Dual active bridge extended phase-shift peak current optimization method, system, medium, and product

Abstract

This invention discloses a method, system, medium, and product for optimizing the peak current of a dual active bridge extended phase-shift converter, belonging to the field of power electronics control. The method involves: obtaining the voltage conversion ratio and internal / external phase-shift duty cycles of the dual active bridge converter; constructing an inductor current peak function under extended phase-shift modulation, a penalized power balance equation, and a zero-voltage turn-on inequality constraint; initializing a particle swarm with the internal / external phase-shift duty cycles as particle positions; constructing a fitness function based on the minimum current peak value and constraints; iteratively solving for the optimal phase-shift parameters using a particle swarm optimization algorithm; and dynamically adjusting the constraint penalty weights based on the particle violation ratio during iteration; finally, generating a modulation signal based on the optimal phase-shift parameters to control the switching transistors, completing the optimization. By implementing this invention, the problem of poor convergence stability of existing optimization algorithms under wide operating conditions can be solved.

Description

Technical Field

[0001] This invention belongs to the field of power electronic control and relates to a method, system, medium, and product for optimizing the peak current of extended phase shifting of dual active bridges. Background Technology

[0002] Dual active bridge converters, as the core power conversion unit for flexible DC interconnect systems and energy storage interfaces, are widely used in DC microgrids and power electronic transformers due to their bidirectional power transmission capability, high-frequency electrical isolation characteristics, and high power density. Under extended phase-shift modulation strategies, by independently adjusting the duty cycles of the inner and outer phase shifters, flexible power control and soft-switching operation of devices can be achieved over a wide voltage ratio and power range, which is a key technical means to improve system efficiency.

[0003] Existing technologies for solving extended phase-shift parameter optimization problems typically construct a mathematical model that includes current stress objectives and power balance constraints. They then employ a penalty function method to transform the constraints into penalty terms, which are then incorporated into the objective function for optimization calculations. Some schemes also combine analytical or iterative methods for secondary correction. However, such constraint handling mechanisms based on fixed penalty coefficients lack responsiveness to the degree of solution group violation during the iterative search process. This leads to situations where the penalty intensity is difficult to match the current convergence state under different voltage matching ratios or load fluctuations. It is highly susceptible to situations where the penalty coefficient is set too small, causing the search process to fall into the infeasible region, or set too large, causing the search to converge prematurely to a local optimum. Consequently, the optimization results lack sufficient convergence robustness under complex operating conditions. Summary of the Invention

[0004] This application provides a method, system, medium, and product for optimizing the peak current of extended phase-shifted dual active bridges, which can solve the problem of poor convergence stability of optimization algorithms under wide operating conditions in existing technologies.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for optimizing the extended phase-shifted peak current of a dual active bridge, comprising: Obtain the electrical parameters of the dual active bridge converter; wherein, the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data, and outer phase shift duty cycle data; Based on the electrical parameters, an inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint under extended phase-shift modulation are constructed; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty. Using the inner and outer phase shift duty cycle data, a combination of phase shift parameters is constructed as the particle position vector to initialize the particle swarm. A fitness function is constructed using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint. With the goal of minimizing the inductor current peak value, the optimal combination of phase shift parameters is obtained through iterative solution using the particle swarm algorithm. In each iteration, the weight coefficients in the equality constraint penalty and inequality constraint penalty are adjusted according to the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint. Based on the optimal combination of phase-shifting parameters, a pulse width modulation signal is generated and the switching transistors of the dual active bridge converter are controlled to complete the optimization.

[0006] Compared with existing technologies, the embodiments of this application have the following beneficial effects: By acquiring the voltage conversion ratio data and the inner and outer phase shift duty cycle data of the dual active bridge converter, the basis of multi-dimensional control variables under extended phase shift modulation is established, providing a complete input dimension for parameter optimization under wide operating conditions; Based on the electrical parameters, inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint are constructed, and the latter two are configured with equality and inequality penalties respectively, transforming the physical-level current stress minimization objective and power transmission accuracy and soft-switching reliability requirements into a calculable mathematical optimization model, ensuring that the optimization process is carried out under the premise of meeting the engineering hard indicators; By constructing the inner and outer phase shift duty cycle data as particle position vectors and initializing the particle swarm, the swarm intelligence search mechanism is used to replace the traditional analytical method or random initialization method, avoiding the risk of the algorithm getting trapped in local optima and improving the global efficiency. The algorithm possesses superior optimization capabilities. During each iteration, it dynamically adjusts the weight coefficients in the equality and inequality penalties based on the proportion of particles violating the power balance equality constraint and the zero-voltage turn-on inequality constraint in the current particle swarm. This feedback mechanism establishes a closed-loop adjustment logic of "constraint violation degree—penalty intensity—search direction," enabling the algorithm to adaptively enhance the repulsion force against infeasible regions or weaken excessive constraints on feasible regions based on the real-time state of the population. This solves the stability problem of existing technologies, where fixed penalty weights cause algorithm convergence oscillations, easy entrapment into local optima, or inability to converge to feasible regions under complex operating conditions such as wide voltage ratios and wide loads. Finally, based on the optimal combination of phase-shifting parameters, a pulse width modulation signal is generated to control the switching transistors, achieving low-current stress operation and zero-voltage turn-on of all devices in the dual active bridge converter across the entire operating range, significantly improving the system's transmission efficiency and operational reliability.

[0007] In some embodiments of the first aspect of this application, obtaining the electrical parameters of the dual active bridge converter includes: Within the feasible solution space of the particle swarm optimization algorithm, initialize the initial values ​​of the inner phase shift duty cycle data and the outer phase shift duty cycle data; Collect primary-side input voltage data, secondary-side output voltage data, and turns ratio data of the high-frequency isolation transformer from the dual active bridge converter; The voltage conversion ratio is calculated based on the primary input voltage data, secondary output voltage data, and turns ratio data.

[0008] Compared with existing technologies, the above embodiments have the following advantages: By initializing the initial values ​​of the inner and outer phase shift duty cycle data within the feasible solution space of the particle swarm optimization algorithm, the search boundary and initial distribution of the optimization variables are clearly defined, preventing computational divergence or invalid searches caused by the initial solution exceeding the physical feasible domain; by collecting the primary input voltage data, secondary output voltage data, and turns ratio data of the high-frequency isolation transformer of the dual active bridge converter, the core electrical quantities reflecting the real-time operating status of the system are directly obtained, providing an accurate physical basis for subsequent calculations; the voltage conversion ratio data is calculated based on the above voltage and turns ratio data, quantifying the degree of primary and secondary voltage matching, eliminating model mismatch errors caused by voltage fluctuations, and enabling the subsequently constructed current peak function and constraint conditions to accurately follow the changes in operating conditions, thereby improving the adaptability and computational accuracy of the optimization model under different voltage transfer ratios.

