Energy rebalancing control method for fault power unit of high-voltage cascaded grid-connection energy storage system

Abstract

The application discloses a high-voltage cascaded grid-connection energy storage system fault power unit energy rebalancing control method, and belongs to the technical field of power electronics and energy storage systems; and aims to solve problems such as the fact that residual energy cannot be recovered after a bypass of a fault power unit, voltage imbalance of healthy units and overcurrent impact caused by a rebalancing process. The method builds a multi-dimensional state-aware dynamic feasible region model, combines weight optimization distribution, phase smooth transition and multi-stage rate control, realizes efficient and smooth transfer of residual energy under the premise of guaranteeing system safety and grid-connection stability, and significantly improves available capacity, voltage quality and dynamic stability of the system under a fault condition.

Description

Technical Field

[0001] This invention relates to the field of power electronics and energy storage system technology, and more specifically, to a method for energy rebalancing control of fault power units in a high-voltage cascaded grid energy storage system. Background Technology

[0002] With the increasing demand for high-proportion renewable energy integration and rapid power support capabilities in new power systems, high-voltage cascaded grid-connected energy storage systems are becoming a key technological path to support grid inertia and stability due to their advantages of direct connection to the high-voltage grid without transformers, autonomous voltage and frequency construction capabilities, and modular expansion. These systems consist of multiple cascaded power cells connected in series. Each cell includes an independent DC capacitor, an H-bridge inverter circuit, and a local controller, achieving overall output voltage synthesis and energy dispatch through coordinated control. In grid-connected operation mode, the system needs to independently establish and maintain the grid voltage amplitude and phase, placing extremely high demands on the energy state consistency and dynamic response accuracy of each power cell.

[0003] When a power unit shuts down due to IGBT failure, drive malfunction, or capacitor failure, the traditional approach is to use hardware bypass to remove it from the cascade chain. However, while this method ensures short-term continuous operation, it introduces three technical challenges: First, the DC-side energy carried by the faulty unit cannot be effectively utilized or transferred, causing a non-linear decrease in the system's available energy storage capacity. Second, the remaining healthy units need to redistribute the total DC voltage, leading to overvoltage or undervoltage in some units, threatening device safety and reducing modulation margin. Third, under the grid control framework, energy imbalance directly disturbs the system's virtual impedance characteristics, inducing output voltage distortion and even frequency instability, especially under weak grid conditions or high-power sudden changes.

[0004] Existing energy rebalancing methods are mostly based on open-loop voltage equalization or fixed-ratio compensation, lacking joint modeling of post-fault system topology reconfiguration, residual energy state, and grid dynamic constraints. On the one hand, a quantitative mapping relationship between the residual capacitor voltage of the faulted unit and the total system energy storage is not established, making it difficult to achieve accurate energy recovery. On the other hand, the rebalancing process often ignores the inherent requirements of grid control for voltage phase continuity and power flow smoothness, leading to excessive rebalancing current surges and triggering protection actions. Furthermore, existing strategies generally do not incorporate real-time optimization mechanisms, failing to adaptively adjust the rebalancing rate and power allocation weights based on load changes and grid strength. Therefore, there is an urgent need for a rebalancing control method that integrates fault energy sensing, dynamic reconfiguration, and closed-loop optimization to maximize the utilization of residual energy and maintain the stability and reliability of grid operation while ensuring system safety. Summary of the Invention

[0005] The purpose of this invention is to provide a method for energy rebalancing control of faulty power units in a high-voltage cascaded grid energy storage system. This method mainly solves the technical problems in the prior art, such as the inability to effectively recover residual energy after the faulty power unit is bypassed, voltage distribution imbalance of healthy units, overcurrent impact caused by the rebalancing process, and deterioration of system dynamic performance.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A method for energy rebalancing control of faulty power units in a high-voltage cascaded grid-connected energy storage system is disclosed. The system consists of N cascaded power units, each including a DC capacitor, an H-bridge inverter circuit, and a local controller. All power units are connected in series to a high-voltage AC bus and synthesize the system output voltage in grid-connected mode. When any power unit fails and is bypassed by hardware, the remaining M healthy power units re-share the total DC link voltage, where M = N - the number of faulty units. The method includes the following steps:

[0008] S1, after a fault occurs, collects the DC capacitor voltage value of the faulty unit and the initial DC voltage, temperature status, historical state of charge and current limiting threshold of each healthy unit in real time.

