Multi-machine parallel current sharing algorithm for 5.7 MW high-power hydrogen production power supply

Abstract

The invention relates to the technical field of hydrogen production power supplies, and discloses a 5.7 MW high-power hydrogen production power supply multi-machine parallel current sharing algorithm, which effectively improves the current sharing precision of a 5.7 MW high-power hydrogen production power supply multi-machine parallel system through droop control and PI regulation current sharing control improved by combining a master-slave architecture, and can control the current imbalance degree of each parallel module within + / -3%. The method is far better than a traditional method, ensures that each power supply module uniformly shares a load, gives full play to the overall power output capability of the system, improves the hydrogen production efficiency, has a rapid dynamic response characteristic, can rapidly adjust current distribution under the condition of sudden change of the load, enables the system to recover a stable current-sharing state within 20ms, avoids equipment damage caused by current impact, and improves the hydrogen production efficiency. Reliable operation of the system is guaranteed, and complex and changeable load requirements in the hydrogen production process are met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydrogen production power supply, in particular to a 5.7MW high-power hydrogen production power supply multi-machine parallel current sharing algorithm. BACKGROUND

[0002] With the rapid development of hydrogen energy industry, the demand for high-power hydrogen production power supply is increasing. In many hydrogen production scenarios, a single power supply often cannot meet the high-power output requirement, so multi-machine parallel connection becomes a common solution. However, when multiple machines are connected in parallel, due to the differences in parameters of each power supply module, inconsistent line impedance and other factors, the current distribution between each parallel module will be uneven, which not only reduces the overall efficiency of the system, but also may cause some modules to be overloaded, affecting the reliability and stability of the system and shortening the service life of the equipment.

[0003] At present, the current sharing methods commonly used in digital power supply applications include droop control, average current sharing control, maximum current automatic sharing control, and thermal stress sharing control. Among them, the droop control is simple to implement and does not require additional hardware, but its control precision is limited and the response is slow, and the line impedance has a great impact on the control precision. The average current sharing control requires detecting the output currents of all parallel modules and calculating the average current through complex circuits or control, and then adjusting the output current of each module to the average current value. This scheme requires complex hardware circuits to ensure the accuracy of current sharing. The maximum current automatic sharing method requires real-time detection of the output currents of each module, finds the maximum current value, and adjusts the currents of other modules to be equal to the maximum current value. This scheme requires a hardware circuit to detect the maximum current, which has a certain delay and the current sharing effect is not good when the load changes rapidly. Therefore, we propose a 5.7MW high-power hydrogen production power supply multi-machine parallel current sharing algorithm. SUMMARY

[0004] The purpose of the present application is to provide a 5.7MW high-power hydrogen production power supply multi-machine parallel current sharing algorithm to solve the problems raised in the background.

[0005] In order to solve the above technical problems, the present application provides the following technical scheme: a 5.7MW high-power hydrogen production power supply multi-machine parallel current sharing algorithm, which is based on a master-slave control architecture and applied to a 5.7MW hydrogen production power supply system composed of multiple power supply modules, including the following steps:

[0006] 1) Master selection: when the system is initialized, each power supply module exchanges output voltage range, rated current and current working state parameters through a pre-set communication protocol, and selects the module with the best performance as the master according to the output capacity, stability, historical operation data and comprehensive selection rules, and the rest are slaves;

[0007] 2) Voltage reference signal generation: the host collects its output voltage through a high-precision voltage sampling circuit at a frequency of 12 kHz, generates a voltage reference signal after filtering by a built-in digital filter, and sends it to each slave through the CAN bus;

[0008] 3) Current sampling and feedback: each slave collects its output current at a frequency of 12 kHz through a sampling circuit composed of a high-precision shunt and an operational amplifier, converts it into a voltage signal suitable for the processor, and feeds it back to the host in real time through the CAN bus;

[0009] 4) Current sharing calculation: the host aggregates the current feedback values of all slaves, calculates the average output current I aυg of the parallel modules, and the formula is The improved droop control combined with PI regulation is used to calculate the voltage compensation amount ΔV i of the i-th slave according to the formula p , where k i is the proportional coefficient, k i is the integral coefficient, and I i is the actual output current of the i-th slave;

[0010] 5) Voltage compensation signal transmission: the host sends the voltage compensation amount ΔV i of each slave to the corresponding slave through the CAN bus, and the slave superimposes ΔV aυg on the voltage reference signal to generate a new voltage reference value for itself;

[0011] 6) Dynamic adjustment: repeat steps 2-5, and the slave adjusts the conduction duty cycle of the power switch tube through PWM technology to adjust the output voltage in real time, so that the output current of each slave tends to I p ;

[0012] 7) Fault detection and processing: the host monitors the current feedback abnormalities and communication interruption states of the slaves in real time, and if a faulty slave is detected, it is immediately removed from the current sharing loop, and the voltage compensation amount of the remaining slaves is recalculated to adjust the load current distribution. The faulty slave is automatically connected to the system after repair.

