Segmented soft start control method for forward charging of bidirectional CLLC resonant converter

Abstract

The invention discloses a sectional type soft start control method for forward charging of a bidirectional CLLC resonant converter, and relates to the technical field of power electronic converter control, the sectional type soft start control method is realized based on a vehicle-mounted charger OBC system, the OBC system comprises a preceding-stage totem pole PFC rectification circuit, a CLLC charging and discharging circuit and an auxiliary power supply circuit, PFC output voltage is used as CLLC primary side bus voltage, and the CLLC charging and discharging circuit is used for charging the CLLC resonant converter. The secondary side output of the CLLC charges the storage battery, and the system is controlled by an external BMS or an upper computer. According to the method, a dynamic parameter setting algorithm based on model prediction is adopted, and the minimum constant current point under different starting working conditions is automatically adjusted, so that the stability of soft starting under each working condition is realized; aiming at the problem of stage switching oscillation, smooth transition of a control mode is ensured, and the problems of gain abrupt change and relatively long soft start time at the end of soft start of an initial fixed slope frequency reduction strategy are solved, so that the gain in the whole soft start process is closer to a gain curve of a converter, and the peak current of a resonant cavity is further reduced.

Description

Technical Field

[0001] This invention relates to the field of power electronic converter control technology, specifically a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter. Background Technology

[0002] Bidirectional CLLC resonant converters have been widely used in the fields of new energy and electric vehicles due to their significant advantages such as soft switching and high efficiency. Similar to other topologies, their startup process requires a soft-start strategy to ensure stable system operation. The startup phase of the converter involves a complex process of energy building up the resonant cavity. If the soft-start strategy is missing or improperly designed, it can easily lead to damage to the switching devices and load equipment, seriously affecting the reliability of the system.

[0003] Currently, the soft-start control strategy for forward buck charging of CLLC resonant converters has formed a certain theoretical system, and its technological evolution has undergone a transformation from single control to composite adaptive methods. Early frequency conversion control (PFM) and variable duty cycle control (PWM), while simple to implement, suffer from poor dynamic response and insufficient adaptability to operating conditions, making it difficult to achieve a balance between startup speed and current suppression. In recent years, segmented hybrid control and load identification adaptive control have become research hotspots, improving system reliability through phased optimization of the startup process. However, existing solutions still have significant technical bottlenecks: voltage and current surges are prone to occur during phase switching; parameters need to be preset and cannot adapt to different operating conditions; gain surges at the end of soft start lead to high peak current in the resonant cavity; startup time is relatively long, resulting in insufficient balance between efficiency and stability. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter, which solves the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter, based on an on-board charger (OBC) system. The OBC system includes a front-stage totem-pole PFC rectifier circuit, a CLLC charging / discharging circuit, and an auxiliary power supply circuit. The PFC output voltage serves as the primary bus voltage of the CLLC, and the CLLC secondary output charges the battery. The system is controlled by an external BMS or a host computer and includes the following steps: S1. Pre-startup and adaptive calculation of minimum constant current value: After receiving an external charging command and stabilizing the PFC rectified voltage, the controller enters the pre-start phase and collects the CLLC primary bus voltage in real time. and secondary battery voltage The minimum constant current value under the current operating condition is adaptively calculated based on the resonant cavity parameters. ; S2, Soft Start Phase 1: Start the CLLC converter and enter the first stage of soft start: the frequency is fixed at the maximum frequency, the duty cycle is changed using the Sigmoid curve, and the Sigmoid function is used to realize the characteristics of gradual rise, acceleration and gradual rise. The duty cycle is increased from the initial duty cycle to the corresponding maximum duty cycle, thus completing the first stage of variable duty cycle soft start. S3, Soft Start Phase Two: After the first stage ends, the second stage of soft start begins: the duty cycle remains unchanged at the final value of the first stage, and the frequency is designed using an inverse proportional function to achieve a non-linear frequency decrease, ensuring a smooth transition of the output gain, while effectively reducing the soft start time of the second stage. S4, Closed-loop soft start: This step is the third stage of soft start. Using the switching frequency at the end of the second stage as the initial closed-loop frequency, it achieves the initial value transfer from open loop to closed loop. Based on the maximum current value of the external command, closed-loop control is used to slowly increase the current command, causing the output current to gradually increase from the external current value. The current is smoothly increased to the target current. Once the output current reaches the target current and stabilizes, the soft start ends and the system enters normal closed-loop operation.

