set of submodules participating in the sorting and action, significantly improves efficiency. By reducing unnecessary switching operations, switching losses are significantly reduced, effectively curbing further increases in junction temperature during overload periods and solving the thermal runaway problem caused by frequent switching under heavy loads in traditional voltage equalization strategies. Otherwise, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected for operation according to modulation requirements to obtain the switching status data of each sub-module in the current control cycle. This ensures that the system can maintain high-precision capacitor voltage balance under non-overheating conditions, guaranteeing power quality on the grid side and DC voltage stability. Finally, based on the switching status data of each sub-module in the current control cycle, a drive pulse signal is generated to control the converter operation, realizing a closed loop from control algorithm to physical execution. The synergistic effect of these features, through the combination of "real-time junction temperature monitoring" and "dual-mode strategy switching," significantly improves the thermal survivability and operational reliability of the modular multilevel converter under short-term overload conditions without increasing hardware heat dissipation redundancy.

[0051] Example 2: Please refer to Figure 2 Based on the same inventive concept, the present invention discloses a voltage equalization control system for a modular multilevel converter, comprising: a data acquisition module M1, a calculation module M2, an adjustment module M3, and a control module M4; The data acquisition module M1 is used to acquire the real-time operating electrical quantities and switching status signals of each sub-module in the modular multilevel converter. The calculation module M2 is used to calculate the real-time junction temperature data of the power devices of each sub-module based on the real-time operating electrical quantities and switch status signals using a preset thermal network model, and to determine the corresponding control mode based on the real-time junction temperature data. The calculation module M2 includes: an extraction unit, a first calculation unit, a second calculation unit, an addition unit, a model calculation unit, and a value retrieval unit; The extraction unit is used to extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities. The first calculation unit is used to determine the type of power device in the conducting state in each submodule at the current moment based on the switch state signal, and calculate the transient conduction loss value of each power device by combining the bridge arm current data and preset device conduction characteristic parameters. The second calculation unit is used to identify the power device that is switching at the current moment based on the switching status signal, and calculate the transient switching loss value of each power device by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. The summing unit is used to add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device. The model calculation unit is used to input the total instantaneous power loss data as a heat source into a preset thermal resistance-capacitance network model, and obtain the estimated transient junction temperature of each power device through iterative calculation. The value-taking unit is used to take the maximum value among the estimated transient junction temperatures of all power devices in each submodule as the real-time junction temperature data of that submodule.

[0052] In this preferred embodiment, the current arm current data and capacitor voltage data of each submodule are extracted from real-time operating electrical quantities to clarify the physical input variables for heat loss calculation. Furthermore, based on the switch status signal, the type of power device in the conducting state of each submodule at the current moment is determined. Combined with the arm current data and preset device conduction characteristic parameters, the transient conduction loss value of each power device is calculated, achieving refined quantification of the steady-state heat source and reflecting the cumulative ohmic loss effect of the device during continuous conduction. Simultaneously, based on the switch status signal, the power device that has switched at the current moment is identified. Combined with capacitor voltage data, arm current data, and preset device switching characteristic parameters, the transient switching loss value of each power device is calculated, accurately capturing the dynamic heating impact caused by state transitions and compensating for the temperature rise estimation deviation caused by only considering conduction losses. By combining the power... The transient conduction loss and transient switching loss of each power device are added together to obtain the total transient power loss data of each power device, thus constructing a complete transient heat source model and providing accurate power excitation for thermal network iteration. This total transient power loss data is then input as a heat source into a preset thermal resistance-capacitor (RTC) network model. Through iterative calculation, the transient junction temperature estimate of each power device is obtained. The inertial characteristics of the RTC model are used to simulate the physical process of heat transfer and accumulation within the device, achieving high-precision reconstruction of the chip junction temperature, which cannot be directly measured. Finally, for each submodule, the maximum value among all transient junction temperature estimates of all power devices within that submodule is taken as the real-time junction temperature data of that submodule. Following the "weakest link" principle, the worst thermal state is used as the control basis, ensuring that the protection mechanism can be triggered in a timely manner even under the most dangerous operating conditions, thus improving the system's thermal safety margin.

[0053] Furthermore, the computing module M2 also includes: a mode determination unit; The mode determination unit is used to determine whether the real-time junction temperature data is greater than or equal to a preset upper temperature threshold. If the determination result is yes, it is determined to be a high-temperature protection mode; otherwise, it determines whether the real-time junction temperature data is less than a preset lower temperature threshold. If the duration for which the real-time junction temperature data is less than the preset lower temperature threshold is greater than a preset delay time, it is determined to be a normal pressure equalization mode; otherwise, the control mode of the previous control cycle remains unchanged.