[0013] Preferably, the output capacity, stability, and historical operation data comprehensive selection rule in step 1) is specifically: preferentially selecting a power module with a rated current ≥ 1.44 MW, a continuous fault-free operation time ≥ 3000 h, and an output voltage fluctuation ≤ ± 1% as the host.

[0014] Preferably, the sampling accuracy of the high-precision voltage sampling circuit in step 2) is ≤ 0.1%, and the built-in digital filter is a second-order Butterworth low-pass filter with a cutoff frequency of 1 kHz.

[0015] Preferably, the range of the high-precision shunt in step 3) is 0-2000A, and the accuracy level is 0.05 level, the operational amplifier uses a low offset voltage operational amplifier, and the amplification multiple is set to 100 times.

[0016] Preferably, the value range of the proportional coefficient k p in step 4) is 0.5-1.2, and the value range of the integral coefficient k i is 0.1-0.3, and the specific value is determined through experimental debugging to adapt to different load characteristics.

[0017] Preferably, the response speed of the dynamic adjustment in step 6) meets: when the load suddenly increases from 50% to 100% or suddenly decreases from 100% to 30%, the time for the system to recover to a stable current sharing state is ≤20ms.

[0018] Preferably, the judgment threshold of the fault detection in step 7) is: when the deviation of the current feedback value of a slave machine from I aυg is ≥15% and the duration is ≥500μs, it is determined that the current feedback is abnormal; and when the CAN bus communication interruption time is ≥100ms, it is determined that the communication is interrupted.

[0019] Preferably, the 5.7MW hydrogen production power supply system is composed of four 1.44MW power supply modules in parallel, and the average output current I aυg =(I1+I2+I3+I4) / 4 in step 4), wherein I1, I2, I3, I4 are the actual output currents of the four power supply modules, respectively.

[0020] Compared with the prior art, the beneficial effects achieved by the present application are:

[0021] Firstly, the present application effectively improves the current sharing precision of the 5.7MW high-power hydrogen production power supply multi-machine parallel system by combining the master-slave architecture with the improved droop control and PI regulation current sharing control, and the current imbalance degree of each parallel module can be controlled within ±3%, which is much better than the traditional method, ensuring that each power supply module shares the load evenly, fully utilizes the overall power output capability of the system, and improves the hydrogen production efficiency.

[0022] Secondly, the present application has a fast dynamic response characteristic, which can quickly adjust the current distribution under load mutation, so that the system recovers to a stable current sharing state within 20ms, avoids equipment damage caused by current impact, ensures reliable operation of the system, and meets the complex and variable load demand in the hydrogen production process.

[0023] Thirdly, the perfect fault detection and processing mechanism of the present application can monitor and automatically isolate the faulty module in real time, and intelligently adjust the working state of the remaining modules, realizing uninterrupted operation of the system, greatly improving the stability and usability of the entire hydrogen production power supply system, reducing maintenance cost and downtime, and providing strong support for large-scale industrial hydrogen production. Attached Figure Description

[0024] Figure 1 Diagram of a hydrogen production power system;

[0025] Figure 2 Master-slave control framework diagram;

[0026] Figure 3 This is a diagram of a four-phase interleaved parallel double closed-loop control for a DC / DC unit. 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] This invention provides the following technical solutions:

[0029] Please see Figure 1 , Figure 2 , Figure 3 A current-sharing algorithm for a 5.7MW high-power hydrogen production power supply system with multiple parallel units is disclosed. The algorithm, based on a master-slave control architecture, is applied to a 5.7MW hydrogen production power supply system composed of multiple power modules connected in parallel. The algorithm includes the following steps:

[0030] 1) Master election: During system initialization, each power module exchanges its output voltage range, rated current and current operating status parameters through a preset communication protocol. Based on the comprehensive election rules of output capacity, stability and historical operating data, the module with the best performance is selected as the master, and the rest are slaves.