[0006] Furthermore, in step S1, the external charging command includes start / stop, maximum charging voltage, and current.

[0007] Furthermore, in step S1, the minimum constant current value under the current operating condition... The calculation formula is as follows: ; The turns ratio of the high-voltage and low-voltage windings. The characteristic impedance of the resonant cavity, The resonant frequency, For maximum operating frequency, Magnetizing inductor With resonant inductor The ratio of .

[0008] Furthermore, in step S2, the duty cycle changes over time as follows: ; The initial duty cycle, This represents the duty cycle when the dead time is minimized, with a value ranging from 0.42 to 0.45, varying with frequency. This is a coefficient of variation used to adjust the rate of change in the middle section of the S-curve and to slow down the rate of increase in the final segment of the S-curve. This is the soft start deadline; In step S2, the PWM generator adopts forced complementary output, and the rising edge of the upper and lower transistors is inserted into the dead time to prevent shoot-through. By using a fixed comparison value, i.e. half a cycle, in conjunction with dead time adjustment, the PWM duty cycle of the upper and lower transistors is kept consistent, thereby achieving symmetrical peak-to-peak values ​​of the positive and negative currents in the resonant cavity.

[0009] Furthermore, in step S3, the inverse proportional function of the frequency change is: ; The maximum frequency value, This is the minimum frequency value allowed for down-frequency soft-start. This is the slope coefficient of the curve, used to adjust the rate of change in the middle section of the S-curve. This is the central time point used to adjust the total duration of the second-stage soft start.

[0010] Furthermore, in steps S1 to S4, after the whole machine is connected to the mains power, the system is in standby mode. After receiving the charging command from the external BMS or host computer, it starts to switch to forward charging mode. When the rectified voltage of the front-end PFC reaches the specified value, the PFC is automatically started and CLLC charging begins.

[0011] Furthermore, CLLC charging consists of two main parts. The first part is segmented soft start, i.e., steps S2-S3. The second part is normal closed-loop control. After the segmented soft start is completed, it will switch to normal closed-loop control, i.e., step S4, and become normal charging. It will enter standby mode after the battery is fully charged or end after receiving an external shutdown command.

[0012] Furthermore, the CLLC forward charging process is as follows: standby state, receiving charging command, PFC voltage stabilization, entering segmented soft start, soft start completion, normal closed-loop charging, battery fully charged or shutdown command, standby.

[0013] Furthermore, in step S1, the current data is collected. and In addition, battery status data is obtained via BMS: temperature Voltage fluctuations in the last three charging cycles Number of loops And store it in the historical database; Load characteristic models are trained based on machine learning algorithms such as BP neural networks. ; in, This is the result of the original static calculation formula. The correction factor is dynamic. At low temperatures and high cycle counts, the correction factor increases by 1.2-1.5 times to compensate for the increased current demand caused by the increased internal resistance. Corrected predictions As the second-stage switching threshold, the coefficient of change of the first-stage Sigmoid curve is adjusted in advance. When load fluctuations exceed a preset threshold, it is considered a large fluctuation. Reduce by 20% to slow the growth rate of duty cycle.

[0014] This invention provides a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter, which has the following advantages: This segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter achieves adaptive parameter adjustment and multi-stage disturbance-free switching during the startup process. First, to address the issue of parameter self-tuning, a model-predictive dynamic parameter tuning algorithm is adopted. By monitoring key parameters such as input and output voltages and load conditions in real time at startup, the minimum constant current point under different startup conditions is automatically adjusted, thereby achieving stability of soft-start under various conditions. Second, to address the oscillation problem during stage switching, a feedforward compensation mechanism and gradual output change ensure a smooth transition of the control mode. The segmented soft-start strategy can solve the problems of sudden gain changes at the end of the soft-start and long soft-start time in the initial fixed-slope frequency reduction strategy, making the gain of the entire soft-start process closer to the gain curve of the converter, thereby further reducing the peak current of the resonant cavity.