[0009] In some embodiments of the first aspect of this application, constructing the peak inductor current function under extended phase-shift modulation based on the electrical parameters includes: Based on the relative magnitudes of the inner and outer phase shift duty cycle data, half a switching cycle is divided into various operating modes; For each operating mode, the instantaneous value of the inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits, and the maximum current amplitude is extracted from the instantaneous value of the inductor current at each switching moment to construct the peak inductor current function.

[0010] Compared with existing technologies, the above embodiments have the following beneficial effects: By dividing half a switching cycle into various working modes according to the relative magnitude of the inner and outer phase shift duty cycle data, the logical classification of complex waveform timing under extended phase shift modulation ensures the accuracy of circuit topology state identification under different modes; for each working mode, the instantaneous value of inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits, restoring the true physical trajectory of inductor current at the moment of switching action, providing high-resolution time-domain data for stress assessment; the maximum current amplitude is extracted from the instantaneous value of inductor current at each switching moment and an inductor current peak function is constructed, mapping discrete current sampling points to continuous optimization target indicators, directly quantifying the ultimate current stress borne by power devices, enabling the optimization algorithm to locate and suppress current spikes, thereby effectively reducing device heat loss and improving system safety margin.

[0011] In some embodiments of the first aspect of this application, constructing the power balance equation constraint under extended phase-shift modulation based on the electrical parameters includes: Based on the power transmission mechanism of the dual active bridge converter under extended phase-shift modulation, a functional relationship is established between the transmitted power and the inner phase-shift duty cycle data and the outer phase-shift duty cycle data. The condition for power balance equation constraint is that the difference between the transmitted power and the preset active power command reference value is equal to zero. The absolute value of the actual difference is taken as the default amount and multiplied by the preset equation constraint penalty weight coefficient as the equation constraint penalty term.

[0012] Compared with existing technologies, the above embodiments have the following beneficial effects: By establishing a functional relationship between the transmitted power and the inner and outer phase shift duty cycle data based on the power transmission mechanism of the dual active bridge converter under extended phase shift modulation, an analytical model describing the energy transmission characteristics is constructed, clarifying the nonlinear mapping law between the control variables and the output power; the difference between the transmitted power and the preset active power command reference value being equal to zero is used as the condition for the power balance equation constraint, strictly limiting the energy conservation principle that the optimization solution must satisfy, and preventing power tracking deviation; the actual absolute value of the difference is used as the violation quantity and multiplied with the preset equation constraint penalty weight coefficient to obtain the equation constraint penalty term, transforming the rigid power balance requirement into a gradient-descent-capable soft constraint cost, so that the subsequent particle swarm algorithm can automatically sense and correct the power deviation during the search process, ensuring that the final output phase shift parameter combination can accurately maintain the power supply and demand balance of the DC microgrid.

[0013] In some embodiments of the first aspect of this application, constructing the zero-voltage turn-on inequality constraint based on the electrical parameters includes: Based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the polarity requirements of the inductor current corresponding to each switching device at the moment of turn-on are transformed into inequality constraint functions. The negative of the function value that does not meet the polarity requirements of the inductor current is taken as the default value and multiplied by the preset inequality constraint penalty weight coefficient as the inequality constraint penalty term.

[0014] Compared with existing technologies, the above embodiments have the following beneficial effects: By transforming the inductor current polarity requirements of each switching device at the moment of turn-on into inequality constraint functions based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the abstract soft-switching physical mechanism is transformed into a computable mathematical criterion, providing a quantitative criterion for the soft-switching operation of all devices; the negative of the function value that does not meet the inductor current polarity requirement is used as the default quantity and multiplied by the preset inequality constraint penalty weight coefficient to obtain the inequality constraint penalty term, thus constructing a dynamic penalty mechanism for the risk of soft-switching failure. This allows the optimization process to pursue low current stress while forcibly guiding the search direction to avoid parameter regions that would lead to increased hard-switching losses, thereby expanding the zero-voltage turn-on operation range of the system while ensuring efficiency optimization, and avoiding a surge in switching losses under light load or high voltage ratio conditions.

[0015] In some embodiments of the first aspect of this application, constructing the fitness function using the inductor current peak function, the power balance equation constraint, and the zero-voltage turn-on inequality constraint includes: The fitness function is obtained by adding the peak inductor current function, the equality constraint penalty term in the power balance equation constraint, and the inequality constraint penalty term in the zero-voltage turn-on inequality constraint.

[0016] Compared with existing technologies, the above embodiments have the following beneficial effects: By adding the inductor current peak function, the equality constraint penalty term in the power balance equality constraint, and the inequality constraint penalty term in the zero-voltage turn-on inequality constraint to obtain the fitness function, a unified evaluation system containing multiple optimization objectives and constraints is constructed; This comprehensive fitness function integrates the three mutually coupled or even conflicting control objectives of minimizing current stress, maintaining power balance, and ensuring soft-switching operation into a single scalar index, enabling the particle swarm algorithm to directly judge the quality of particles by comparing fitness values, achieving synchronous convergence of multi-objective collaborative optimization, avoiding the logical fragmentation and suboptimal solution problems caused by step-by-step optimization, and ensuring that the final phase-shifting parameter combination achieves a globally optimal balance among current stress, power accuracy, and soft-switching performance.

[0017] In some embodiments of the first aspect of this application, adjusting the weighting coefficients in the equality constraint penalty and the inequality constraint penalty based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and the zero voltage turn-on inequality constraint includes: Count the number of first particles that do not satisfy the power balance equation constraint and the number of second particles that do not satisfy the zero voltage turn-on inequality constraint in the current iteration. Calculate the ratios of the first and second particle counts to the total size of the particle swarm to obtain the first ratio and the second ratio. Based on the first ratio and the second ratio respectively, the weight coefficients in the equality constraint penalty and the inequality constraint penalty are adjusted linearly with positive correlation.

[0018] Compared with existing technologies, the above embodiments have the following beneficial effects: By statistically analyzing the number of first particles that do not satisfy the power balance equation constraint and the number of second particles that do not satisfy the zero-voltage turn-on inequality constraint in the current iteration, the overall default scale of the current population in the power dimension and soft-switching dimension is quantified in real time, providing accurate data support for adaptive adjustment; the ratios of the number of first particles and the number of second particles to the total size of the particle swarm are calculated to obtain the first ratio and the second ratio, respectively, normalizing the absolute number of defaults into a relative probability index reflecting the convergence state of the population, eliminating the influence of the particle swarm size on the adjustment strategy, and enhancing the universality of the algorithm; based on the first ratio and the second ratio, the weight coefficients in the equality penalty and inequality penalty are linearly adjusted with positive correlation, respectively, establishing a linear growth mapping relationship between the penalty intensity and the default ratio, so that when the population default ratio is high, the penalty weight is automatically increased to accelerate convergence to the feasible region, while when the default ratio is low, the penalty weight is appropriately reduced to retain the global search capability. This dynamic linear adjustment mechanism significantly improves the convergence robustness of the algorithm under wide operating condition perturbations and effectively overcomes the convergence stagnation or oscillation defects caused by fixed weights.

