Optimal Strategy of Triple Phase Shift Transient DC Bias for Dual Active Bridge DC-DC Converter

. Under the traditional phase shift control, dual active bridge DC-DC converter (DAB) has a transient bias on the inductance current when the phase shift ratio changes suddenly, which leads to increase in the current stress of switch tubes and even unidirectional saturation of the transformer. All those will affect the stable operation of the converter. This paper focuses on quantifying the transient DC bias model on the control of triple phase shift when phase shift ratios change, and then the strategy proposed by the paper can eliminate the DC bias within one switching period and reduce current stress and response time by adjusting the drive signal during the transient transition process. Finally, the proposed strategy is verified by simulation results from Matlab/Simulink software.


Introduction
In recent years, with the development of renewable energy, it is necessary to use bidirectional DC-DC converters to achieve bidirectional energy transmission when the renewable energy power generation system and energy storage system are connected to the power grid [1]- [3]. The dual active bridge DC-DC converter (DAB) is a typical topology for bidirectional energy transmission. It has the advantage of electrical isolation, easy to achieve ZVS soft switching and high power density [4]- [6]. Naturally, DAB converters are commonly used in the fields of microgrid, energy storage system and new energy vehicles. Figure 1 shows the topology of DAB converter. This converter is mainly composed of four parts: primary H-bridge, secondary H-bridge, high-frequency isolation transformer TF and inductance Lr. In DAB converter, the inductance Lr acts as an energy storage and release element, and the power of converter is determined by the internal and external phase shift ratios of the primary and secondary sides. In the process of converter power regulation, the sudden change of phase shift ratios will cause an instantaneous current shock in the circuit, and DC bias will also appear at this moment. Current shock will cause the circuit current stress to increase, and the long-term DC bias may even cause deep saturation of the transformer in one direction, affecting the stable operation of the converter [7]- [8]. Triple phase shift (TPS) is the phase shift control with the highest degree of freedom. Single phase shift (SPS), extended phase shift (EPS) and dual phase shift (DPS) can be regarded as special cases of TPS. Therefore, this paper focuses on analyzing the converter under TPS control, through quantitative calculation to adjust the drive signal of the next period after phase shift ratios change, so that the converter can transition to the next stable state within one switching period after phase shift ratios change. This strategy can reduce current stress, eliminate DC bias and speed up dynamic response when phase shift ratios change.

TPS control of DAB converter
When the converter energy is transferred in the forward direction, the energy is transferred from the input side Vin to the output side Vout ; when the energy is transferred in the reverse direction, the energy is transferred from the output side Vout to the input side Vin . In order to simplify the analysis process, this paper only analyzes the transient optimization strategy when the energy is E3S Web of Conferences 185, 01088 (2020) ICEEB 2020 http://doi.org/10.1051/e3sconf/202018501088 transferred in the forward direction, and there is a similar analysis method when the energy is transferred in the reverse direction. As shown in Figure 1, VAB is the square wave voltage output by the primary H-bridge, and VCD is the square wave voltage output by the secondary H-bridge and converted to the primary. D1 and D3 are defined as the internal phase shift ratios of the primary side and secondary side respectively, and D2 is the external phase shift ratio between primary side and secondary side.
When the converter energy is transferred in the forward direction, according to the relationship between D2+D3 and 1, the steady working state of TPS can be divided into two types: when D2+D3<1, the converter is considered to be in Mode 1, and its steady state operating waveform is shown in Figure 2; when D2+D3≥1, the converter is considered to be in Mode 2, and its steady state operating waveform is shown in Figure 3.
When the converter is in the steady state Mode 1, it satisfies I(tx)=-I(tx+Ths), where Ths is half a switching period. According to this condition, the current at the moment of the switching action within half a switching period can be obtained, where the initial value of the current:  When the converter is in the steady state mode 2, it also satisfies I(tx)=-I(tx+Ths). According to this condition, the current at the moment of the switching action within half a switching period can also be obtained, where the initial value of the current:

Transient DC bias phenomenon
The sudden change of phase shift ratios will cause current shock and DC bias of the inductance current. In order to quantitatively calculate the DC bias of the inductance current, let the phase shift ratios before the change be D1, D2, D3, and the phase shift ratios after the change are D1", D2", D3" respectively. If the maximum value of the inductance current after change is Imax, the minimum value is Imin, the transient DC bias is defined as Ibias=0.5 (Imax+Imin). The sudden change of phase shift ratios can be divided into the following four types because the steady state under TPS control is divided into two types. Transient DC bias Ibias under various conditions:  I  I  V  nV  I  D  D  T  D  D  D  D  T  L  L  I  I  V  nV  I  D  D  T  D  D  D  D  T  L  L  I  I  V  nV  I  D  D  T  D  D  T  L  L  I  I After phase shift ratios change, the inductance current ILr exhibits asymmetry between positive and negative, Ibias≠0, which is called transient DC bias. Long-term DC bias will cause transformer unidirectional magnetic saturation. At the moment of change, current shock will also occur, which will endanger the safe operation of the converter.

Optimized control strategy for transient DC bias
After optimization, let the phase shift ratios before change be D1, D2, D3, and the converter is considered to be in steady state 1; let the phase shift ratios after change be D1", D2", D3", and the converter is considered to be in steady state 2. The moment between steady state 1 and steady state 2 is called transient transition process. In the following, we can quantitatively adjust the turn-on and turn-off moments of drive signal during the transient transition process by setting an unknown quantity to solve, in order to eliminate transient DC bias within one switching period.

Mode 1 before change, Mode 2 after change
As shown in Figure 9, assuming that the drive signal of the 4, 5 and 8 tubes are turned on at the time of D1´T hs , D2´T hs , (D2+D3)Ths , ,and the turn-off time of the 4 and It0" itself can be solved by the steady state process afterwards: Combining (7)、(8),we can get: If D2=0.5(D2+D2"), then D3=0.5(D3+D3").

Mode 2 before change and Mode 1 after change
As shown in Figure 10, assuming that the drive signal of the 4 and 5 tubes are turned on at the time of D1´T hs , D2´T hs , and the turn-off time is the same as steady state 2, and the turn-off time of tube 8 is xThs during transient transition process, the converter can eliminate transient DC bias. Through It0 and the set value of the transient transition process, we can obtain: (13)

Mode 2 before and after change
As shown in Figure 11, assuming that the drive signal of the 4 and 5 tubes are turned on at the time of D1´T hs , D2´T hs , and the turn-off time is the same as steady state 2, and the turn-on time of tube 8 is the same as steady state 2, and the turn-off time is xThs during transient transition process, the converter can eliminate transient DC bias. Through It0 and the set value of the transient transition process, we can obtain: It0" itself can be solved by the steady state process afterwards: In the four cases discussed, we can eliminate the transient DC bias by adjusting the turn-on and turn-off time of the switch tube respectively during the transient transition process.

Simulation verification
In order to verify the transient optimization control strategy proposed for TPS, this paper uses Matlab/simulation software to verify the idea. The simulation parameters are shown in Table 1. Switch frequency 100KHz Transformer ratio 8:1

Mode 1 before change, Mode 2 after change
The simulation conditions of Figure

Mode 2 before and after change
The simulation conditions of Figure 15: the transition is from D1=0.7, D2=0.4, D3=0.8 of Mode 2 to D1=0.2, D2=0.8, D3=0.5 of Mode 2. From the simulation results, a larger current shock will be generated at the moment of phase shift ratios change and the response time is longer when the converter does not have a transient optimization strategy. When the transient optimization strategy exists, the stable state is restored within a switching period after phase shift ratios change. This optimization strategy reduces the transient DC bias and current shock caused by the sudden change of phase shift ratios, and speeds up the dynamic response.

Conclusion
In this paper, a transient optimization strategy under the control of TPS with dual active bridge DC-DC converters is proposed by classifying the TPS control strategy. The main idea of this strategy is to adjust the moment when the drive signal is turned on and off during the transient transition process, so that the current can be adjusted from the initial value of steady state 1 to the initial value of steady state 2. We can solve the specific moment by setting the unknown quantity. This strategy E3S Web of Conferences 185, 01088 (2020) ICEEB 2020 http://doi.org/10.1051/e3sconf/202018501088 6 can reduce current shock and transient DC bias of the converter and speed up the dynamic response after phase shift ratios change.