[0015] This segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter introduces a load dynamic prediction algorithm. During the pre-start phase, it collects historical battery data in real time, including voltage fluctuation rate, temperature, and cycle count, to establish a load characteristic model. It pre-sets the minimum constant current value and stage switching threshold, achieving dual-dimensional parameter optimization of static parameters and dynamic prediction. This addresses the problem of delayed start-up parameter adaptation under extreme operating conditions caused by dynamic load characteristics of the battery, such as sudden changes in internal resistance at low temperatures and capacity decay of aging batteries. Attached Figure Description

[0016] Figure 1 This is a block diagram of the OBC system for a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention. Figure 2 This is a flowchart illustrating the bidirectional CLLC forward charging process of a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention. Figure 3 This is a flowchart illustrating the entire soft-start process of a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention. Figure 4The adaptive current-voltage variation curve is shown in the figure for the segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention. Figure 5 The waveforms of duty cycle versus time in the first and second stages of the soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention are shown. Figure 6 The waveforms of the switching frequency versus time in the first and second stages of the soft-start control method for forward charging of a bidirectional CLLC resonant converter according to the present invention are shown. Figure 7 The experimental waveforms of the soft-start (primary-side drive waveform and secondary-side resonant cavity current) scheme of the CLLC converter are for the segmented soft-start control method for forward charging of the bidirectional CLLC resonant converter of the present invention. Detailed Implementation

[0017] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0018] like Figures 1-7 As shown, this invention provides a technical solution: a segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter, implemented based on an on-board charger (OBC) system. The OBC system includes a front-stage totem-pole PFC rectifier circuit, a CLLC charging / discharging circuit, and an auxiliary power supply circuit. The PFC output voltage serves as the primary bus voltage of the CLLC, and the CLLC secondary output charges the battery. The system is controlled by an external BMS or a host computer and includes the following steps: S1. Pre-startup and adaptive calculation of minimum constant current value: After receiving an external charging command and the PFC rectified voltage stabilizes, the controller enters the pre-start phase. The external charging command includes start / stop, maximum charging voltage, and current. Real-time acquisition of CLLC primary bus voltage and secondary battery voltage The minimum constant current value under the current operating condition is adaptively calculated based on the resonant cavity parameters. ; Minimum constant current value under current operating conditions The calculation formula is as follows: ; The turns ratio of the high-voltage and low-voltage windings. The characteristic impedance of the resonant cavity, The resonant frequency, For maximum operating frequency, Magnetizing inductor With resonant inductor The ratio; S2, Soft Start Phase 1: The CLLC converter is started, entering the first stage of soft start: the frequency is fixed at the maximum frequency, and the duty cycle is varied using a Sigmoid curve. The Sigmoid function is used to achieve the characteristics of gradual rise, acceleration, and gradual rise, increasing the duty cycle from the initial duty cycle to the corresponding maximum duty cycle, thus completing the first stage of variable duty cycle soft start. The duty cycle changes over time as follows: ; The initial duty cycle, This represents the duty cycle when the dead time is minimized, with a value ranging from 0.42 to 0.45, varying with frequency. This is a coefficient of variation used to adjust the rate of change in the middle section of the S-curve and to slow down the rate of increase in the final segment of the S-curve. This is the soft start deadline; The PWM generator uses forced complementary output, and the rising edge of the upper and lower transistors is inserted into the dead time to prevent shoot-through. By using a fixed comparison value, i.e. half a cycle, in conjunction with dead time adjustment, the PWM duty cycle of the upper and lower transistors is kept consistent, so as to achieve symmetrical peak-to-peak values ​​of positive and negative currents in the resonant cavity. S3, Soft Start Phase Two: After the first stage, the second stage of soft start begins: The duty cycle remains unchanged from the final value of the first stage. The frequency is designed using an inverse proportional function to achieve a non-linear frequency decrease, ensuring a smooth transition of the output gain. This also effectively reduces the soft start time of the second stage. The inverse proportional function for frequency change is: ; The maximum frequency value, This is the minimum frequency value allowed for down-frequency soft-start. This is the slope coefficient of the curve, used to adjust the rate of change in the middle section of the S-curve. This is the central time point used to adjust the total duration of the second-stage soft start; S4. Closed-loop soft start (this step is the third stage of soft start): Using the switching frequency at the end of the second stage as the initial frequency of the closed loop, the initial value transfer from the open loop to the closed loop is realized. Based on the maximum current value of the external command, the current command is slowly increased using closed-loop control, so that the output current changes from... The current is smoothly increased to the target current. Once the output current reaches the target current and stabilizes, the soft start ends and the system enters normal closed-loop operation. In steps S1 to S4, after the whole machine is connected to the mains power, the system is in standby mode. After receiving the charging command from the external BMS or host computer, it starts to switch to forward charging mode. When the rectified voltage of the front-end PFC reaches the specified value, the PFC is automatically started and CLLC charging begins. CLLC charging consists of two parts. The first part is segmented soft start, i.e. steps S2-S3. The second part is normal closed-loop control. After the segmented soft start is completed, it will switch to normal closed-loop control, i.e. step S4, and become normal charging. It will enter standby mode after the battery is fully charged or end after receiving an external shutdown command. In summary, the CLLC forward charging process is as follows: standby state, receiving charging command, PFC voltage stabilization, entering segmented soft start, soft start completion, normal closed-loop charging, battery fully charged or shutdown command, standby. Example 1:

