I. INTRODUCTION

The use of the fossil fuels causes environmental pollution and ecological problems. In addition, carbon-dioxide emissions result in global warming. Therefore, the development and application of renewable energy sources has become very important [1]-[10]. Renewable energies include wind power, solar power, ocean energy, hydrogen energy, hydro energy and so on. Fig. 1(a) shows the conventional renewable energy and battery hybrid power supply system. It can be seen that the battery is charged from a renewable energy source via two converters [11]-[17]. Therefore, the conventional hybrid power supply system results in energy waste. For this reason, this paper presents a new DC-DC converter for three-port renewable energy and battery hybrid power supplies systems, as shown in Fig. 1(b). It can be seen that the battery is only via one converter at the charging or discharging mode. Fig. 2 shows the circuit configuration of the proposed converter. The pulse-width modulation (PWM) technique is used for the proposed converter. At the charging mode, the renewable energy source provides its energy to charge the battery via the proposed converter. Here the proposed converter is a conventional DC-DC buck converter [18], [19]. At the discharging mode, the battery releases its energy to the DC bus via the proposed converter. Meanwhile, this converter is a non-isolated high step-up DC-DC converter [20]. The coupled-inductor technique is used to achieve a high step-up voltage gain by adjusting the turns ratio. Moreover, the leakage-inductor energies of the primary and secondary windings can be recycled. Thus, only one power converter is utilized at the charging and discharging modes. The operating principles and steady-state analyses will be described in the following sections for the charging and discharging modes.

Fig. 1.Renewable energy and battery hybrid power supplies system. (a) Conventional type, (b) Three-port type.

Fig. 2.Circuit configuration of proposed DC-DC converter.

In order to analyze the steady-state characteristics of the proposed converter, some conditions are assumed: (1) Because the capacitors C1, CB, and Cbus are sufficiently large, the voltages across these capacitors can be treated as constant during each switching period. (2) The ON-state resistance of the switches, the forward voltage drop of the diodes, and the equivalent series resistances of the coupled inductor and capacitors are ignored.

II. CHARGING MODE

When the proposed converter is operated in the charging mode, the primary winding of the coupled inductor is used for a general inductor. The equivalent circuit of the proposed converter at the charging mode is shown in Fig. 3, where RB is the equivalent load for the battery. The PWM technique is used to control switch S1. Switch S2 is used for the synchronous rectifier. Fig. 4 shows typical waveforms of the proposed converter with the continuous conduction mode (CCM) operation at the charging mode. The operating principles and steady-state analysis are described as follows:

1) Mode 1: During this time interval [t0, t1], switch S1 is turned on and switch S2 is turned off. The current flow path is shown in Fig. 5(a). The renewable-energy source Vin supplies its energy to for the coupled inductor Lm, capacitor CB, and load RB. Meantime, the coupled inductor Lm stores its energy. Thus, the current iLm is increased. The voltage across the coupled inductor Lm is given by:

2) Mode 2: During this time interval [t1, t2], switch S1 is turned off and switch S2 is turned on. The current flow path is shown in Fig. 5(b). Meanwhile, switch S2 is utilized for a synchronous rectifier. The energy stored in the coupled inductor Lm is released to the capacitor CB and load RB. The voltage across the coupled inductor Lm is derived as:

By using the voltage-second balance principle on the coupled inductor Lm, the following equation is obtained:

Substituting (1) and (2) into (3), the voltage gain in the charging mode is found to be:

When the proposed converter is operated in the boundary conduction mode (BCM), typical waveforms are shown in Fig. 6. Thus, the peak value of the coupled-inductor current is written as:

By utilizing the ampere-second balance principle on the capacitor CB, the following equation can be expressed as:

From (6), the peak value of the coupled-inductor current can be rewritten as:

Then, the normalized inductor time constant is defined as:

Substituting (4), (5), and (8) into (7), the boundary normalized inductor time constant is given by:

If τLm1 is larger than τLm1,B, the proposed converter is operated in the CCM at the charging mode.

Fig. 3.Equivalent circuit of proposed converter at charging mode.

Fig. 4.Some typical waveforms of proposed converter with CCM operation at charging mode.

Fig. 5.Current flow path of proposed converter at charging mode. (a) Mode 1. (b) Mode 2.

Fig. 6.Some waveforms of proposed converter with BCM operation at charging mode.

