Four-degree-of-freedom Efficiency Optimization Control Strategy For Isolated Dual-active-full-bridge DC-DC Converter

In the context of global energy shortage and increasingly serious climate problems,renewable energy power generation technology continues to receive extensive attention from countries around the world.In the field of distributed renewable energy generation and its application,bi-directional DC-DC converters are commonly used for voltage adaptation and bidirectional energy flow control.Due to the advantages of isolated dual active bridge(DAB)DCDC converter,such as simple structure,high power density,wide voltage regulation range,and easy-to-realize soft switching,it has become a research hotspot of bidirectional DC-DC converter topology.In this dissertation,the comprehensive loss model,efficiency optimization modulation strategy,and closed-loop control strategy of the DAB converter are deeply studied.Increasing the modulating degree can improve the control flexibility and working efficiency of DAB.In order to solve the problem of indirect indicators and ignoring high frequency transformer loss in efficiency optimization with triple-phase-shift modulation,this dissertation proposes a direct efficiency optimization strategy based on the comprehensive loss model of DAB.The working characteristics of DAB converter under triple-phase-shift modulation are analyzed by discrete time-domain analysis method,considering the parasitic capacitance of switch devices and the commutation process within the dead time.Then a forward and backward computation process of instantaneous value of inductor current is proposed.Considering the non-linear characteristics of high-frequency transformer excitation inductance,a comprehensive loss model of DAB including switching device loss and high-frequency transformer loss is established,and a four-degree-of-freedom(4-DF)modulation efficiency optimization strategy with triple-phaseshift modulation plus variable frequency strategy is proposed.The accuracy of the established loss model is verified by PLECS thermal simulation and COMSOL finite element simulation.The experimental results demonstrate that the forward and backward computation process of instantaneous value of inductor current obtained by the proposed method can better approximate the actual value than that obtained by the traditional lossless model.Moreover,the 4-DF optimization modulation can effectively improve the working efficiency.The comprehensive loss model of DAB has the characteristics of discretization and nonlinearity.When the operating conditions vary widely,the varying range of the four-degreeof-fredom parameters is larger,and the parameter optimization time is longer.Aiming at the problem of 4-DF parameter operation constraints and efficiency optimization,this dissertation proposes a method to optimize and solve the 4-DF parameters of DAB under all operating conditions using the cuckoo search algorithm,which can quickly obtain the discrete space mapping relationship between the converter operating variables and the 4-DF modulation parameters under efficiency optimization.On this basis,a 4-DF parameter mapping implementation method based on the neural network is proposed.According to the real-time requirements of converter power and gain,the optimal 4-df modulation parameters are directly given,which solves the contradiction between storage capacity and control accuracy of the discrete look-up table method when working conditions change continuously.A control strategy is proposed for closed-loop control of DAB converter,which combines traditional PI regulator with 4-DF efficiency optimization modulation based on neural network mapping,achieving constant-voltage or constant-power control with efficiency optimization.In addition,anti-disturbance rejection control(ADRC)is introduced into the control loop of DAB under 4-DF modulation to improve DAB control performance,and an improved ADRC strategy is proposed that realizes adaptive adjustment of the ADRC parameters through virtual current.Furthermore,an improved ADRC parameters optimization design method based on Parteo multiobjective optimization is proposed.Simulation results show that the improved controller effectively solves the contradiction between system dynamic performance and steady-state performance in PI control,not only enhances the tracking performance of output voltage to command,but also effectively reduces the peak-to-peak fluctuation of inductor current in highfrequency chain at steady-state.An experimental prototype of isolated dual-active-full-bridge DC-DC converter under 4-DF efficiency optimization control is established,and efficiency optimization and control performance experiments are carried out under different operating conditions.The experimental results show that DAB converter under improved ADRC,neural network mapping and 4-DF modulation has relatively high operating efficiency and excellent dynamic and static performance.