[0009] S2, Based on the collected state information, a dynamic feasible region model is constructed to calculate the maximum energy increment that each healthy unit can safely absorb during the rebalancing process;

[0010] S3. Based on the comparison between the sum of the maximum energy increments and the residual energy of the faulty unit, determine the total amount of energy that actually needs to be transferred;

[0011] S4. A closed-loop optimization scheduling strategy based on weight allocation is adopted to allocate the total amount of energy that actually needs to be transferred to each healthy unit in proportion, and generate a corresponding target DC voltage command to send to each local controller, so as to realize energy injection by adjusting the modulation wave of the H-bridge inverter.

[0012] Furthermore, in this invention, the maximum energy increment that each of the health units can safely absorb is... Determined by the following formula:

[0013]

[0014] Where C is the nominal value of the DC capacitor. The maximum allowable DC voltage threshold for the i-th healthy cell. Its initial DC voltage This is a correction factor for energy absorption efficiency obtained by looking up tables based on real-time temperature and historical state of charge. To estimate power loss, is the thermal recovery time constant.

[0015] Furthermore, in this invention, if the residual energy of the fault unit is greater than the upper limit of the total energy that the system can absorb, only a portion of the energy corresponding to the upper limit of the total energy is recovered, and the remaining energy is discharged through the controllable discharge resistor built into the fault unit.

[0016] Furthermore, in this invention, the weight allocation for the i-th healthy unit in the weight allocation is... Determined by the following formula:

[0017]

[0018] in, For temperature sensitivity coefficient, This is the state-of-charge equilibrium coefficient. For reference temperature, The maximum allowable junction temperature, The historical charge state is represented by j, which indicates the j-th healthy unit.

[0019] Furthermore, in this invention, the target DC voltage Calculated using the following formula:

[0020]

[0021] in, This represents the total amount of energy that actually needs to be transferred.

[0022] Furthermore, in this invention, at the instant of rebalancing startup, a phase offset command is synchronously sent to all healthy units. This is used to compensate for reactive power disturbances caused by energy injection and maintain voltage-current phase continuity. The phase offset command is calculated by the following formula:

[0023]

[0024] in, The system's rated angular frequency, For equivalent series inductance, Let be the additional current component of the i-th unit during the rebalancing process. This represents the current grid voltage amplitude.

[0025] Furthermore, in this invention, the rebalancing process employs a multi-stage rate control strategy, including: a first stage is a rapid injection stage, the duration of which is... The first stage allows the rebalancing current to reach 80% of its limit; the second stage is the fine adjustment stage, which lasts for a certain period of time. The current limit is reduced to 50%; the third stage is the steady-state convergence stage, which continues until the DC voltage deviation of all healthy units is less than the preset threshold of 0.5% of the rated voltage.

[0026] Furthermore, in this invention, the duration of the rapid injection phase is... Duration of fine-tuning phase Determined by the following formula:

[0027]

[0028] in, For the system damping ratio, The natural oscillation frequency of the system. and All are empirical coefficients, and the damping ratio is... The short-circuit ratio (SCR) to the power grid is estimated in real time by the online identification module and smoothed by a low-pass filter before being used for timing decisions.

[0029] Furthermore, in this invention, if the DC voltage of any healthy unit exceeds its maximum DC voltage threshold or the output current exceeds its current limit threshold, the local controller of that unit immediately sends an emergency current limiting signal to the central coordinating controller. The central coordinating controller then reduces the global rebalancing power command and recalculates the remaining allocable energy, entering a safety degradation mode.

