Design and Implementation of Z-Source Inverter by Simple Boost Control Technique for Laboratory Scale Micro-Hydropower Application

In a power plant such as micro-hydropower (MHP), an induction generator (IG) is usually employed to produce electrical power. Therefore, an inverter is needed to deliver it with high efficiency. Z-source inverter (ZSI) has been introduced as a topology with many advantages over conventional inverters. This research aims to investigate the performance of ZSI based simple boost control (SBC) in laboratory-scale MHP systems using a rewinding induction generator. This research has been conducted both from simulations and experiments. Based on the result, the waveform characteristic and value of ZSI are close to the desired design. A shoot-through duty ratio of 17% can reach 60 Vrms output voltage, and this condition has a conversion ratio of about 2.05 times. Also, SBC can significantly reduce the Total Harmonic Distortion (THD). ZSI efficiency has a value of 84.78% at 50% of rating load 100 W and an average value of 80%. Compared to the previous study, the proposed design has more economical with the same component for the higher rating power. Moreover, it has a smoother and entire output waveform of the voltage. Keywords— micro-hydropower, induction generator, simple boost control, z-source inverter


I. INTRODUCTION
Electricity generation based on fossil energy alone is appraised as inadequate for the growth of demand energy capacity. Besides, the utilization of fossil energy adversely affects the increase of environmental pollution and the greenhouse effect. This problem encourages many researchers to utilize renewable energy resources more and promote nonemission electricity for better human social development. [1], [2]. Some renewable energy such as solar energy, geothermal, wind energy, biomass, and hydro become options to generate electricity from sustainable energy in the future [3]. Among all renewable energy resources, micro-hydropower (MHP) is considered the most favorable energy source, which has an efficient, clean, and uncomplicated technology [4]. Hence, this power generation technology deserves to implement in remote areas with intricate geographical contours, such as mountainous areas.
The utilization of renewable energy can be installed either standalone or grid system [5]. The MHP system implemented in an isolated area is possible for a standalone system. This concept used in specific locales is mostly not connected to the grid. Standalone systems of MHP constitute renewable solutions that can give a low-cost power supply with more good social and environmental sustainability than a grid-connected system [6]. The MHP system generally utilizes an induction generator (IG) to convert mechanical energy into electricity. It is selected because of some merits such as costless, reliable, simple operation, brushless, without DC excitation, and a long lifetime without any severe problem [7].
After the electrical power has been converted, the IG requires an electrical circuit to deliver the load power optimally. Therefore, the use of electrical circuits with high efficiency is essential. One of the most important electrical devices is an inverter, which the popular technology is Voltage Source Inverter (VSI) and Current Source Inverter (CSI) [8]. The other inverter topology is introduced that is familiar enough for renewable energy application. This topology is known as Zsource Inverter (ZSI), which was investigated in [9] - [11]. ZSI has a unique circuit because it consists of two identical inductors and capacitors crossing each other, utilizes the shoot-through state to increase performance, and operates as a voltage/current source type of inverter [12].
Moreover, ZSI can also operate either buck mode or boost mode, which is more beneficial to overcome the drawbacks of two-stage configurations [13]. The other advantage of ZSI is no need for time delay regulation [14], able to avert the output waveform from the distortion [15], vast voltage operation [16], and adding a shoot-through to avoid inverter short circuit [17]. Consequently, it can boost the input voltage with good output voltage quality [18].
Many researchers have been testing ZSI on renewable energy applications. In [19], ZSI is applied for the MHP emulator using standard PWM switching. The ZSI maximum 63 Jurnal Teknik Elektro Vol. 13 No. 2 Juli -Desember 2021 efficiency is obtained around 80% and is almost constant for wide-range power. In [20], ZSI has been implemented in photovoltaic applications with the result that this topology is more appropriate for both standalone and grid applications. The comparison between ZSI and other topologies for small wind turbines has been investigated [21]. It has good reliability for the conversion systems and reduces the voltage stresses. Another comparison between ZSI and conventional inverter topology has been presented in [22]. This comparison is applied in a fuel cell with the result that ZSI gives a more suitable DClink voltage. For ZSI applications, most studies focus on conventional switching with simulation schemes. Therefore, experimental validation is required to confirm the performance result of the simulation. Different strategy control is needed to improve the performance of ZSI. One of the best strategies for ZSI topology is simple boost control (SBC). This strategy has some advantages, such as being easy to control [23] and having good characteristics [24]. The merit of SBC is also emphasized in [25], wherein this strategy can reduce the harmonic output properly and increase the power density. Hence, it is suitable for grid power quality improvement.
This paper aims to design and implement a ZSI topology controlled by a simple boost strategy. The simulation and experimental investigation were carried out to confirm its performance. Furthermore, the experimental design of ZSI is applied to a micro-hydropower for laboratory-scale based on induction generators to validate the simulation result. The application of ZSI in MHP with a standalone system is predicted to increase the inverter roughness and maintain the characteristics of the turbine in a suitable condition.
II. METHOD Figure 1 shows the proposed configuration design system. The systems consist of an induction generator rotated by a prime mover to produce AC voltage. Then a 3-phase rectifier device converts it into DC voltage. A bank capacitor is utilized to supply field current on the induction generator. The output voltage has functioned as input of ZSI that consists of two identical inductors and capacitors connected to the 3-phase inverter. An LC filter installs on the output side to produce pure sinusoidal voltage. The 3-phase inverter switching uses simple boost control to maintain the constant output voltage with shoot-through duty ratio adjustment.
ZSI topology depicted in Figure 2 only uses two identical inductors and capacitors crossing each other to produce the desired voltage [12]. In ZSI impedance, the inductor components work to reduce current ripples, while the capacitor component is used to compensate for these ripples to keep the constant output voltage [10]. ZSI can overcome the drawback of conventional inverters such as VSI and CSI. The conventional inverter cannot operate on both modes, as step-up and step-down. Meanwhile, ZSI can increase input voltage (boost mode) and otherwise (buck mode) without auxiliary switching. The total switching on ZSI is written in (1).
= 0 + 1 (1) where T0 is a shoot-through period and T1 is a non-shoot through period.
There are three operating modes on ZSI: active conditions, null conditions, and shoot-through zero states. An analysis is carried out by assuming the ideal component. The theoretical waveforms are shown in Figure 3. In shoot-through zero states or during the T 0 period, the inductor current will increase, the capacitor in discharge mode, and the voltage across the inductor will appear. Meanwhile, in non-shoot through zero states or during the T1 period, the inductor current will decrease, the capacitor in charge mode, and the inductor voltage will be zero. In this circumstance, the DC source is charged to the ZSI circuit that consists of two identical inductors and capacitors. The capacitor is charged, and the energy will flow to the load through the inductor component. Based on the condition, the inductor suffers discharging. The equivalent circuit of ZSI on mode I is shown in Figure 4(a).
2) Mode II (Null Condition) ZSI operates in one of two null conditions. DC voltage flows to the inductor and capacitor. The switch suffers a short circuit either on the bottom or top. As long as this mode, the inverter bridge becomes an open circuit condition because the current flows to the load equivalent reach zero value (Ix = 0). Figure 4(b) shows the equivalent circuit on this mode.