[0019] Secondly, the present invention also provides a dual active bridge extended phase-shifting peak current optimization system, comprising: a data acquisition module, a configuration module, a solution module and an output module; The data acquisition module is used to acquire the electrical parameters of the dual active bridge converter; wherein the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data, and outer phase shift duty cycle data; The configuration module is used to construct, based on the electrical parameters, the peak function of inductor current under extended phase-shift modulation, the power balance equation constraint, and the zero-voltage turn-on inequality constraint; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty. The solution module is used to construct a combination of phase-shifting parameters as a particle position vector and initialize the particle swarm using the inner and outer phase-shifting duty cycle data. It then constructs a fitness function using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint. With the goal of minimizing the inductor current peak value, it iteratively solves the problem using a particle swarm optimization algorithm to obtain the optimal combination of phase-shifting parameters. In each iteration, the weighting coefficients in the equality constraint penalty and inequality constraint penalty are adjusted based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint. The output module is used to generate a pulse width modulation signal based on the optimal phase shift parameter combination and to regulate the switching transistors of the dual active bridge converter to complete the optimization.

[0020] Compared with the prior art, the above embodiments of this application have the following beneficial effects: By acquiring the voltage conversion ratio data and the inner and outer phase shift duty cycle data of the dual active bridge converter, the basis of multi-dimensional control variables under extended phase shift modulation is established, providing a complete input dimension for parameter optimization under wide operating conditions; Based on the electrical parameters, the inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint are constructed, and the latter two are configured with equality and inequality penalties respectively, transforming the physical-level current stress minimization objective and power transmission accuracy and soft-switching reliability requirements into a calculable mathematical optimization model, ensuring that the optimization process is carried out under the premise of meeting the engineering hard indicators; By constructing the inner and outer phase shift duty cycle data as particle position vectors and initializing the particle swarm, the swarm intelligence search mechanism is used to replace the traditional analytical method or random initialization method, avoiding the risk of the algorithm getting trapped in local optima, and improving the overall efficiency. The algorithm possesses strong local optimization capabilities. During each iteration, it dynamically adjusts the weight coefficients in the equality and inequality penalties based on the proportion of particles violating the power balance equality constraint and the zero-voltage turn-on inequality constraint in the current particle swarm. This feedback mechanism establishes a closed-loop adjustment logic of "constraint violation degree—penalty intensity—search direction," enabling the algorithm to adaptively enhance the repulsion force against infeasible regions or weaken excessive constraints on feasible regions based on the real-time state of the population. This solves the stability problem of existing technologies, where fixed penalty weights cause algorithm convergence oscillations, easy entrapment into local optima, or inability to converge to feasible regions under complex operating conditions such as wide voltage ratios and wide loads. Finally, based on the optimal phase-shift parameter combination, a pulse width modulation signal is generated to control the switching transistors, achieving low-current stress operation and zero-voltage turn-on of all devices in the dual active bridge converter across the entire operating range, significantly improving the system's transmission efficiency and operational reliability.

[0021] Thirdly, the present invention also provides a computer program product, including a computer program or instructions, characterized in that, when the computer program or instructions are executed, they implement any one of the dual active bridge extended phase-shifting peak current optimization methods of the present invention.

[0022] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any one of the dual active bridge extended phase-shifting peak current optimization methods of the present invention. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a dual active bridge extended phase-shifting peak current optimization method provided in some embodiments of the present invention.

[0024] Figure 2 This is a schematic diagram of a dual active bridge extended phase-shifting peak current optimization system provided in some embodiments of the present invention. Detailed Implementation

[0025] 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.

[0026] Example 1: Please refer to Figure 1 To address the problem of poor convergence stability of existing optimization algorithms under wide operating conditions, an embodiment of the present invention provides a dual active bridge extended phase-shifting peak current optimization method, comprising steps S1 to S4: Step S1: Obtain the electrical parameters of the dual active bridge converter; wherein the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data, and outer phase shift duty cycle data.

[0027] Furthermore, step S1 can be implemented through the following preferred embodiments, including steps S11-S13, as follows: S11: Within the feasible solution space of the particle swarm optimization algorithm, initialize the initial values ​​of the inner phase shift duty cycle data and the outer phase shift duty cycle data; S12: Collect primary input voltage data, secondary output voltage data, and turns ratio data of the high-frequency isolation transformer of the dual active bridge converter; S13: Calculate the voltage conversion ratio data based on the primary input voltage data, secondary output voltage data, and turns ratio data.

[0028] In this preferred embodiment, by initializing the inner and outer phase shift duty cycle data within the feasible solution space of the particle swarm optimization algorithm, the search boundary and initial distribution of the optimization variables are clearly defined, preventing computational divergence or invalid searches caused by the initial solution exceeding the physical feasible region. The primary-side input voltage data, secondary-side output voltage data of the dual active bridge converter, and the turns ratio data of the high-frequency isolation transformer are collected, directly obtaining the core electrical quantities reflecting the real-time operating status of the system, providing accurate physical basis for subsequent calculations. The voltage transfer ratio data is calculated based on the aforementioned voltage and turns ratio data, quantifying the primary and secondary voltage matching degree, eliminating model mismatch errors caused by voltage fluctuations, and enabling the subsequently constructed current peak function and constraints to accurately follow changes in operating conditions, thereby improving the adaptability and computational accuracy of the optimization model under different voltage transfer ratios.

[0029] Step S2: Based on the electrical parameters, construct the inductor current peak function under extended phase-shift modulation, the power balance equation constraint, and the zero-voltage turn-on inequality constraint; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty.

[0030] Furthermore, step S2 can be implemented through the following preferred embodiments, including steps S21-S25, as detailed below: S21: Based on the relative magnitude of the inner phase shift duty cycle data and the outer phase shift duty cycle data, half a switching cycle is divided into various operating modes; S22: For each operating mode, the instantaneous value of the inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits, and the maximum current amplitude is extracted from the instantaneous value of the inductor current at each switching moment to construct the peak inductor current function.

[0031] In this preferred embodiment, half a switching cycle is divided into various operating modes based on the relative magnitude of the inner and outer phase shift duty cycle data. This logical classification of the complex waveform timing under extended phase shift modulation ensures the accuracy of circuit topology state identification under different modes. For each operating mode, the instantaneous value of the inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits. This reconstructs the true physical trajectory of the inductor current at the moment of switching action, providing high-resolution time-domain data for stress assessment. The maximum current amplitude is extracted from the instantaneous value of the inductor current at each switching moment, and an inductor current peak function is constructed. This maps discrete current sampling points to continuous optimization target indicators, directly quantifying the ultimate current stress that the power device can withstand. This allows the optimization algorithm to locate and suppress current spikes, thereby effectively reducing device heat loss and improving system safety margin.