[0030] Furthermore, in this invention, status information acquisition is accomplished through a time-triggered communication protocol, with a message period of 500. s, sampling period not greater than 100 Status data is shared via dual-port RAM, written by the DMA controller and read by a timer interrupt, with an update latency not exceeding 200ms. s.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] (1) This invention constructs a multi-dimensional state-aware dynamic feasible domain model to calculate the safe energy absorption limit of each healthy power unit in real time, and matches the residual energy of the faulty unit with the total absorbable energy of the system to achieve full or optimal recovery of residual energy, thus solving the problem of nonlinear decrease in system available capacity caused by energy waste in traditional bypass strategies. At the same time, based on the weighted allocation strategy of voltage margin, temperature state and state of charge, energy is preferentially injected into healthy units with better operating conditions to avoid local overcharging or hot spot effects. In the example operating condition, the residual energy recovery rate of the faulty unit reaches 100%, and the system available capacity is increased by about 32.5% compared with the traditional open-loop voltage equalization strategy, significantly extending the system endurance under fault conditions.

[0033] (2) This invention collects multi-dimensional information such as the initial DC voltage, temperature, and historical state of charge of healthy units in real time, and sets personalized energy absorption thresholds for each unit, thus fundamentally avoiding uneven voltage distribution. Simultaneously, a phase smoothing transition mechanism is introduced at the moment of rebalancing startup. By synchronously sending phase offset commands, reactive power disturbances caused by energy injection are compensated, maintaining voltage-current phase continuity and effectively suppressing output voltage distortion. Data from the embodiment show that after rebalancing, the DC voltage deviation of all healthy units is controlled within ±4V, and the total harmonic distortion (THD) of the system output voltage only rises to 1.1%, far below the national standard limit of 2%, ensuring voltage quality and frequency stability in the grid configuration mode.

[0034] (3) This invention adopts a multi-stage rate control strategy of "rapid injection - fine adjustment - steady-state convergence". The duration of each stage is dynamically adjusted according to the damping ratio and grid short-circuit ratio identified online by the system, taking into account both energy transfer efficiency and process stability, and solving the problems of large rebalancing current impact and easy triggering of protection action in traditional methods. At the same time, a complete safety degradation mechanism is established. When overvoltage or overcurrent occurs in a healthy unit, the global rebalancing power is immediately reduced and energy is redistributed to avoid cascading failures caused by local overload. In the embodiment, the total rebalancing process takes about 300ms, and the current impact amplitude is only 40% of that of the traditional open-loop strategy, realizing uninterrupted energy transfer and continuous system operation under fault conditions. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0036] Figure 2 This is a schematic diagram of the dynamic feasible domain model and energy allocation weight calculation process based on multi-dimensional state awareness in this invention. Detailed Implementation

[0037] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0038] like Figure 1As shown, this invention discloses an energy rebalancing control method for faulty power units in a high-voltage cascaded grid-connected energy storage system. The high-voltage cascaded grid-connected energy storage system consists of N cascaded power units, each containing a DC capacitor, an H-bridge inverter circuit, and a local controller. All power units are connected in series to the high-voltage AC bus and collectively synthesize the system output voltage in grid-connected operation mode. When an unrecoverable fault occurs in a power unit, its internal hardware bypass switch is triggered, and the unit is electrically isolated and no longer participates in the energy conversion process. At this time, the system topology is reconfigured, and the total DC link voltage is redistributed by the remaining M healthy power units, where M equals N minus the number of faulty units. Since residual energy is still stored in the DC capacitor of the faulty unit, if left untreated, it will not only cause a nonlinear loss of the system's available energy storage capacity but also lead to an imbalance in the DC voltage distribution among the healthy units, thereby affecting the overall grid-connected stability and dynamic response performance of the system.