[0031] 2) Voltage reference signal generation: The host acquires its own output voltage at a frequency of 12kHz through a high-precision voltage sampling circuit, and generates a voltage reference signal after filtering by a built-in digital filter. The signal is then sent to each slave device via the CAN bus.

[0032] 3) Current sampling and feedback: Each slave device collects its own output current at a frequency of 12kHz through a sampling circuit composed of a high-precision shunt and operational amplifier, and converts it into a voltage signal adapted to the processor, which is then fed back to the host in real time via the CAN bus.

[0033] 4) Current sharing calculation: The master unit aggregates the current feedback values ​​of all slave units and calculates the average output current I of the parallel modules. aυg The formula is An improved droop control combined with PI regulation is adopted, according to the formula. Calculate the voltage compensation amount ΔV for the i-th slave device.i where k p is the proportional coefficient, k i is the integral coefficient, I i is the actual output current of the i-th slave;

[0034] 5) Voltage compensation signal transmission: the master sends the voltage compensation amount AV i of each slave to the corresponding slave through the CAN bus, and the slave superimposes AV i on the voltage reference signal to generate a new voltage reference value of itself;

[0035] 6) Dynamic adjustment: repeat steps 2-5, and the slave adjusts the power switch tube conduction duty ratio through the PWM technique to adjust the output voltage in real time, so that the output current of each slave approaches I aυg ;

[0036] 7) Fault detection and processing: the master real-time monitors the current feedback abnormality and communication interruption state of the slave, and if a faulty slave is detected, it is immediately removed from the current sharing loop, and the voltage compensation amount of the remaining slaves is recalculated to adjust the load current distribution. The faulty slave is automatically connected to the system after repair.

[0037] The comprehensive selection rule of output capacity, stability and historical operation data in step 1) is: preferentially select the power module with rated current ≥1.44MW, continuous fault-free running time ≥3000h, and output voltage fluctuation ≤±1% as the master.

[0038] The sampling accuracy of the high-precision voltage sampling circuit in step 2) is ≤0.1%, and the built-in digital filter is a second-order Butterworth low-pass filter with a cutoff frequency of 1kHz.

[0039] The range of the high-precision shunt in step 3) is 0-2000A, and the accuracy level is 0.05 level. The operational amplifier uses a low-offset voltage operational amplifier with an amplification factor of 100.

[0040] The value range of the proportional coefficient k p in step 4) is 0.5-1.2, and the value range of the integral coefficient k i is 0.1-0.3. The specific value is determined through experimental debugging to adapt to different load characteristics.

[0041] The response speed of dynamic adjustment in step 6) meets: when the load suddenly increases from 50% to 100% or suddenly decreases from 100% to 30%, the time for the system to recover to the stable current sharing state is ≤20ms.

[0042] The judgment threshold of fault detection in step 7) is: when the current feedback value of a certain slave is I aυgWhen the deviation of the current feedback is greater than or equal to 15% and the duration is greater than or equal to 500 microseconds, the current feedback is determined to be abnormal; when the CAN bus communication interruption time is greater than or equal to 100 milliseconds, the communication is determined to be interrupted.

[0043] The 5.7 MW hydrogen production power supply system is composed of four 1.44 MW power supply modules in parallel, and the average output current I of step 4) is aυg =(I1+I2+I3+I4) / 4, wherein I1, I2, I3, and I4 are the actual output currents of the four power supply modules, respectively.

[0044] Through the above technical solution, steps 2-5 are repeatedly performed during system operation to realize real-time dynamic current sharing. When the load changes suddenly, the host can quickly detect the voltage change. Due to its high sampling frequency and fast computing and processing capability, the current sharing calculation is completed in a very short time (less than [X] ms), and the adjustment signal is sent to the slave. The slave also has fast response capability and can complete voltage adjustment within [X] ms after receiving the signal, ensuring that the system current distribution remains uniform and stable during load changes, maintaining the efficient operation of the system. The host monitors the working state of each slave in real time, including whether the current feedback value is abnormal and whether the communication is interrupted. Once a fault is found in a slave, the host takes appropriate measures, on the one hand, to remove the faulty slave from the current sharing control loop to avoid the expansion of the fault affecting the entire system; on the other hand, to adjust the voltage reference value of the remaining normal slaves and redistribute the load current to ensure uninterrupted operation of the system and maintain stable total output power to meet the continuity requirements of the hydrogen production process. After the faulty slave is repaired, it can be automatically reconnected to the system and participate in current sharing control. The entire process does not require manual intervention and has high intelligence.