[0032] S23: Based on the power transmission mechanism of the dual active bridge converter under extended phase-shift modulation, establish the functional relationship between the transmitted power and the inner phase-shift duty cycle data and the outer phase-shift duty cycle data; S24: The difference between the transmitted power and the preset active power command reference value is equal to zero as the condition for the power balance equation constraint. The absolute value of the actual difference is used as the default amount and multiplied by the preset equation constraint penalty weight coefficient as the equation constraint penalty term.

[0033] In this preferred embodiment, a functional relationship between the transmitted power and the inner and outer phase shift duty cycle data is established based on the power transmission mechanism of the dual active bridge converter under extended phase shift modulation. An analytical model describing the energy transmission characteristics is constructed, clarifying the nonlinear mapping law between the control variables and the output power. The condition that the difference between the transmitted power and the preset active power command reference value is equal to zero is used as the condition for the power balance equation constraint, strictly limiting the energy conservation principle that the optimization solution must satisfy to prevent power tracking deviation. The absolute value of the actual difference is used as the violation quantity and multiplied with the preset equation constraint penalty weight coefficient to obtain the equation constraint penalty term. The hard power balance requirement is transformed into a soft constraint cost that can be gradient-descented, so that the subsequent particle swarm algorithm can automatically sense and correct the power deviation during the search process, ensuring that the final output phase shift parameter combination can accurately maintain the power supply and demand balance of the DC microgrid.

[0034] S25: Based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the polarity requirements of the inductor current corresponding to each switching device at the moment of turn-on are transformed into inequality constraint function form. The negative of the function value that does not meet the polarity requirements of the inductor current is taken as the default quantity and multiplied by the preset inequality constraint penalty weight coefficient as the inequality constraint penalty term.

[0035] In this preferred embodiment, based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the inductor current polarity requirements corresponding to each switching device at the moment of turn-on are transformed into inequality constraint functions. This transforms the abstract soft-switching physical mechanism into a computable mathematical criterion, providing a quantitative criterion for the soft-switching operation of all devices. The negative of the function value that does not meet the inductor current polarity requirement is used as the default value and multiplied by the preset inequality constraint penalty weight coefficient to obtain the inequality constraint penalty term. This constructs a dynamic penalty mechanism for the risk of soft-switching failure, so that while pursuing low current stress, the optimization process forces the search direction to avoid parameter regions that will lead to increased hard-switching losses. This expands the zero-voltage turn-on operation range of the system while ensuring efficiency optimization, and avoids a surge in switching losses under light load or high voltage ratio conditions.

[0036] Step S3: Using the inner and outer phase shift duty cycle data, construct a combination of phase shift parameters as the particle position vector and initialize the particle swarm. Construct a fitness function using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint. With the goal of minimizing the inductor current peak value, iteratively solve the problem using the particle swarm algorithm to obtain the optimal combination of phase shift parameters. In each iteration, adjust the weight coefficients in the equality constraint penalty and inequality constraint penalty according to the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint.

[0037] Furthermore, in step S3, constructing the fitness function can be implemented through the following preferred embodiment, including step S31: S31: Add the inductor current peak function, the equality constraint penalty term in the power balance equality constraint, and the inequality constraint penalty term in the zero voltage turn-on inequality constraint to obtain the fitness function.

[0038] In this preferred embodiment, a fitness function is obtained by adding the peak inductor current function, the equality constraint penalty term in the power balance equation constraint, and the inequality constraint penalty term in the zero-voltage turn-on inequality constraint. This constructs a unified evaluation system containing multiple optimization objectives and constraints. This comprehensive fitness function integrates three mutually coupled and even conflicting control objectives—minimizing current stress, maintaining power balance, and ensuring soft-switching operation—into a single scalar index. This allows the particle swarm algorithm to directly determine the quality of particles by comparing fitness values, achieving synchronous convergence of multi-objective collaborative optimization. This avoids the logical fragmentation and suboptimal solution problems caused by step-by-step optimization, ensuring that the final phase-shifting parameter combination achieves a globally optimal balance among current stress, power accuracy, and soft-switching performance.

[0039] Furthermore, adjusting the weighting coefficients can be achieved through the following preferred implementation method, including steps S31-S33, as detailed below: S31: Count the number of first particles that do not satisfy the power balance equation constraint and the number of second particles that do not satisfy the zero voltage turn-on inequality constraint in the current iteration. S32: Calculate the ratios of the first particle number and the second particle number to the total size of the particle swarm, respectively, to obtain the first ratio and the second ratio; S33: Based on the first ratio and the second ratio respectively, adjust the weight coefficients in the equality constraint penalty and the inequality constraint penalty linearly with the positive correlation.

[0040] In this preferred embodiment, by counting the number of first particles that do not satisfy the power balance equation constraint and the number of second particles that do not satisfy the zero-voltage turn-on inequality constraint in the current round of iteration, the overall default scale of the current population in the power dimension and the soft-switching dimension is quantified in real time, providing accurate data support for adaptive adjustment; the ratios of the number of first particles and the number of second particles to the total scale of the particle swarm are calculated respectively to obtain the first ratio and the second ratio, normalizing the absolute default numbers into relative probability indicators reflecting the convergence state of the population, eliminating the influence of the particle swarm scale on the adjustment strategy, and enhancing the universality of the algorithm; based on the first ratio and the second ratio respectively, the weight coefficients in the positive-correlation linear adjustment equation penalty and inequality penalty are adjusted, establishing a linear growth mapping relationship between the penalty intensity and the default ratio, so that when the population default ratio is high, the penalty weight is automatically increased to accelerate the convergence to the feasible region, and when the default ratio is low, the penalty weight is moderately reduced to retain the global search ability. This dynamic linear adjustment mechanism significantly improves the convergence robustness of the algorithm under wide operating condition disturbances and effectively overcomes the defects of convergence stagnation or oscillation caused by fixed weights.

[0041] Step S4: Based on the optimal phase-shifting parameter combination, generate a pulse width modulation signal and regulate the switching tubes of the dual-active-bridge converter to complete the optimization.

[0042] In specific implementation, the present invention mainly uses extended phase-shift modulation as the control framework. On the premise of meeting the power transmission requirements and the zero-voltage turn-on constraints of all power devices, the internal phase-shift duty cycle and the external phase-shift duty cycle are jointly optimized by the particle swarm algorithm to minimize the peak current and broaden the zero-voltage turn-on interval within the full operating range.