[0039] like Figure 2 As shown, the central coordinating controller initiates the status acquisition process immediately after a fault occurs. This process acquires the DC capacitor voltage value of the faulty unit in real time via a high-speed serial communication link. Simultaneously, the initial DC voltage of each healthy unit is collected. (i=1,2,…,M), Temperature state Historical state of charge and the current limiting threshold set by the local controller. The communication link uses a time-triggered protocol with a fixed message period of 500. s, ensuring that all state information is synchronously sampled under a unified time reference. The sampling period is no more than 100. This is to meet the stringent timeliness requirements of subsequent dynamic reconfiguration control. All local controllers write the aforementioned status parameters into a designated buffer in the dual-port RAM, and the central coordinating controller triggers a read operation via the DMA controller. The status data update latency is strictly controlled within 200ms. Within s.

[0040] After completing the status information acquisition, the central coordinating controller constructs a dynamic feasible domain model for system energy rebalancing after a fault. The core of this model lies in quantifying the maximum energy increment that each healthy unit can safely absorb during the rebalancing process. This energy limit is determined by the DC voltage safety margin, heat accumulation effect, and power loss, and its mathematical expression is:

[0041]

[0042] Where C is the nominal capacitance value of the DC capacitor in each power unit. The maximum allowable DC voltage threshold for the i-th healthy unit is determined by the withstand voltage rating of the power device and the system safety margin. The estimated power loss of this unit during the rebalancing process is calculated online by the local controller based on the current switching frequency, modulation ratio, and RMS current value. The thermal recovery time constant reflects the time characteristics of the heat dissipation capability of the power unit. This is an energy absorption efficiency correction factor used to compensate for the decrease in energy conversion efficiency caused by increased temperature and device aging. This correction factor is determined by the local controller based on real-time temperature. Compared with historical states of charge The value is determined by a two-dimensional lookup table pre-stored in non-volatile memory, and its range is strictly limited to between 0.85 and 1.0 to ensure the model's conservatism and engineering feasibility.

[0043] In a preferred embodiment of the present invention, the central collaborative controller completes various tasks. After calculation, the total energy that the system can absorb as a whole is further summed to obtain the upper limit. Meanwhile, the residual energy of the faulty unit Calculated as:

[0044]

[0045] Central collaborative controller and Compare and judge. If If the system is deemed capable of full energy recovery, the full energy recovery process is initiated; if Then only recycling Part of the energy is discharged slowly through the controllable discharge resistor built into the fault unit in constant power or constant current mode. The discharge rate is dynamically adjusted by the local controller according to the slope of the capacitor voltage drop to ensure that no secondary overvoltage or electromagnetic interference problems are caused.

[0046] Determining the actual total amount of energy to be transferred Subsequently, the central coordinating controller executes a closed-loop optimization scheduling strategy based on weight allocation. The core of this strategy is to assign an energy injection weight to each healthy unit. To achieve the optimal distribution of energy among healthy units. The weights... Calculated using the following formula:

[0047]

[0048] in, and These are preset empirical coefficients, with values ​​of 0.3 and 0.4 respectively, used to adjust the relative importance of temperature sensitivity and state of charge balance; This is a reference temperature, usually set to the ambient temperature or the system's rated operating temperature; The maximum allowable junction temperature for the power cell is determined by the thermal design specifications of the semiconductor device. The design of this weighting function ensures that energy is preferentially injected into healthy cells with a large voltage margin, low temperature, and a state of charge close to the median (i.e., 0.5), thereby effectively suppressing local overcharging, hot spot effects, and SOC polarization.

[0049] Based on the calculated weights With actual transferred energy The central coordinating controller further generates the target DC voltage for each healthy unit. Its expression is:

[0050]

[0051] The target voltage value is encapsulated as a control command and sent to the corresponding local controller via a time-triggered communication protocol. Each local controller, within its internal network control loop, will... The DC voltage outer loop serves as the setpoint. Its output acts as the reference value for the active current inner loop. By adjusting the modulation wave phase and amplitude of the H-bridge inverter, the energy flowing into the DC capacitor is precisely controlled, achieving target voltage tracking. The entire control loop employs a dual closed-loop structure: an outer loop uses a PI controller, and an inner loop uses a PR controller, balancing steady-state accuracy and dynamic response.