3) Mode III (Shoot-Through Zero States Condition)
ZSI operates on one of seven shoot-through zero states conditions. The inverter bridge circuit suffers short circuit again, resulting in zero output voltage on the inverter bridge. The capacitor voltage will gradually boost related to the shootthrough duty ratio. Figure 4(c) shows the equivalent circuit on mode III.  On the ZSI topology, the DC-link voltage is written in (2), while the boosting factor (B) is given in (3).
V rec is the rectifier output voltage as ZSI input. Then, equation (4) gives the peak output phase voltage as follows: M is the modulation index with a value of M ≤ 1. At the time of shoot-through condition with a period of T0, it can be found that the shoot-through duty ratio has a value, which is expressed in (5).
If T is the switching period, then the switching frequency (fs) is 1/T. Hence, D0 can satisfy the formula in (6).
During the non-shoot through time, the inductor current decreases linearly, and its voltage is different from the input and capacitor voltage. The average inductor current is formulated in (7).
The maximum current ripple that passes through the inductor occurs when the shoot-through is maximum, so the peak to peak current ripple (ΔIL) on the inductor needs to be determined by maximum and minimum inductor current, which is calculated by (8) and (9), respectively. As a result, ΔIL can be found in (10).
so that the inductor and capacitor value is formulated in (12) and (13) respectively [13].
The filter that used in ZSI circuit can be determined in (14) for inductor and (15) for capacitor.
where RL is load resistance.
SBC can regulate the output voltage by adding shootthrough zero states for each switches component [26]. Two DC signals employ the shoot-through duty ratio (D0). The first signal is equal to the peak value of the three-phase reference signal, while the second signal is the negative value of the first signal. When the triangular carrier signal is higher than the positive DC signal (VPO) or less than the lower limit DC signal (VNE), the circuit will be in a shoot-through state [27]. Finally, the process above will generate the gate driving signal, which has characteristics in As shown in Figure 6, to obtain the output of the SBC system, the load voltage is read by the voltage sensor and converted to the RMS value. Next, it is compared with the reference voltage to result in the error value as an input of the PI controller. The PI controller generates the shoot-through duty ratio (D0). To produce a shoot-through signal, the value of D0 is multiplied by 1 and -1 to produce the required DC voltage on the SBC. The switching characteristics of the MOSFET inverter are highly dependent on the results of the comparison between the DC voltage and the carrier signal to produce shoot-through conditions. When the carrier signal is higher than the positive DC signal or less than the lower limit DC signal, the circuit will be in a shoot-through state. Outside of both conditions, it becomes a non-shoot-through state. Then, a shoot-through condition is added to the SPWM circuit for switching in all MOSFETs whose mechanism is regulated using OR logic gates. Three conditions worked in the switching process: active, null, and shoot-through zero state conditions.
The experiment of ZSI using simple boost control for laboratory-scale MHP was carried out by supplying IG rotation using a prime mover. A variable frequency drive (VFD) controls the prime mover to obtain the desired rotation. It is necessary to rotate IG above a synchronous speed of about 1490 rpm based on its rating. Moreover, the capacitor bank of 257 μF is used to generate the field current of IG. Figure 7 shows the complete setup of IG as the input of ZSI. This configuration can ensure the production of an input voltage for the ZSI circuit. The IG used is a rewinding machine, so it has a rated phase voltage of 56 V, as shown in Table I. The ZSI circuit is designed with the initial parameters shown in Table II. Then, Table III summarizes ZSI impedance, simple boost control, and driver components. The simulation design of ZSI using the SBC technique is conducted in PSIM software, as shown in Figures 8(a) and 8(b), respectively. Then, Figure 9 shows the experimental setup of the ZSI device.