[0019] The PFC is connected to the mains power input, and the rectified output voltage is used as the primary bus voltage of the CLLC. The secondary output voltage and current of the CLLC charge the external battery. The working status of the entire OBC system is controlled and set by the external BMS or host computer. The CLLC single forward charging process is as follows: Figure 2 As shown, after the whole machine is connected to the mains power, the system is in standby mode. After receiving the charging command from the external BMS or host computer, it starts to switch to forward charging mode. After the front-end PFC rectified voltage reaches the specified value, the PFC starts up by default and CLLC charging begins. The entire soft-start workflow is as follows: Figure 3 As shown in the figure, the segmented soft-start process for CLLC forward charging consists of three steps. The specific function and implementation method of each step are as follows: Step 1: During the pre-start phase, the controller first adaptively calculates the minimum constant current value under the current bus voltage and battery voltage. The purpose is to calculate the current value at the maximum operating frequency and maximum duty cycle under this condition, i.e., to fix the minimum gain. This current value ensures that the CLLC remains in a continuous, uninterrupted state at the end of the second-stage soft start and before entering the closed-loop soft start phase. This prevents abnormal closed-loop control caused by incomplete duty cycle soft start or sudden current changes during intermittent states, which could lead to sudden changes in resonant cavity current and charging abnormalities. This strategy effectively prevents such situations. The relationship between the minimum constant current value and the battery voltage value is as follows: ; In the above function, The input bus voltage is the primary bus voltage of the CLLC. This is the secondary battery voltage. The turns ratio of the high-voltage and low-voltage windings. The characteristic impedance of the resonant cavity, The resonant frequency, For maximum operating frequency, , for excitation inductor With resonant inductor The ratio; Based on the function described above, substitute the relevant parameters for this example: = ; The range is ; =35.2uH; =5; =130KHz; =250KHz; =4, substituting the above parameters into the relationship function between the minimum constant current value and the battery voltage value, the calculated result is approximately 8A. Figure 4 The measured values ​​are consistent; The final adaptive current-voltage variation curve is shown below. Figure 4 As shown in the figure, after the bus voltage value is adaptively matched to the current battery voltage value, the minimum constant current value of the controller's soft start second stage will decrease as the battery voltage increases. In the first stage of soft boot, the frequency is fixed at the maximum frequency. The duty cycle variation uses a Sigmoid curve, employing the Sigmoid function to achieve gradual rise, acceleration, and gradual rise characteristics, increasing from the initial duty cycle to... The corresponding maximum duty cycle is then used to complete the first stage of variable duty cycle soft start. The duty cycle changes over time as follows: ; In the inverse function mentioned above, the relevant parameters in this example are: The initial duty cycle is 0.08. This represents the duty cycle when the dead time is minimized, with a value ranging from 