[0052] To ensure that the rebalancing process does not conflict with the main network control tasks, this invention introduces a phase smooth transition mechanism. At the instant rebalancing starts, the central coordinating controller synchronously sends a phase offset command to all healthy units. This offset is used to compensate for the disturbance to the system's reactive power balance caused by the additional current component due to energy injection, maintain the continuity of the voltage-current phase relationship, and prevent an increase in the total harmonic distortion rate. Phase offset command Calculated by the following formula:

[0053]

[0054] in, The system's rated angular frequency, The equivalent series inductance includes the power unit output filter inductance and the line distributed inductance; The additional current component generated by the i-th unit during the rebalancing process is calculated by the outer loop controller from the difference between the target voltage and the current voltage. The current grid voltage amplitude is extracted in real time by the phase-locked loop module. This phase offset is superimposed on the modulation wave phase reference of the local controller to achieve seamless switching.

[0055] As another key feature of this invention, the rebalancing process employs a multi-stage rate control strategy to balance speed and stability. This strategy consists of three stages: the first stage is a rapid injection stage, lasting for [duration missing]. During this stage, the rebalancing current is allowed to reach 80% of its limit to achieve efficient energy transfer; the second stage is the fine-tuning stage, lasting for [duration missing]. The current limit is reduced to 50% to eliminate residual oscillations caused by voltage deviation; the third stage is the steady-state convergence stage, which continues until the DC voltage deviation of all healthy units is less than a preset threshold ε, typically 0.5% of the rated voltage. The duration of each stage is determined by the central co-controller based on the damping ratio identified online by the system. The short-circuit ratio (SCR) is dynamically adjusted relative to the power grid. Specifically, the formula for calculating the stage duration is:

[0056]

[0057] in, The system's natural oscillation frequency is determined by the system's equivalent LC parameters; and These are empirical coefficients, with values ​​of 1.2 and 2.5 respectively. Damping ratio The short-circuit ratio (SCR) is estimated in real time by the system's online identification module based on voltage-current step response data, and smoothed by a low-pass filter with a cutoff frequency of 10Hz to suppress measurement noise from interfering with timing decisions.

[0058] Throughout the rebalancing process, the system is equipped with a comprehensive safety protection mechanism. If the DC voltage of any healthy unit exceeds its set limit... The threshold, or its output current exceeds the local controller setting. Upon receiving the limit value, the local controller of the affected unit immediately sends an emergency current limiting signal to the central coordinating controller via a high-speed interrupt channel. Upon receiving this signal, the central coordinating controller immediately reduces the global rebalancing power command and recalculates the allocable energy limit for the remaining healthy units, entering a safety degradation mode. In this mode, the energy transfer rate is limited within safety boundaries, and the system prioritizes ensuring the continuity of basic network functions to avoid cascading shutdowns caused by localized overloads.

[0059] To verify the effectiveness of the technical solution of the present invention, a specific embodiment and comparative example are provided below, along with relevant performance data comparison.

[0060] In one specific embodiment, a high-voltage cascaded grid energy storage system consisting of 10 power units is constructed, with a rated output voltage of 10kV and a nominal DC capacitance of 4700 kWh for each power unit. F, rated DC voltage is 1000V. The system operates in network mode, with an initial SOC of 0.6 for each unit and a temperature of 45℃. During operation, power unit number 5 experiences an IGBT drive failure and is bypassed by hardware; at the time of the failure, the DC capacitor voltage is 950V. The central coordinating controller is at 100... The status acquisition is completed within s, and the calculation is obtained. =2.12kJ. Of the remaining 9 healthy cells, 3 cells have a temperature of 50℃ and a SOC of 0.7; the remaining 6 cells have a temperature of 42℃ and a SOC of 0.55. Based on the dynamic feasible region model, the calculated... =2.35 kJ, greater than Therefore, full recovery was initiated. After weight allocation, the low-temperature, low-SOC unit achieved a higher injection ratio. The rebalancing process employs a three-stage control. It takes 80ms. The initial time was 160ms, with a total latency of approximately 300ms. Ultimately, the DC voltage deviation of all healthy cells was controlled within... Within 4V, the system output voltage THD increased from 0.8% to 1.1%, which did not exceed the national standard limit of 2%.