[0054] In this preferred embodiment, by determining whether the real-time junction temperature data is greater than or equal to the upper temperature threshold, if the determination result is yes, the high-temperature protection mode is determined, realizing an immediate response to overheating risks and ensuring that the device quickly switches to a low-power control strategy to prevent thermal breakdown when it approaches the safety limit; otherwise, it is determined whether the real-time junction temperature data is less than the lower temperature threshold. If the duration of the real-time junction temperature data being less than the lower temperature threshold is greater than the preset delay time, the conventional voltage equalization mode is determined, introducing a time-dimensional stability criterion to prevent frequent oscillations of the control mode caused by temperature measurement noise or short-term fluctuations, ensuring the smoothness of system state switching; if none of the above conditions are met, the control mode of the previous control cycle remains unchanged, constructing a hysteresis control logic, maintaining the continuity of the control strategy in the transition zone between the upper and lower temperature limits, avoiding the ping-pong effect near the critical point, thereby improving the control stability and robustness of the converter during dynamic load changes.

[0055] The adjustment module M3 is used to select, when the control mode is high temperature protection mode, target sub-modules that meet the voltage conditions from the sub-modules in either the bypass state or the on state according to the current modulation requirements, and switch them on or off to obtain the switching status data of each sub-module in the current control cycle; otherwise, sort the capacitor voltages of all sub-modules and select the corresponding number of sub-modules to be on according to the modulation requirements to obtain the switching status data of each sub-module in the current control cycle. Furthermore, the adjustment module M3 includes: a first required quantity calculation unit, a difference unit, and a selection execution unit; The first required quantity calculation unit is used to calculate the total number of sub-modules that need to be put into the bridge arm of the converter at the current moment according to the nearest level approximation modulation algorithm. The difference unit is used to subtract the total number of sub-modules that the bridge arm needs to be put into operation at the current moment from the total number of sub-modules actually put into operation in the previous control cycle to obtain the quantity difference. The selection execution unit is configured to: if the quantity difference is positive, select the sub-modules currently in bypass state according to the order of capacitor voltage from low to high, select the sub-modules with the aforementioned quantity difference, change their state to active state, and keep the states of the remaining sub-modules in bypass state and all sub-modules in active state unchanged; if the quantity difference is negative, select the sub-modules currently in active state according to the order of capacitor voltage from high to low, select the sub-modules with the absolute value of the aforementioned quantity difference, change their state to bypass state, and keep the states of the remaining sub-modules in active state and all sub-modules in bypass state unchanged; if the quantity difference is zero, keep the switching state of all sub-modules consistent with the previous control cycle.

[0056] In this preferred embodiment, the total number of submodules that need to be engaged in the bridge arm of the converter at the current moment is calculated according to the nearest-level approximation modulation algorithm, establishing the basic constraint condition to meet the grid voltage following requirement. Furthermore, the total number of submodules that need to be engaged in the bridge arm at the current moment is subtracted from the total number of submodules actually engaged in the previous control cycle to obtain the quantity difference, quantifying the state change requirements of the current control cycle relative to the previous cycle, providing a calculation benchmark for local optimization. If the quantity difference is positive, the submodules currently in the bypass state are sorted in ascending order of capacitor voltage, and the top few submodules with the largest quantity difference are selected and their states are changed to the engaged state. Simultaneously, the states of the remaining bypass submodules and all engaged submodules remain unchanged. This incremental engagement strategy ensures charging balance (prioritizing low voltage) while maximizing the retention of the conduction of the original engaged modules. The system reduces switching losses caused by the turn-on action. If the quantity difference is negative, the sub-modules currently in the active state are sorted in descending order of capacitor voltage. The sub-modules with the absolute value of the aforementioned quantity difference are selected and their states are changed to bypass state. At the same time, the states of the remaining active sub-modules and all sub-modules in the bypass state remain unchanged. This reduction-type cut-off strategy ensures discharge balance (prioritizing high voltage cut-off) while avoiding large-scale state rearrangement, further suppressing the heat generated by the turn-off action. If the quantity difference is zero, the switching state of all sub-modules is kept consistent with the previous control cycle, achieving an ideal low-power state with zero switching action. Overall, through the "state maintenance + local screening" mechanism, the switching frequency is reduced to the theoretical minimum in high-temperature protection mode, directly serving the core technical objective of suppressing rapid junction temperature rise.