[0043] First, calculate the voltage conversion ratio k according to the input voltage V1 and the output voltage V2 of the DAB converter; then establish an extended phase-shift modulation mathematical model. Under extended phase-shift modulation, the control variables include the internal phase-shift duty cycle D1 and the external phase-shift duty cycle D2. According to the size relationship between D1 and D2, the classification order of the inductor voltage within half of the switching period is divided into two major categories: Case A: D2 < D1 and Case B: D2 > D1. For any Case, since the arm voltage is piecewise constant within half a cycle and the inductor current iL(t) changes linearly in three segments within half a cycle, it can be expressed in segments in the intervals [t0, t1], [t1, t2], [t2, t3]. On this basis, each Case is further subdivided into 8 working modes by using the sign combination sgn(i(t0), i(t1), i(t2)) of the three segmented boundary currents for subsequent extraction of peak current candidate points and discrimination of the ZVS conditions of all switching devices.

[0044] Specifically, obtain the circuit parameters and operating conditions of the dual-active-bridge converter, including the input-side DC voltage V1, the output-side DC voltage V2, and the turns ratio n of the primary and secondary sides of the high-frequency transformer. Then, the voltage matching ratio k = V1 / n / V2, the primary-side full-bridge inverter AC port voltage Vh1, the equivalent transmission inductance L of the dual-active bridge, and the switching period or half-switching period ( ), the extended phase-shift control variables (the inner phase-shift duty cycle D1 and the outer phase-shift duty cycle D2), the equivalent on-state parameter Rdson of the power semiconductor device, and the system equivalent resistance Req are used to calculate the losses.

[0045] Here, taking D2 < D1 as an example, the basis for dividing its operating modes is as follows: The piecewise expression of its inductor current is as follows: .

[0046] Secondly, based on the piecewise analytical expression of the extended phase-shift modulation inductor current, further construct the functional mapping relationship between the control variables (D1, D2) and the system operating state, which is used to form the objective function and constraint condition system required for the optimization algorithm. This functional mapping relationship at least includes the following three types of analytical models: (1) Power analytical model: According to the power transmission mechanism of the dual-active-bridge converter under extended phase-shift modulation, establish the functional relationship between the DAB transmission power P and the phase-shift control variables D1 and D2: ; where represents the DAB switching frequency.

[0047] (2) Inductor current analytical model: Based on the piecewise linear current expression within half of the switching period, solve the analytical values of the inductor current at the piecewise boundary times t0, t1, and t2: i(t0), i(t1), and i(t2). These current values can all be expressed as functions of D1 and D2, and then the peak current function can be obtained.

[0048] (3) Zero-voltage turn-on criterion model: According to the physical conditions for the power devices of the dual-active bridge to achieve zero-voltage turn-on, convert the requirements for the polarity of the inductor current corresponding to each switching device at the instant of turn-on into an inequality constraint form: g(D1, D2) ≥ 0. The inequality constraint is used to ensure that during the optimization process, the direction of the inductor current at the turn-on moment of each switching device satisfies the conditions of reverse-parallel diode freewheeling or junction capacitance charge and discharge, so as to achieve zero-voltage turn-on of all devices.

[0049] Next, based on the above mathematical model, the instantaneous values ​​of the inductor current at each switching moment within half a switching cycle are extracted; the maximum current amplitude is defined as the peak current index, and this peak current is used as the optimization objective function to minimize the current stress, with the following constraints: Equality constraints: The combination of D1 and D2 is constrained by the power transfer model to ensure that the converter's output power meets the preset per-unit power requirement, as shown below: ;in, This indicates the reference value for the active power command transmitted by DAB.

[0050] Inequality constraint: Establish a zero-voltage turn-on constraint for all power devices, requiring that at the turn-on time of each switch, the inductor current direction satisfies the conduction condition of the reverse parallel diode, to ensure that all power devices achieve zero-voltage turn-on, as shown below: .

[0051] Finally, the particle swarm optimization algorithm is used to begin iterative optimization, and the steps are explained below: (1) Initialize the particle swarm within the feasible solution space that satisfies the constraints; (2) Particle position definition: The phase shift parameter combination (D1, D2) is used as the particle position vector; each (D1, D2) is used as a candidate control solution.

[0052] (3) Definition of particle search range: The physically feasible interval of extended phase-shift modulation is used as the boundary condition for variables, that is: The power feasible region is used as an additional search range constraint.

[0053] (4) Particle swarm size and initialization: The swarm size is set to the preset number of particles Np, and the particle position and velocity vectors are initialized in the feasible region in a random distribution manner.

[0054] (5) Fitness function definition: The optimization objective is to minimize the peak current, and the peak current function is defined as follows: As a fitness function; (6) Equality constraint embedding: based on power balance conditions As an equality constraint, if a particle does not satisfy the power constraint, a penalty term is constructed using the power error, and the penalty term is added to the fitness function to form a modified fitness.

[0055] (7) Inequality constraint embedding: The zero-voltage turn-on condition of all devices is used as the inequality constraint; the polarity of the current at the turn-on instant is used as the judgment basis; if a particle does not meet the ZVS condition, the particle is directly removed, or its fitness value is increased by the penalty function so that it does not participate in the optimal update.

[0056] (8) Velocity and position update mechanism: the best position in the history of a particle is the individual optimum; the best position in the history of the group is the global optimum; iterative updates are performed according to the velocity update formula of the particle swarm algorithm; the inertia weight is used as the velocity adjustment factor; and the learning factor is used as the search guidance parameter.

[0057] (9) Convergence determination: The fitness change is less than the preset threshold or the maximum number of iterations is reached as the termination condition; the optimal phase shift parameter combination that satisfies the constraint condition with the minimum peak current is output.

[0058] Regarding the punishment section, the details are as follows: During the particle swarm optimization (PSO) iteration process, particle positions that do not meet the constraints (i.e., combinations of D1 and D2) are handled according to the following penalty function: (1) The original form of the fitness function: The objective function is defined as the peak value of the inductor current: ; (2) Equality constraint penalty term: The power balance constraint Introducing equality constraints for breach of contract penalties: , where α is the equality constraint penalty weight coefficient.

[0059] (3) Inequality constraint penalty term: During the optimization process, for the case where the inductor current polarity violates the ZVS condition at the turn-on time of each power device, we define g(D1, D2)<0 and map it to the inequality constraint penalty. , where β is the inequality constraint penalty weight coefficient.

[0060] ; (4) Overall fitness correction formula: The original objective function is combined with each constraint penalty term to obtain the final modified fitness function. : ; During the iteration process, particles that do not meet the zero-voltage turn-on constraint are eliminated or penalized until the algorithm converges and outputs the optimal combination of phase shift parameters.

[0061] Furthermore, to improve the convergence speed and feasible solution search capability under different operating conditions, this invention introduces an adaptive adjustment mechanism for penalty weights. The expression for the adaptive adjustment formula can be as follows: The initial penalty weights are all defined to be greater than 0. As the iteration number t increases, they are adaptively adjusted as follows: ; ;in, This represents the number of particles that do not satisfy the equality constraint during the t-th iteration. This represents the number of particles in generation t that do not satisfy the ZVS inequality constraint. γ represents the particle swarm size; both γ and δ are greater than 0, representing the adaptive speed control coefficients.