0.42 to 0.45, varying with frequency. This is a coefficient of variation used to adjust the rate of change in the middle section of the S-curve and to slow down the rate of increase in the final segment of the S-curve. The soft start cutoff time is 20ms; During the first stage of soft start, the duty cycle versus time curve is shown in the figure below. Figure 5 As shown, this method can reduce the soft-start time of the first stage, thereby further reducing the overall soft-start time, achieving a balance between efficiency and stability while suppressing the opening-tube impact and maintaining a relatively fast start-up speed. Step Two: In the second stage of soft start, the duty cycle remains constant, and the frequency is designed using an inverse proportional function to achieve a non-linear frequency decrease, ensuring a smooth transition in output gain and effectively reducing the soft start time of the second stage. The inverse proportional function for frequency change is: ; In the inverse function mentioned above, the relevant parameters in this example are: The maximum frequency value is 250kHz. The minimum allowed frequency for soft-start at reduced frequency is 160kHz. The slope coefficient of the curve is 300, used to adjust the rate of change in the middle section of the S-curve. The central time point of 40ms is used to adjust the total duration of the second-stage soft start. The frequency versus time curve during the second-stage soft start is shown in the figure below. Figure 6 As shown in the figure, the frequency remains constant within 0-20ms, and the duty cycle changes in an S-shaped curve, showing a trend of fast in the middle and slow at both ends. Within 20-60ms, the frequency changes in an inverse function trend, consistent with the gain curve. Step 3: Closed-loop soft start begins. The initial frequency is the end frequency of the open-loop frequency reduction at the previous moment. Since the switching point used in Step 2 is the minimum constant current point obtained by adaptive calculation, the end frequency value is in a stable changing state before switching. This ensures a smooth transition from open-loop soft start to closed-loop soft start without voltage or current jitter. Then, based on the maximum current value set by the external command, the closed-loop output current is given to gradually increase or decrease soft start. The purpose is to further smooth the output current waveform. The resonant cavity current will increase slowly without current spikes. At the same time, this function does not rely on the dynamic recovery capability of the front-end PFC, which can greatly reduce the bus voltage jitter during the start-up process, thereby further reducing the smooth increase of charging power at the same frequency. Meanwhile, the current increases gradually, which can reduce the impact on the battery and extend the battery's lifespan. Figure 7 The figure shows the experimental waveforms of the soft-start (primary-side drive waveform and secondary-side resonant cavity current) scheme of the bidirectional CLLC resonant converter. As can be seen from the figure, after using the segmented soft-start scheme of this invention, the secondary-side resonant cavity current rises smoothly without current spikes during the adaptive change of frequency and duty cycle. The startup process is relatively rapid, and the positive and negative current values ​​remain symmetrical. Finally, it stably enters the closed-loop system, proving the rationality of this scheme.