[0057] Furthermore, the adjustment module M3 also includes: a second required quantity calculation unit, an arrangement unit, a first marking unit, and a second marking unit; The second required quantity calculation unit is used to calculate the number of sub-modules required to be put into operation by the bridge arm of the converter in the current control cycle based on the modulation requirements and the rated voltage of the sub-modules. The arrangement unit is used to perform a full arrangement of the real-time capacitor voltage data of all sub-modules in the bridge arm to obtain a global voltage sorting sequence. The first marking unit is used to determine the charging and discharging requirements based on the current direction of the current in the current arm. If charging is required, several sub-modules with the lowest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. If discharging is required, several sub-modules with the highest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. The second marking unit is used to mark the unselected submodules as bypass states, thereby obtaining the switching status data of each submodule in the current control cycle.

[0058] In this preferred embodiment, based on modulation requirements and the rated voltage of the submodules, the number of submodules required to be engaged in the converter arm during the current control cycle is calculated, transforming abstract control commands into a specific number of discrete execution units. Furthermore, the real-time capacitor voltage data of all submodules within the arm are fully sorted to obtain a global voltage ranking sequence, providing a data foundation for globally optimal voltage equalization and ensuring the transparency and controllability of voltage distribution. Next, the charging and discharging requirements are determined based on the current direction of the current arm. If charging is required, the submodules with the lowest capacitor voltage values ​​are selected from the global voltage ranking sequence and marked as engaged. In the current control cycle, the submodules with the highest capacitor voltage values ​​are selected from the global voltage ranking sequence and marked as in operation. The high-voltage modules then release energy to lower their voltage, conforming to the capacitor discharge balancing principle. Finally, the unselected submodules are marked as bypassed, thus obtaining the switching status data of each submodule in the current control cycle. Through optimal matching within the global range, rapid convergence and high balance of capacitor voltages in all submodules within the bridge arm are achieved, ensuring minimal DC-side voltage ripple and high-quality AC-side output waveform.

[0059] The control module M4 is used to generate drive pulse signals to control the operation of the converter based on the switching status data of each sub-module in the current control cycle.

[0060] Furthermore, the control module M4 includes: a signal generation unit and a driving unit; The signal generation unit is used to generate a corresponding pulse width modulation (PWM) drive signal based on the switching state data of each submodule in the current control cycle. The driving unit is used to send the PWM driving signal to the gate driving unit of each submodule to control the on and off of the insulated gate bipolar transistor (IGBT) in order to control the operation of the converter.

[0061] In this preferred embodiment, by generating corresponding pulse width modulation (PWM) drive signals based on the switching state data of each submodule in the current control cycle, the conversion from digital logic level to analog drive waveform is completed, ensuring the physical executability of the control commands. Furthermore, the PWM drive signals are sent to the gate drive units of each submodule to control the on and off of the insulated gate bipolar transistor (IGBT) to control the operation of the converter, realizing precise regulation of power flow and ensuring that the converter can stably output power according to the preset voltage equalization strategy and thermal protection logic, thus completing the final execution stage of the entire control closed loop.

[0062] In summary, compared with the prior art, the embodiments of this application have the following beneficial effects: They acquire real-time operating electrical quantities and switching status signals of each submodule in the converter, providing accurate underlying data support for subsequent thermal state assessment and control decisions; based on the real-time operating electrical quantities and switching status signals, they calculate the real-time junction temperature data of the power devices in each submodule, realizing online soft measurement of the internal thermal state of the power devices, avoiding the increased hardware cost and response lag problems caused by directly installing temperature sensors, and accurately capturing the transient temperature rise trend under overload conditions; based on the real-time junction temperature data, they determine the corresponding control mode. When the control mode is the high-temperature protection mode, according to the current modulation requirements, they select target submodules that meet the voltage conditions from the submodules in either the bypass state or the active state for activation or deactivation, obtaining the switching status data of each submodule in the current control cycle. This feature significantly reduces the time required by limiting the set of submodules participating in the sorting and action. By reducing unnecessary switching operations, switching losses are significantly reduced, effectively curbing further increases in junction temperature during overload periods and solving the thermal runaway problem caused by frequent switching in traditional voltage equalization strategies under heavy loads. Otherwise, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected for operation according to modulation requirements to obtain the switching status data of each sub-module in the current control cycle. This ensures that the system can maintain high-precision capacitor voltage balance under non-overheating conditions, guaranteeing the power quality on the grid side and the stability of the DC side voltage. Finally, based on the switching status data of each sub-module in the current control cycle, a drive pulse signal is generated to control the converter operation, realizing a closed loop from control algorithm to physical execution. The synergistic effect of these features, through the combination of "real-time junction temperature monitoring" and "dual-mode strategy switching," significantly improves the thermal survivability and operational reliability of the modular multilevel converter under short-term overload conditions without increasing hardware heat dissipation redundancy.