[0061] As a comparative example, a traditional open-loop voltage equalization strategy was adopted: after a fault, the energy of the faulty unit was directly and evenly distributed to the remaining healthy units, without considering temperature, SOC, and current limits. Under the same operating conditions, because the differences in state between units were not considered, the DC voltage of the two high-temperature, high-SOC units rapidly approached the 1050V upper limit during the rebalancing process, triggering overvoltage protection and causing the rebalancing to be interrupted. Ultimately, only 1.6kJ of energy was recovered, the system's usable capacity loss reached 25%, and the output voltage THD rose to 2.8%, exceeding the acceptable range. The key performance indicators of this embodiment and the comparative example are compared in Table 1 below.

[0062] Table 1 Comparison of key performance indicators between the examples and comparative examples

[0063]

[0064] The above data shows that, while ensuring system safety, the present invention significantly improves energy recovery efficiency and voltage quality, and effectively avoids the problems of local overload and performance degradation caused by ignoring multidimensional state differences in traditional methods.

[0065] Furthermore, at the engineering implementation level, the central coordinating controller employs a high-performance embedded processor based on the ARM Cortex-M7 core, with a clock speed of no less than 480MHz, equipped with a hardware floating-point unit and DMA controller to meet the requirements of real-time state modeling and optimization calculations. The local controller uses a TI C2000 series digital signal controller, integrating a high-resolution PWM module and a fast ADC, with a control loop update cycle of 20... All controllers communicate via a CAN FD bus with a baud rate configured at 5 Mbit / s, ensuring low latency and high reliability in command transmission. The dual-port RAM is the IDT 70V25 series, with a capacity of 32KB, supporting zero-wait status access to meet the real-time requirements of status data sharing.

[0066] In terms of software architecture, the central collaborative controller runs an operating system based on time-partitioned scheduling, dividing status acquisition, model calculation, instruction generation, and security monitoring into independent tasks, each with a cycle of 100 seconds. s, 500 s, 500 s and 200 The local controller adopts a foreground-background architecture. The foreground is a timed interrupt service routine responsible for the execution of the control loop; the background is the main loop, handling communication and status reporting. All algorithm modules have undergone fixed-point processing to reduce computational resource consumption and improve execution efficiency.

[0067] In summary, the energy rebalancing control method for faulted power units in high-voltage cascaded grid-connected energy storage systems described in this invention constructs a complete, robust, and efficient rebalancing control system by integrating multi-dimensional state perception, dynamic feasible region modeling, weight optimization allocation, phase smooth transition, and multi-stage rate control. This method not only achieves the safe, efficient, and smooth transfer of residual energy from faulted units but also significantly improves the system's available energy storage capacity, voltage waveform quality, and dynamic stability under fault conditions, providing solid technical support for the highly reliable operation of high-voltage cascaded grid-connected energy storage systems. Those skilled in the art, after reading this specific embodiment, can fully reproduce the technical solution of this invention and apply it to actual product development based on the described technical details and conventional engineering practices.

[0068] The above embodiments are merely one of the preferred embodiments of the present invention and should not be used to limit the scope of protection of the present invention. Any modifications or refinements made to the main design concept and spirit of the present invention that are not of substantial significance, but solve the same technical problem as the present invention, should be included within the scope of protection of the present invention.