[0062] The final penalty function consists of the objective function penalty term and the constraint penalty term, as follows: In summary, compared with the prior art, the above embodiments of this application have the following beneficial effects: By acquiring the voltage conversion ratio data and the inner and outer phase shift duty cycle data of the dual active bridge converter, the basis of multi-dimensional control variables under extended phase shift modulation is established, providing a complete input dimension for parameter optimization under wide operating conditions; Based on the electrical parameters, inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint are constructed, and the latter two are configured with equality and inequality penalties respectively, transforming the physical-level current stress minimization objective and power transmission accuracy and soft-switching reliability requirements into a calculable mathematical optimization model, ensuring that the optimization process is carried out under the premise of meeting the engineering hard indicators; By constructing the inner and outer phase shift duty cycle data as particle position vectors and initializing the particle swarm, the swarm intelligence search mechanism is used to replace the traditional analytical method or random initialization method, avoiding the risk of the algorithm getting trapped in local optima, and improving efficiency. The algorithm enhances global optimization capabilities. During each iteration, it dynamically adjusts the weight coefficients of equality and inequality penalties based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and the zero-voltage turn-on inequality constraint. This feedback mechanism establishes a closed-loop adjustment logic of "constraint violation degree - penalty intensity - search direction," enabling the algorithm to adaptively strengthen the repulsion force on the infeasible region or weaken the excessive constraint on the feasible region according to the real-time state of the population. This solves the stability problem of existing technologies, which suffer from convergence oscillations, easy getting trapped in local optima, or failure to converge to the feasible region due to fixed penalty weights under complex operating conditions such as wide voltage ratios and wide loads. Finally, based on the optimal phase-shift parameter combination, a pulse width modulation signal is generated to control the switching transistors, realizing low current stress operation and zero-voltage turn-on of all devices in the dual active bridge converter across the entire operating range, significantly improving the transmission efficiency and operational reliability of the system.

[0063] Example 2: Please refer to Figure 2 Based on the same inventive concept, the present invention discloses a dual active bridge extended phase-shifting peak current optimization system, comprising: a data acquisition module M1, a configuration module M2, a solution module M3, and an output module M4; The data acquisition module M1 is used to acquire the electrical parameters of the dual active bridge converter; the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data and outer phase shift duty cycle data.

[0064] Furthermore, the data acquisition module M1 includes: an initialization unit, a data acquisition unit, and a first calculation unit; The initial unit is used to initialize the initial values ​​of the inner phase shift duty cycle data and the outer phase shift duty cycle data within the feasible solution space of the particle swarm algorithm. The acquisition unit is used to acquire the primary input voltage data, secondary output voltage data, and turns ratio data of the high-frequency isolation transformer of the dual active bridge converter. The first calculation unit is used to calculate the voltage conversion ratio data based on the primary side input voltage data, secondary side output voltage data, and turns ratio data.

[0065] In this preferred embodiment, by initializing the inner and outer phase shift duty cycle data within the feasible solution space of the particle swarm optimization algorithm, the search boundary and initial distribution of the optimization variables are clearly defined, preventing computational divergence or invalid searches caused by the initial solution exceeding the physical feasible region. The primary-side input voltage data, secondary-side output voltage data of the dual active bridge converter, and the turns ratio data of the high-frequency isolation transformer are collected, directly obtaining the core electrical quantities reflecting the real-time operating status of the system, providing accurate physical basis for subsequent calculations. The voltage transfer ratio data is calculated based on the aforementioned voltage and turns ratio data, quantifying the primary and secondary voltage matching degree, eliminating model mismatch errors caused by voltage fluctuations, and enabling the subsequently constructed current peak function and constraints to accurately follow changes in operating conditions, thereby improving the adaptability and computational accuracy of the optimization model under different voltage transfer ratios.

[0066] The configuration module M2 is used to construct, based on the electrical parameters, the peak function of inductor current under extended phase-shift modulation, the power balance equation constraint, and the zero-voltage turn-on inequality constraint; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty.

[0067] Furthermore, the configuration module M2 includes: a mode division unit and a current extraction unit; The mode division unit is used to divide half a switching cycle into various working modes according to the relative size relationship between the inner phase shift duty cycle data and the outer phase shift duty cycle data. The current extraction unit is used to calculate the instantaneous value of the inductor current at each switching moment within half a switching cycle for each working mode using the principle of piecewise linear circuits, and to extract the maximum current amplitude from the instantaneous value of the inductor current at each switching moment to construct the peak inductor current function.

[0068] In this preferred embodiment, half a switching cycle is divided into various operating modes based on the relative magnitude of the inner and outer phase shift duty cycle data. This logical classification of the complex waveform timing under extended phase shift modulation ensures the accuracy of circuit topology state identification under different modes. For each operating mode, the instantaneous value of the inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits. This reconstructs the true physical trajectory of the inductor current at the moment of switching action, providing high-resolution time-domain data for stress assessment. The maximum current amplitude is extracted from the instantaneous value of the inductor current at each switching moment, and an inductor current peak function is constructed. This maps discrete current sampling points to continuous optimization target indicators, directly quantifying the ultimate current stress that the power device can withstand. This allows the optimization algorithm to locate and suppress current spikes, thereby effectively reducing device heat loss and improving system safety margin.

[0069] Furthermore, the configuration module M2 also includes: a power function construction unit and an equality constraint construction unit; The power function construction unit is used to establish a functional relationship between the transmitted power and the inner phase shift duty cycle data and the outer phase shift duty cycle data based on the power transmission mechanism of the dual active bridge converter under extended phase shift modulation. The equality constraint construction unit is used to take the difference between the transmission power and the preset active power command reference value being equal to zero as the condition for the power balance equality constraint, and to take the absolute value of the actual difference as the default amount and multiply it by the preset equality constraint penalty weight coefficient as the equality constraint penalty term.

[0070] In this preferred embodiment, a functional relationship between the transmitted power and the inner and outer phase shift duty cycle data is established based on the power transmission mechanism of the dual active bridge converter under extended phase shift modulation. An analytical model describing the energy transmission characteristics is constructed, clarifying the nonlinear mapping law between the control variables and the output power. The condition that the difference between the transmitted power and the preset active power command reference value is equal to zero is used as the condition for the power balance equation constraint, strictly limiting the energy conservation principle that the optimization solution must satisfy to prevent power tracking deviation. The absolute value of the actual difference is used as the violation quantity and multiplied with the preset equation constraint penalty weight coefficient to obtain the equation constraint penalty term. The hard power balance requirement is transformed into a soft constraint cost that can be gradient-descented, so that the subsequent particle swarm algorithm can automatically sense and correct the power deviation during the search process, ensuring that the final output phase shift parameter combination can accurately maintain the power supply and demand balance of the DC microgrid.