[0063] Example 3: This invention also provides a computer program product, including a computer program or instructions, capable of running on a computing device or stored in any available medium. When the computer program product is run on at least one computing device, it causes the at least one computing device to execute any of the voltage equalization control methods of the modular multilevel converter of this invention.

[0064] Example 4: This invention also provides a computer-readable storage medium storing at least one executable instruction that, when executed on the voltage equalization control system of a modular multilevel converter, causes the voltage equalization control system of the modular multilevel converter to perform a voltage equalization control method for a modular multilevel converter in any of the above method embodiments.

[0065] The specific working processes of each module described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. The division of modules is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple modules may be combined or integrated into another system.

[0066] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A voltage equalization control method for a modular multilevel converter, characterized in that, include: Acquire real-time operating electrical quantities and switching status signals of each submodule in the modular multilevel converter; Based on the real-time operating electrical quantities and switch status signals, the real-time junction temperature data of the power devices in each sub-module is calculated using a preset thermal network model, and the corresponding control mode is determined based on the real-time junction temperature data. When the control mode is high temperature protection mode, according to the current modulation requirements, only the target sub-module that meets the voltage conditions is selected from the sub-modules that are in either the bypass state or the on state to be put into or cut out, so as to obtain the switching status data of each sub-module in the current control cycle; otherwise, the capacitor voltages of all sub-modules are sorted, and a corresponding number of sub-modules are selected to be put into operation according to the modulation requirements, so as to obtain the switching status data of each sub-module in the current control cycle. Based on the switching status data of each submodule in the current control cycle, drive pulse signals are generated to control the operation of the converter.

2. The voltage equalization control method for a modular multilevel converter as described in claim 1, characterized in that, The step of calculating the real-time junction temperature data of the power devices in each submodule using a preset thermal network model based on the real-time operating electrical quantities and switch status signals includes: Extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities; The type of power device in the conducting state in each submodule is determined based on the switch status signal at the current moment, and the transient conduction loss value of each power device is calculated by combining the bridge arm current data and the preset device conduction characteristic parameters. The power device that is switching at the current moment is identified based on the switching status signal, and the transient switching loss value of each power device is calculated by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. Add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device; The total instantaneous power loss data is input as a heat source into a preset thermal resistance-capacitance network model, and the transient junction temperature estimate of each power device is obtained through iterative calculation. For each submodule, the maximum value among the estimated transient junction temperatures of all power devices within that submodule is taken as the real-time junction temperature data for that submodule.

3. The voltage equalization control method for a modular multilevel converter as described in claim 1, characterized in that, The step of determining the corresponding control mode based on the real-time junction temperature data includes: Determine whether the real-time junction temperature data is greater than or equal to the preset upper temperature threshold. If the determination result is yes, then it is determined to be in high temperature protection mode; otherwise, determine whether the real-time junction temperature data is less than the preset lower temperature threshold. If the duration for which the real-time junction temperature data is less than the preset lower temperature threshold is greater than the preset delay time, then it is determined to be in normal pressure equalization mode; otherwise, the control mode of the previous control cycle remains unchanged.

4. The voltage equalization control method for a modular multilevel converter as described in claim 1, characterized in that, When the control mode is high-temperature protection mode, based on the current modulation requirements, a target submodule that meets the voltage conditions is selected from the submodules in either bypass or active state for activation or deactivation, thereby obtaining the switching status data of each submodule in the current control cycle, including: Based on the nearest level approximation modulation algorithm, calculate the total number of sub-modules that need to be put into operation in the bridge arm of the converter at the current moment; The difference in quantity is obtained by subtracting the total number of submodules that need to be deployed in the bridge arm at the current moment from the total number of submodules actually deployed in the previous control cycle. If the difference in quantity is positive, then from the sub-modules currently in the bypass state, sort them in order of capacitor voltage from low to high, select the sub-modules with the difference in quantity mentioned above, change their state to the engaged state, and keep the states of the remaining sub-modules in the bypass state and all sub-modules in the engaged state unchanged. If the quantity difference is negative, then from the sub-modules currently in the input state, sort them in descending order of capacitor voltage, select the sub-modules with the absolute value of the quantity difference mentioned above, change their state to bypass state, and keep the states of the remaining sub-modules in the input state and all sub-modules in the bypass state unchanged. If the difference in quantity is zero, then the switching state of all sub-modules remains consistent with the previous control cycle.