[0045] Although embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made thereto without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A current-sharing algorithm for a 5.7MW high-power hydrogen production power supply with multiple units in parallel, characterized in that... The algorithm, based on a master-slave control architecture, is applied to a 5.7MW hydrogen production power system composed of multiple power modules connected in parallel, and includes the following steps: 1) Master election: During system initialization, each power module exchanges its output voltage range, rated current and current operating status parameters through a preset communication protocol. Based on the comprehensive election rules of output capacity, stability and historical operating data, the module with the best performance is selected as the master, and the rest are slaves. 2) Voltage reference signal generation: The host acquires its own output voltage at a frequency of 12kHz through a high-precision voltage sampling circuit, and generates a voltage reference signal after filtering by a built-in digital filter. The signal is then sent to each slave device via the CAN bus. 3) Current sampling and feedback: Each slave device collects its own output current at a frequency of 12kHz through a sampling circuit composed of a high-precision shunt and operational amplifier, and converts it into a voltage signal adapted to the processor, which is then fed back to the host in real time via the CAN bus. 4) Current sharing calculation: The master unit aggregates the current feedback values ​​of all slave units and calculates the average output current I of the parallel modules. aυg The formula is An improved droop control combined with PI regulation is adopted, according to the formula. Calculate the voltage compensation amount ΔV for the i-th slave device. i , where k p For proportionality coefficient, k i I is the integral coefficient. i This represents the actual output current of the i-th slave device. 5) Voltage compensation signal transmission: The host transmits the voltage compensation amount ΔV of each slave device via the CAN bus. i Send to the corresponding slave device, the slave device will send ΔV i It is superimposed on the voltage reference signal to generate its own new voltage reference value; 6) Dynamic Adjustment: Repeat steps 2-5. The slave device adjusts the duty cycle of the power switch transistor using PWM technology to adjust the output voltage in real time, so that the output current of each slave device approaches I. aυg ; 7) Fault detection and handling: The host monitors the current feedback abnormality and communication interruption status of the slave in real time. If a faulty slave is detected, it is immediately disconnected from the current sharing loop, and the voltage compensation of the remaining slaves is recalculated. The load current distribution is adjusted, and the faulty slave is automatically reconnected to the system after being repaired.

2. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation according to claim 1, characterized in that: The specific selection rules for step 1) based on a comprehensive evaluation of output capacity, stability, and historical operating data are as follows: priority is given to selecting power modules with a rated current ≥1.44MW, continuous fault-free operation time ≥3000h, and output voltage fluctuation ≤±1% as the host.

3. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation as described in claim 1, characterized in that: In step 2), the sampling accuracy of the high-precision voltage sampling circuit is ≤0.1%, and the built-in digital filter is a second-order Butterworth low-pass filter with a cutoff frequency of 1kHz.

4. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation as described in claim 1, characterized in that: In step 3), the high-precision shunt has a range of 0-2000A and an accuracy class of 0.

05. The operational amplifier uses a low offset voltage operational amplifier with a gain of 100.

5. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation as described in claim 1, characterized in that: In step 4), the proportionality coefficient k p The value range is 0.5-1.2, and the integral coefficient k i The value ranges from 0.1 to 0.

3. The specific value is determined through experimental debugging to adapt to different load characteristics.

6. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation as described in claim 1, characterized in that: The dynamic adjustment response speed in step 6) satisfies the following condition: when the load suddenly increases from 50% to 100% or suddenly decreases from 100% to 30%, the time for the system to recover to a stable current sharing state is ≤20ms.

7. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation as described in claim 1, characterized in that: The fault detection threshold in step 7) is: when the current feedback value of a slave device is equal to I... aυg When the deviation is ≥15% and the duration is ≥500μs, it is judged as an abnormal current feedback; when the CAN bus communication interruption time is ≥100ms, it is judged as a communication interruption.

8. The current sharing algorithm for a 5.7MW high-power hydrogen production power supply in parallel operation according to claim 1, characterized in that: The 5.7MW hydrogen production power system consists of four 1.44MW power modules connected in parallel. The average output current I in step 4) is... aυg = (I1+I2+I3+I4) / 4, where I1, I2, I3, and I4 are the actual output currents of the four power modules, respectively.

Images