III. DISCHARGING MODE

The equivalent circuit of the proposed converter at the discharging mode is shown in Fig. 7, where RL is the equivalent load for the DC bus. The PWM technique is used to control switch S2. Switch S1 is turned off at the discharging mode. Fig. 8 shows typical waveforms of the proposed converter with the CCM operation at the discharging mode. The operating principles and steady-state analysis are described as follows:

1) Mode 1: During this time interval [t0, t1], switch S2 is turned on. The current flow path is shown in Fig. 9(a). The energy of the battery is released to the magnetizing inductor Lm and primary leakage inductor Lk1. Thus, the currents iLm and iLk1 are increased. The secondary leakage inductor Lk2, secondary winding N2, capacitor C1, and battery are in series to release their energies for the load RL. The energy of the capacitor Cbus is also provided to the load RL. Therefore, the current iLk2 decreases. At iLk2 = 0, the energy stored in the leakage inductor Lk2 is completely recycled to the load RL.

2) Mode 2: During this time interval [t1, t2], switch S2 is still turned on. The current flow path is shown in Fig. 9(b). The energy of the battery is still released to the magnetizing inductor Lm and primary leakage inductor Lk1. The energy of the capacitor Cbus is provided to the load RL.

3) Mode 3: During this time interval [t2, t3], switch S2 is turned off. The current flow path is shown in Fig. 9(c). The energies stored in the magnetizing inductor Lm and primary leakage inductor Lk1 are released the capacitor C1. Thus, the currents iLm and iLk1 are decreased. The secondary winding N2, capacitor C1, and battery are in series to release their energies for the leakage inductor Lk2, capacitor Cbus, and load RL. Therefore, the current iLk2 is increased. At iLk1 = 0, the energy stored in the leakage inductor Lk1 is completely recycled to the capacitor C1.

4) Mode 4: During this time interval [t3, t4], switch S2 is still turned off. The current flow path is shown in Fig. 9(d). The secondary leakage inductor Lk2, secondary winding N2, capacitor C1, and battery are in series to release their energies for the capacitor Cbus and load RL. Therefore, the currents iLm and iLk2 are decreased.

When switch S2 is turned on, the following equation can be represented as:

Thus:

where the turns ratio of the coupled inductor n = N2/N1, and the coupled coefficient k = Lm/(Lm+Lk1).

From the operating principle, it is known that the energy stored in the primary leakage inductor Lk1 is recycled to the capacitor C1. By using the ampere-second balance principle on the capacitor C1, the released time duration of the primary leakage-inductor energy can be expressed as [20]:

By utilizing the voltage-second balance principle on the primary leakage inductor Lk1 and magnetizing inductor Lm, the following equations can be obtained:

where vLk1(tr) is the voltage across the primary leakage inductor Lk1 during the time duration [t2, t3] and the voltage vN1(OFF) across the magnetizing inductor Lm during the time duration [t2, t4].

Substituting (11) and (14) into (15), yields:

By substituting (12) into (16), it is possible to derive:

From Fig. 9(c), the following equation can be obtained:

Substituting (17) and (18) into (19), the voltage across the capacitor C1 is written as follows:

During the switch S2 OFF-period, the following voltage equation is found as:

Therefore:

Using the voltage-second balance principle on the secondary winding of the coupled inductor N2, the equation can be represented as:

Substituting (13), (20), and (22) into (23), the voltage gain can be found as follows:

At k = 1, equation (24) is rewritten as:

The voltage gain with parasitic components is analyzed as follows. In order to simplify the analysis, the leakage inductors of the coupled inductor are neglected. The equivalent circuit is shown in Fig. 10. rN1 and rN2 represent the equivalent series resistances (ESR) of the primary and secondary windings of the coupled inductor. VFD2 and rD2 are the ON-state forward voltage drop and resistance of D2. rS2 denotes the ON-state resistance of S2.

Fig. 7.Equivalent circuit of proposed converter at discharging mode.

Fig. 8.Some typical waveforms of proposed converter with CCM operation at discharging mode.

Fig. 9.Current flow path of proposed converter at discharging mode. (a) Mode1. (b) Mode 2. (c) Mode 3. (d) Mode 4.

Fig. 10.Equivalent circuit including ESR of coupled inductor, ON-state forward voltage drop and resistance of diodes, and ON-state resistance of switch. (a) S2 ON. (b) S2 OFF.