5. The voltage equalization control method for a modular multilevel converter as described in claim 1, characterized in that, The process involves sorting the capacitor voltages of all sub-modules and selecting a corresponding number of sub-modules for activation based on modulation requirements to obtain the switching status data of each sub-module in the current control cycle, including: Based on the modulation requirements and the rated voltage of the submodules, calculate the number of submodules required to be put into operation in the bridge arm of the converter during the current control cycle. The real-time capacitor voltage data of all sub-modules within the bridge arm are fully sorted to obtain a global voltage sorting sequence. The charging and discharging requirements are determined based on the current direction of the current in the current arm. If charging is required, several sub-modules with the lowest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. If discharging is required, several sub-modules with the highest capacitor voltage value are selected from the global voltage sorting sequence and marked as in operation. The unselected submodules are marked as bypassed, thereby obtaining the switching status data of each submodule in the current control cycle.

6. The voltage equalization control method for a modular multilevel converter as described in claim 1, characterized in that, The step of generating drive pulse signals to control the converter operation based on the switching status data of each submodule in the current control cycle includes: Based on the switching status data of each submodule in the current control cycle, generate the corresponding pulse width modulation (PWM) drive signal; The PWM drive signal is sent to the gate drive unit of each submodule to control the on and off of the insulated gate bipolar transistor (IGBT) in order to control the operation of the converter.

7. A voltage equalization control system for a modular multilevel converter, characterized in that, include: Data acquisition module, calculation module, adjustment module, and control module; The data acquisition module is used to acquire the real-time operating electrical quantities and switching status signals of each sub-module in the modular multilevel converter. The calculation module is used to calculate the real-time junction temperature data of the power devices of each sub-module based on the real-time operating electrical quantities and switch status signals using a preset thermal network model, and to determine the corresponding control mode based on the real-time junction temperature data. The adjustment module is used to select, when the control mode is high temperature protection mode, target sub-modules that meet the voltage conditions from the sub-modules in either the bypass state or the active state according to the current modulation requirements, and switch them on or off to obtain the switching status data of each sub-module in the current control cycle; otherwise, sort the capacitor voltages of all sub-modules and select the corresponding number of sub-modules to be activated according to the modulation requirements to obtain the switching status data of each sub-module in the current control cycle. The control module is used to generate drive pulse signals to control the operation of the converter based on the switching status data of each sub-module in the current control cycle.

8. The voltage equalization control system for a modular multilevel converter as described in claim 7, characterized in that, The calculation module includes: an extraction unit, a first calculation unit, a second calculation unit, an addition unit, a model calculation unit, and a value retrieval unit; The extraction unit is used to extract the current arm current data and the capacitor voltage data of each submodule from the real-time operating electrical quantities. The first calculation unit is used to determine the type of power device in the conducting state in each submodule at the current moment based on the switch state signal, and calculate the transient conduction loss value of each power device by combining the bridge arm current data and preset device conduction characteristic parameters. The second calculation unit is used to identify the power device that is switching at the current moment based on the switching status signal, and calculate the transient switching loss value of each power device by combining the capacitor voltage data, bridge arm current data and preset device switching characteristic parameters. The summing unit is used to add the transient conduction loss value and the transient switching loss value of each power device to obtain the total instantaneous power loss data of each power device. The model calculation unit is used to input the total instantaneous power loss data as a heat source into a preset thermal resistance-capacitance network model, and obtain the estimated transient junction temperature of each power device through iterative calculation. The value-taking unit is used to take the maximum value among the estimated transient junction temperatures of all power devices in each submodule as the real-time junction temperature data of that submodule.

9. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement a voltage equalization control method for a modular multilevel converter as described in any one of claims 1-6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a voltage equalization control method for a modular multilevel converter as described in any one of claims 1-6.

Images