[0071] Furthermore, the configuration module M2 also includes: an inequality constraint construction unit; The inequality constraint construction unit is used to convert the inductor current polarity requirement corresponding to each switching device at the moment of turn-on into an inequality constraint function form according to the physical conditions for the dual active bridge power devices to achieve zero-voltage turn-on. The negative number of the function value that does not meet the inductor current polarity requirement is taken as the default quantity and multiplied by the preset inequality constraint penalty weight coefficient as the inequality constraint penalty term.

[0072] In this preferred embodiment, based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the inductor current polarity requirements corresponding to each switching device at the moment of turn-on are transformed into inequality constraint functions. This transforms the abstract soft-switching physical mechanism into a computable mathematical criterion, providing a quantitative criterion for the soft-switching operation of all devices. The negative of the function value that does not meet the inductor current polarity requirement is used as the default value and multiplied by the preset inequality constraint penalty weight coefficient to obtain the inequality constraint penalty term. This constructs a dynamic penalty mechanism for the risk of soft-switching failure, so that while pursuing low current stress, the optimization process forces the search direction to avoid parameter regions that will lead to increased hard-switching losses. This expands the zero-voltage turn-on operation range of the system while ensuring efficiency optimization, and avoids a surge in switching losses under light load or high voltage ratio conditions.

[0073] The solution module M3 is used to construct a combination of phase-shifting parameters as a particle position vector and initialize the particle swarm using the inner and outer phase-shifting duty cycle data. It constructs a fitness function using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint, and iteratively solves the problem using a particle swarm optimization algorithm with the objective of minimizing the inductor current peak value to obtain the optimal combination of phase-shifting parameters. In each iteration, the weighting coefficients in the equality constraint penalty and inequality constraint penalty are adjusted based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint.

[0074] Furthermore, the solution module M3 includes: a fitness function construction unit; The fitness function construction unit is used to add the inductor current peak function, the equality constraint penalty term in the power balance equality constraint, and the inequality constraint penalty term in the zero voltage turn-on inequality constraint to obtain the fitness function.

[0075] In this preferred embodiment, a fitness function is obtained by adding the peak inductor current function, the equality constraint penalty term in the power balance equation constraint, and the inequality constraint penalty term in the zero-voltage turn-on inequality constraint. This constructs a unified evaluation system containing multiple optimization objectives and constraints. This comprehensive fitness function integrates three mutually coupled and even conflicting control objectives—minimizing current stress, maintaining power balance, and ensuring soft-switching operation—into a single scalar index. This allows the particle swarm algorithm to directly determine the quality of particles by comparing fitness values, achieving synchronous convergence of multi-objective collaborative optimization. This avoids the logical fragmentation and suboptimal solution problems caused by step-by-step optimization, ensuring that the final phase-shifting parameter combination achieves a globally optimal balance among current stress, power accuracy, and soft-switching performance.

[0076] Furthermore, the solution module M3 also includes: a statistical unit, a proportion calculation unit, and a linear adjustment unit; The statistical unit is used to count the number of first particles that do not meet the power balance equation constraint and the number of second particles that do not meet the zero voltage turn-on inequality constraint in the current iteration. The ratio calculation unit is used to calculate the ratios of the first particle number and the second particle number to the total size of the particle swarm, respectively, to obtain the first ratio and the second ratio. The linear adjustment unit is used to adjust the weight coefficients in the equality constraint penalty and the inequality constraint penalty linearly with positive correlation based on the first ratio and the second ratio, respectively.

[0077] In this preferred embodiment, by statistically analyzing the number of first particles that fail to satisfy the power balance equation constraint and the number of second particles that fail to satisfy the zero-voltage turn-on inequality constraint in the current iteration, the overall default scale of the current population in the power dimension and soft-switching dimension is quantified in real time, providing accurate data support for adaptive adjustment. The ratios of the number of first particles and the number of second particles to the total size of the particle swarm are calculated to obtain the first ratio and the second ratio, respectively. The absolute number of defaults is normalized into a relative probability index reflecting the convergence state of the population, eliminating the influence of the particle swarm size on the adjustment strategy and enhancing the universality of the algorithm. Based on the first ratio and the second ratio, the weight coefficients in the equality penalty and inequality penalty are adjusted linearly with positive correlation, respectively, to establish a linear growth mapping relationship between the penalty intensity and the default ratio. This allows the penalty weight to be automatically increased when the population default ratio is high to accelerate convergence to the feasible region, while the penalty weight is appropriately reduced when the default ratio is low to retain global search capability. This dynamic linear adjustment mechanism significantly improves the convergence robustness of the algorithm under wide operating condition perturbations and effectively overcomes the convergence stagnation or oscillation defects caused by fixed weights.

[0078] The output module M4 is used to generate a pulse width modulation signal based on the optimal phase shift parameter combination and to regulate the switching transistors of the dual active bridge converter to complete the optimization.

[0079] In summary, compared with the prior art, the embodiments of this application have the following beneficial effects: By acquiring the voltage conversion ratio data and the inner and outer phase shift duty cycle data of the dual active bridge converter, the basis of multi-dimensional control variables under extended phase shift modulation is established, providing a complete input dimension for parameter optimization under wide operating conditions; Based on the electrical parameters, inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint are constructed, and the latter two are configured with equality and inequality penalties respectively, transforming the physical-level current stress minimization objective and power transmission accuracy and soft-switching reliability requirements into a calculable mathematical optimization model, ensuring that the optimization process is carried out under the premise of meeting the engineering hard indicators; By constructing the inner and outer phase shift duty cycle data as particle position vectors and initializing the particle swarm, the swarm intelligence search mechanism is used to replace the traditional analytical method or random initialization method, avoiding the risk of the algorithm getting trapped in local optima, and improving the efficiency of the algorithm. Global optimization capability: In each iteration, based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and the zero-voltage turn-on inequality constraint, the weight coefficients in the equality penalty and inequality penalty are dynamically adjusted. This feedback mechanism establishes a closed-loop adjustment logic of "constraint violation degree - penalty intensity - search direction", enabling the algorithm to adaptively enhance the repulsion force on the infeasible region or weaken the excessive constraint on the feasible region according to the real-time state of the population. This solves the stability problem of existing technologies in complex operating conditions such as wide voltage ratio and wide load, where the algorithm is prone to convergence oscillation, getting trapped in local optima, or failing to converge to the feasible region due to fixed penalty weights. Finally, based on the optimal phase shift parameter combination, a pulse width modulation signal is generated to control the switching transistors, realizing low current stress operation and zero-voltage turn-on of all devices in the dual active bridge converter across the entire operating range, significantly improving the transmission efficiency and operational reliability of the system.