Claims

1. A method for energy rebalancing control of faulty power units in a high-voltage cascaded grid-connected energy storage system, wherein the system consists of N cascaded power units, each power unit including a DC capacitor, an H-bridge inverter circuit, and a local controller, all power units are connected in series to a high-voltage AC bus and synthesize the system output voltage in grid-connected mode; when any power unit fails and is bypassed by hardware, the remaining M healthy power units reshare the total DC link voltage, where M = N − the number of faulty units; characterized in that, The method includes the following steps: S1, after a fault occurs, collects the DC capacitor voltage value of the faulty unit and the initial DC voltage, temperature status, historical state of charge and current limiting threshold of each healthy unit in real time. S2, Based on the collected state information, a dynamic feasible region model is constructed to calculate the maximum energy increment that each healthy unit can safely absorb during the rebalancing process; wherein, the maximum energy increment that each healthy unit can safely absorb... Determined by the following formula: Where C is the nominal value of the DC capacitor. The maximum allowable DC voltage threshold for the i-th healthy cell. Its initial DC voltage This is a correction factor for energy absorption efficiency obtained by looking up tables based on real-time temperature and historical state of charge. To estimate power loss, The thermal recovery time constant; S3. Based on the comparison between the sum of the maximum energy increments and the residual energy of the faulty unit, determine the total amount of energy that actually needs to be transferred; S4. A closed-loop optimization scheduling strategy based on weight allocation is adopted to allocate the total amount of energy that actually needs to be transferred to each healthy unit in proportion, and generate a corresponding target DC voltage command to send to each local controller, so as to realize energy injection by adjusting the modulation wave of the H-bridge inverter.

2. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 1, characterized in that, If the residual energy of the fault unit is greater than the upper limit of the total energy that the system can absorb, only the portion of energy corresponding to the upper limit of the total energy will be recovered, and the remaining energy will be discharged through the controllable discharge resistor built into the fault unit.

3. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 2, characterized in that, The weight allocation for the i-th healthy unit in the weight allocation. Determined by the following formula: in, For temperature sensitivity coefficient, This is the state-of-charge equilibrium coefficient. For reference temperature, The maximum allowable junction temperature, The historical charge state is represented by j, which indicates the j-th healthy unit.

4. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 3, characterized in that, The target DC voltage Calculated using the following formula: in, This represents the total amount of energy that actually needs to be transferred.

5. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 4, characterized in that, At the moment of rebalancing startup, a phase offset command is simultaneously sent to all healthy units. This is used to compensate for reactive power disturbances caused by energy injection and maintain voltage-current phase continuity. The phase offset command is calculated by the following formula: in, The system's rated angular frequency, For equivalent series inductance, Let be the additional current component of the i-th unit during the rebalancing process. This represents the current grid voltage amplitude.

6. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 5, characterized in that, The rebalancing process employs a multi-stage rate control strategy, including: the first stage is a rapid injection stage, lasting t1, during which the rebalancing current is allowed to reach 80% of its limit value; the second stage is a fine adjustment stage, lasting t2, during which the current limit is reduced to 50%; and the third stage is a steady-state convergence stage, which continues until the DC voltage deviation of all healthy units is less than a preset threshold of 0.5% of the rated voltage.

7. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 6, characterized in that, Duration of the rapid injection phase Duration of fine-tuning phase Determined by the following formula: in, For the system damping ratio, The natural oscillation frequency of the system. and All are empirical coefficients, and the damping ratio is... The short-circuit ratio (SCR) to the power grid is estimated in real time by the online identification module and smoothed by a low-pass filter before being used for timing decisions.

8. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 7, characterized in that, If the DC voltage of any healthy unit exceeds its maximum DC voltage threshold or the output current exceeds its current limit threshold, the local controller of that unit immediately sends an emergency current limiting signal to the central coordinating controller. The central coordinating controller then reduces the global rebalancing power command and recalculates the remaining allocable energy, entering a safety degradation mode.

9. The energy rebalancing control method for fault power units in a high-voltage cascaded grid energy storage system according to claim 8, characterized in that, Status information is collected via a time-triggered communication protocol with a message period of 500. s, sampling period not greater than 100 Status data is shared via dual-port RAM, written by the DMA controller and read by a timer interrupt, with an update latency not exceeding 200ms. s.

Images