[0020] Based on the above description, this invention achieves adaptive parameter adjustment and multi-stage disturbance-free switching during the startup process, effectively solving several major problems faced by traditional methods. First, regarding the problem of parameter self-tuning, this method adopts a dynamic parameter tuning algorithm based on model prediction. By monitoring key parameters such as input and output voltage and load conditions at the startup moment in real time, it automatically adjusts the minimum constant current point under different startup conditions, thereby achieving stability of soft startup under each condition. Second, regarding the oscillation problem during stage switching, a feedforward compensation mechanism and gradual change in output setpoint ensure a smooth transition of the control mode. The segmented soft-start strategy can solve the problems of sudden gain change at the end of the soft-start and long soft-start time in the initial fixed slope frequency reduction strategy, making the gain of the entire soft-start process closer to the gain curve of the converter, thereby further reducing the peak current of the resonant cavity.

[0021] In step S1, the current data is collected. and In addition, battery status data is obtained via BMS: temperature Voltage fluctuation during the last three charging cycles, from -40℃ to 60℃. Number of loops And store it in the historical database; Load characteristic models are trained based on machine learning algorithms such as BP neural networks. ; in, This is the result of the original static calculation formula. The correction factor is dynamic. At low temperatures and high cycle counts, the correction factor increases by 1.2-1.5 times to compensate for the increased current demand caused by the increased internal resistance. Corrected predictions As the second-stage switching threshold, the coefficient of change of the first-stage Sigmoid curve is adjusted in advance. When load fluctuations exceed a preset threshold, it is considered a large fluctuation. Reduce by 20% to slow the growth rate of duty cycle; Based on the above description, by introducing a load dynamic prediction algorithm, historical battery data, namely voltage fluctuation rate, temperature, and cycle number, are collected in real time during the pre-start-up phase. A load characteristic model is established, and the minimum constant current value and stage switching threshold are pre-tuned. This achieves dual-dimensional parameter optimization of static parameters and dynamic prediction, in order to solve the problem that the dynamic load characteristics of the battery, such as sudden changes in internal resistance at low temperatures and capacity decay of aging batteries, may lead to lag in the adaptation of start-up parameters under extreme operating conditions.

[0022] Example 2:

[0023] For electric vehicle batteries aged at low temperatures (-20℃), with 1200 charge cycles, the voltage fluctuation during the last three charging cycles was analyzed. The effectiveness of the dynamic parameter pre-tuning mechanism was verified in the forward charging start-up scenario. The controller collects the following parameters in real time (corresponding to step S1, the pre-start phase): Primary bus voltage = Current voltage of the secondary battery = Battery status data, including temperature, is obtained through the BMS. =-20℃, voltage fluctuation during the last 3 charging cycles Far exceeding the normal fluctuation threshold Number of loops The aging threshold is 800 cycles; Load model establishment and dynamic correction coefficient calculation: Based on the preset BP neural network load characteristic model, the dynamic correction coefficient is calculated by inputting the above parameters: Static minimum constant current value The result calculated using the original formula is approximately equal to... ; Dynamic correction coefficient : because =-20℃, low temperature correction factor 1.3; The fluctuation correction factor is 1.2. The aging correction factor is 1.4; the overall correction factor is: The result is a weighted average to eliminate overcorrection; Predicted minimum constant current value after correction: ; Parameter pre-tuning and execution: Second-stage switching threshold update: Set the current judgment threshold at the end of the second stage to adapt to low temperature and high internal resistance scenarios. Phase 1: Adjusting Sigmoid curve parameters due to large load fluctuations. The duty cycle variation coefficient The duty cycle rate is reduced by 20% from 0.3 to 0.24, slowing down the rate of increase. The duty cycle function is adjusted as follows: ; The technical effectiveness verification is shown in the table below:

[0024] This embodiment solves the parameter adaptation lag problem caused by sudden changes in internal resistance during the startup of low-temperature aged batteries by dynamically predicting battery load characteristics and pre-tuning parameters. It verifies the effectiveness of the dual-dimensional optimization of static calculation and dynamic prediction, and significantly improves the startup stability under extreme conditions.