[0080] Example 3: This invention also provides a computer program product, including a computer program or instructions, capable of running on a computing device or stored in any available medium. When the computer program product is run on at least one computing device, it causes the at least one computing device to execute any of the dual active bridge extended phase-shifted peak current optimization methods of this invention.

[0081] Example 4: This invention also provides a computer-readable storage medium storing at least one executable instruction that, when executed on a dual active bridge extended phase-shifted peak current optimization system, causes the dual active bridge extended phase-shifted peak current optimization system to perform one of the dual active bridge extended phase-shifted peak current optimization methods described in any of the above method embodiments.

[0082] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. Similarly, for the purpose of simplification and aiding understanding of one or more aspects of the invention, in the above description of exemplary embodiments of this application, various features of the embodiments are sometimes grouped together in a single embodiment, figure, or description thereof. The claims, which follow the detailed description, are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of this application.

[0083] Those skilled in the art will understand that the modules in the system of the embodiments can be adaptively changed and placed in one or more systems different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.

Claims

1. A method for optimizing the extended phase-shifted peak current of a dual active bridge, characterized in that, include: Obtain the electrical parameters of the dual active bridge converter; wherein, the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data, and outer phase shift duty cycle data; Based on the electrical parameters, an inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint under extended phase-shift modulation are constructed; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty. Using the inner and outer phase shift duty cycle data, a combination of phase shift parameters is constructed as the particle position vector to initialize the particle swarm. A fitness function is constructed using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint. With the goal of minimizing the inductor current peak value, the optimal combination of phase shift parameters is obtained through iterative solution using the particle swarm algorithm. In each iteration, the weight coefficients in the equality constraint penalty and inequality constraint penalty are adjusted according to the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint. Based on the optimal combination of phase-shifting parameters, a pulse width modulation signal is generated and the switching transistors of the dual active bridge converter are controlled to complete the optimization.

2. The dual active bridge extended phase-shifting peak current optimization method as described in claim 1, characterized in that, The acquisition of the electrical parameters of the dual active bridge converter includes: Within the feasible solution space of the particle swarm optimization algorithm, initialize the initial values ​​of the inner phase shift duty cycle data and the outer phase shift duty cycle data; Collect primary-side input voltage data, secondary-side output voltage data, and turns ratio data of the high-frequency isolation transformer from the dual active bridge converter; The voltage conversion ratio is calculated based on the primary input voltage data, secondary output voltage data, and turns ratio data.

3. The dual active bridge extended phase-shift peak current optimization method as described in claim 1, characterized in that, The construction of the inductor current peak function under extended phase-shift modulation based on the electrical parameters includes: Based on the relative magnitudes of the inner and outer phase shift duty cycle data, half a switching cycle is divided into various operating modes; For each operating mode, the instantaneous value of the inductor current at each switching moment within half a switching cycle is calculated using the principle of piecewise linear circuits, and the maximum current amplitude is extracted from the instantaneous value of the inductor current at each switching moment to construct the peak inductor current function.

4. The dual active bridge extended phase-shifting peak current optimization method as described in claim 3, characterized in that, The construction of power balance equation constraints under extended phase-shift modulation based on the electrical parameters includes: Based on the power transmission mechanism of the dual active bridge converter under extended phase-shift modulation, a functional relationship is established between the transmitted power and the inner phase-shift duty cycle data and the outer phase-shift duty cycle data. The condition for power balance equation constraint is that the difference between the transmitted power and the preset active power command reference value is equal to zero. The absolute value of the actual difference is taken as the default amount and multiplied by the preset equation constraint penalty weight coefficient as the equation constraint penalty term.

5. The dual active bridge extended phase-shift peak current optimization method as described in claim 4, characterized in that, The construction of zero-voltage turn-on inequality constraints under extended phase-shift modulation based on the electrical parameters includes: Based on the physical conditions for zero-voltage turn-on of dual active bridge power devices, the polarity requirements of the inductor current corresponding to each switching device at the moment of turn-on are transformed into inequality constraint functions. The negative of the function value that does not meet the polarity requirements of the inductor current is taken as the default value and multiplied by the preset inequality constraint penalty weight coefficient as the inequality constraint penalty term.

6. The dual active bridge extended phase-shift peak current optimization method as described in claim 5, characterized in that, The construction of the fitness function based on the inductor current peak function, power balance equation constraint, and zero-voltage turn-on inequality constraint includes: The fitness function is obtained by adding the peak inductor current function, the equality constraint penalty term in the power balance equation constraint, and the inequality constraint penalty term in the zero-voltage turn-on inequality constraint.

7. The dual active bridge extended phase-shifting peak current optimization method as described in claim 1, characterized in that, The step of adjusting the weighting coefficients in the equality constraint penalty and inequality constraint penalty based on the violation ratio of particles in the current particle swarm that violate the power balance equality constraint and the zero voltage turn-on inequality constraint includes: Count the number of first particles that do not satisfy the power balance equation constraint and the number of second particles that do not satisfy the zero voltage turn-on inequality constraint in the current iteration. Calculate the ratios of the first and second particle counts to the total size of the particle swarm to obtain the first ratio and the second ratio. Based on the first ratio and the second ratio respectively, the weight coefficients in the equality constraint penalty and the inequality constraint penalty are adjusted linearly with positive correlation.

8. A dual active bridge extended phase-shifting peak current optimization system, characterized in that, include: Data acquisition module, configuration module, solution module, and output module; The data acquisition module is used to acquire the electrical parameters of the dual active bridge converter; wherein the electrical parameters include: voltage conversion ratio data, inner phase shift duty cycle data, and outer phase shift duty cycle data; The configuration module is used to construct, based on the electrical parameters, the peak function of inductor current under extended phase-shift modulation, the power balance equation constraint, and the zero-voltage turn-on inequality constraint; wherein, the power balance equation constraint is configured with an equality constraint penalty, and the zero-voltage turn-on inequality constraint is configured with an inequality constraint penalty. The solution module is used to construct a combination of phase-shifting parameters as a particle position vector and initialize the particle swarm using the inner and outer phase-shifting duty cycle data. It then constructs a fitness function using the inductor current peak function, power balance equality constraint, and zero-voltage turn-on inequality constraint. With the goal of minimizing the inductor current peak value, it iteratively solves the problem using a particle swarm optimization algorithm to obtain the optimal combination of phase-shifting parameters. In each iteration, the weighting coefficients in the equality constraint penalty and inequality constraint penalty are adjusted based on the proportion of particles in the current particle swarm that violate the power balance equality constraint and zero-voltage turn-on inequality constraint. The output module is used to generate a pulse width modulation signal based on the optimal phase shift parameter combination and to regulate the switching transistors of the dual active bridge converter to complete the optimization.

9. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement the dual active bridge extended phase-shifting peak current optimization method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a dual active bridge extended phase-shift peak current optimization method as described in any one of claims 1-7.

Images