When switch S2 is turned on, the equivalent circuit is shown in Fig. 10(a). The average values of ibus and vN1 are written as:

When switch S2 is turned off, the equivalent circuit is shown in Fig. 10(b). The average values of ibus and vN1 are found as:

Since the leakage inductors of the coupled inductor are neglected, the coupling coefficient k is equal to 1. From (20), the voltage Vc1 can be rewritten as:

Substituting (30) into (29), yields:

By using the ampere-second balance principle on Cbus, the following equations are obtained as:

Substituting (26) an (28) into (32), Ib(OFF) is derived as:

In addition, Ib(ON) can be given as:

Using the volt-second balance principle on Lm, yields:

Substituting (27), (31), (33), and (34) into (35), the actual voltage gain can be obtained as follows:

The curves of the ideal and actual voltage gain under n=3, rN1=rN2=100 mΩ, rD2=50 mΩ, rS2=50 mΩ, VFD2=1.25 V, Vbat=24 V, and RL=200 Ω are plotted in Fig. 11. It can be seen that the proposed converter in the discharging mode can achieve a high step-up voltage gain.

Fig. 11.Ideal and actual voltage gain in discharging mode.

When the proposed converter is operating in the BCM, typical waveforms are shown in Fig. 12. Thus, the peak value of the magnetizing-inductor current is given as:

Applying the ampere-second balance principle on the capacitor Cbus, the following equation is found:

From the above equation, the peak value of the magnetizing-inductor current is rewritten as:

Then, the normalized inductor time constant is defined as:

Substituting (24), (37), (40) into (39), the boundary inductor time constant is obtained as follows:

At k = 1, τLm2 can be rewritten as:

If τLm2 is larger than τLm2,B, the proposed converter in the discharging mode is operated in the CCM.

Fig. 12.Some waveforms of proposed converter with BCM operation at discharging mode.

IV. EXPERIMENTAL RESULTS

In order to verify the feasibility of the proposed converter, a prototype circuit is built for a fuel-cell and battery hybrid power supply system. The electric specifications and circuit components are selected as the input voltage Vin = 28~36.5 V, output power Po = 200~20 W, battery voltage Vbat = 24 V, DC-bus voltage Vbus = 200 V, switching frequency fs = 50 kHz, coupled inductor Lm = 72 μH and n = 3, capacitors CB = C1 = Cbus = 220 μF, switches S1 (IXFH80N085) and S2 (IXTQ96N20P), and diodes D1 (DSSK60-02AR) and D2 (DSEP30-06A).

Fig. 13 shows the experimental results at the charging mode. The electrical specifications are Vin = 28 V, Vbat = 24 V, and full load Po = 200 W. From Figs. 13(a) and 13(c), it can be seen that the current waveforms, iS1 and iLm, are same during the S1 ON-period. As can be seen from Figs. 13(b) and 13(c), the current waveforms, iS2 and iLm, are the same during the S1 OFF-period. Fig. 13(c) shows that the proposed converter is operated in the CCM. The measured efficiency at the charging mode is shown in Fig. 15. The measured efficiency is around 95.2%~97.5% under Vin = 28~36.5 V and Po = 200~20 W. Fig. 14 shows the experimental results at the discharging mode. The electrical specifications are Vbat = 24 V, Vbus = 200 V, and full load Po = 200 W. The waveforms, vgs2, vS2, and iS2, are shown in Fig. 14(a). It can be seen from the waveform iS2 that the proposed converter is operated in the CCM. From the waveform vS2, the voltage across S2 is clamped at approximately 70 V during the S2 OFF-period. Therefore, a low rated voltage MOSFET can be adopted to reduce the conduction loss. As shown in Fig. 14(b), the voltage across D1 is clamped at approximately 70 V during the S2 OFF-period. The waveforms, iD1 and iLk1, are shown in Figs. 14(b) and 14(c). It can be seen that they consist with the operating principle. The measured efficiency at the discharging mode is shown in Fig. 15. The measured efficiency is around 93.3%~95.4%.

Fig. 13.Experimental waveforms of proposed converter at charging mode. (a) vgs1, vS1, iS1, (b) vgs2, vS2, iS2, (c) Vbat, iLm.

Fig. 14.Experimental waveforms of proposed converter at discharging mode. (a) vgs2, vS2, iS2, (b) vD1, iD1, (c) Vbus, iLk1.

Fig. 15.Measured efficiency of proposed converter.

V. CONCLUSIONS

In the conventional hybrid power supplied system, two power converters are used for the battery charging or discharging modes. The conventional system results in energy waste. A new DC-DC converter for renewable energy and battery hybrid power supplied systems is investigated in this paper. At the charging mode, a renewable energy source can provide its energy to charge the battery via the proposed converter. At the discharging mode, the battery can release its energy to the DC bus via the proposed converter. Thus, only one power converter is utilized at the charging or discharging modes. The proposed converter can increase the conversion efficiency. The measured efficiency is around 95.2%~97.5% at the charging mode and it is around 93.3%~95.4% at the discharging mode.