[0025] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter, implemented based on an on-board charger (OBC) system. The OBC system includes a front-stage totem-pole PFC rectifier circuit, a CLLC charging / discharging circuit, and an auxiliary power supply circuit. The PFC output voltage serves as the primary bus voltage of the CLLC, and the CLLC secondary output charges the battery. The system is controlled by an external BMS or a host computer. Its features include: Includes the following steps: S1. Pre-startup and adaptive calculation of minimum constant current value: After receiving an external charging command and the PFC rectified voltage stabilizes, the controller enters the pre-start phase: Real-time acquisition of CLLC primary bus voltage and secondary battery voltage The minimum constant current value under the current operating condition is adaptively calculated based on the resonant cavity parameters. ; S2, Soft Start Phase 1: Start the CLLC converter and enter the first stage of soft start: the frequency is fixed at the maximum frequency, the duty cycle is changed using the Sigmoid curve, and the Sigmoid function is used to realize the characteristics of gradual rise, acceleration and gradual rise. The duty cycle is increased from the initial duty cycle to the corresponding maximum duty cycle, thus completing the first stage of variable duty cycle soft start. S3, Soft Start Phase Two: After the first stage ends, the second stage of soft start begins: the duty cycle remains unchanged at the final value of the first stage, and the frequency is designed using an inverse proportional function to achieve a non-linear frequency decrease, ensuring a smooth transition of the output gain, while effectively reducing the soft start time of the second stage. S4, Closed-loop soft start: This step is the third stage of soft start. Using the switching frequency at the end of the second stage as the initial closed-loop frequency, it achieves the initial value transfer from open loop to closed loop. Based on the maximum current value of the external command, closed-loop control is used to slowly increase the current command, causing the output current to gradually increase from the external current value. The current is smoothly increased to the target current. Once the output current reaches the target current and stabilizes, the soft start ends and the system enters normal closed-loop operation.

2. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In step S1, the external charging command includes start / stop, maximum charging voltage, and current.

3. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In step S1, the minimum constant current value under the current operating condition The calculation formula is as follows: ; The turns ratio of the high-voltage and low-voltage windings. The characteristic impedance of the resonant cavity, The resonant frequency, For maximum operating frequency, Magnetizing inductor With resonant inductor The ratio of .

4. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In step S2, the duty cycle changes with time as follows: ; The initial duty cycle, This represents the duty cycle when the dead time is minimized, with a value ranging from 0.42 to 0.45, varying with frequency. This is a coefficient of variation used to adjust the rate of change in the middle section of the S-curve and to slow down the rate of increase in the final segment of the S-curve. This is the soft start deadline; In step S2, the PWM generator adopts forced complementary output, and the rising edge of the upper and lower transistors is inserted into the dead time to prevent shoot-through. By using a fixed comparison value, i.e. half a cycle, in conjunction with dead time adjustment, the PWM duty cycle of the upper and lower transistors is kept consistent, thereby achieving symmetrical peak-to-peak values ​​of the positive and negative currents in the resonant cavity.

5. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In step S3, the inverse proportional function of the frequency change is: ; The maximum frequency value, This is the minimum frequency value allowed for down-frequency soft-start. This is the slope coefficient of the curve, used to adjust the rate of change in the middle section of the S-curve. This is the central time point used to adjust the total duration of the second-stage soft start.

6. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In steps S1 to S4, after the whole machine is connected to the mains power, the system is in standby mode. After receiving the charging command from the external BMS or host computer, it starts to switch to forward charging mode. When the rectified voltage of the front-end PFC reaches the specified value, the PFC is automatically started and CLLC charging begins.

7. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 6, characterized in that: CLLC charging consists of two main parts. The first part is segmented soft start, i.e., steps S2-S3. The second part is normal closed-loop control. After the segmented soft start is completed, it will switch to normal closed-loop control, i.e., step S4, and become normal charging. It will enter standby mode after the battery is fully charged or end after receiving an external shutdown command.

8. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: The CLLC forward charging process is as follows: standby state, receiving charging command, PFC voltage stabilization, entering segmented soft start, soft start completion, normal closed-loop charging, battery fully charged or shutdown command, standby.

9. The segmented soft-start control method for forward charging of a bidirectional CLLC resonant converter according to claim 1, characterized in that: In step S1, the current data is collected. and In addition, battery status data is obtained via BMS: temperature Voltage fluctuations in the last three charging cycles Number of loops And store it in the historical database; Load characteristic models are trained based on machine learning algorithms such as BP neural networks. ; in, This is the result of the original static calculation formula. The correction factor is dynamic. At low temperatures and high cycle counts, the correction factor increases by 1.2-1.5 times to compensate for the increased current demand caused by the increased internal resistance. Corrected predictions As the second-stage switching threshold, the coefficient of change of the first-stage Sigmoid curve is adjusted in advance. When load fluctuations exceed a preset threshold, it is considered a large fluctuation. Reduce by 20% to slow the growth rate of duty cycle.

Images