III. RESULTS AND DISCUSSION
A. Testing of Rewinding Induction Generator Output Induction generator testing aims to determine the output voltage produced by an induction motor when operated as a generator. The rewinding induction generator was applied in this campaign. Hence, this machine has an RMS voltage of 56 V in normal operation. As a generator, the induction motor must be rotated above its synchronous speed and installed a capacitor bank as a reactive power supplier. The rating speed of the induction machine is 1390 rpm, so in this test, 1492 rpm speed is applied. From this test, the AC output voltage is about 45 V RMS which means below the machine voltage rating as shown in Figure 10(a). Furthermore, this voltage is rectified by a 3-phase rectifier and smoothed by a filter capacitor to obtain the DC voltage around 52 V as the voltage input of ZSI. Figure  10

B. Testing of Simple Boost Control
The test is conducted by analyzing a simple boost control waveform with a switching frequency input of 7.842 kHz. Based on the test results, the switching period is 128 μs, so the switching frequency is around 7.8 kHz. This frequency is also close to the desired value from the simulation, which is 7.842 kHz or comparable to 127.5 μs. Figure 11 shows the characteristics of a simple boost control waveform. Figure 11. Simple boost control waveform C. Testing of ZSI Waveform Figure 12 shows the comparison between the capacitor and inductor voltage waveforms. When the switch is in non-shoot through zero states, capacitors C1 and C2 are charging and have zero inductor voltage. Meanwhile, when the switch is in shoot-through zero states, the capacitors C 1 and C 2 discharge, and a voltage will appear on the inductor. Hence, this condition leads to the inductor voltage is similar to the capacitor voltage. The magnitude of the voltage across the inductor and the capacitor has the same value of about 72 V, which is close to the calculation result of 65. 4   In Figure 13, the peak DC-link voltage of the experiment reaches up to 88 V. The DC-link voltage has zero value at shootthrough condition, which means the inverter leg is in a short circuit. There is non-linear relation between the peak of DC-link and capacitor voltage, but it can be regulated using capacitor voltage control as proposed in [28]. This experiment did not insert the DC-link capacitor to make the distortion of DC-link voltage. Hence, the inverter side is not supplied by pulsed currents. Consequently, it distorts the input source due to the reflect phenomena. Moreover, super-imposed spikes and noises in the DC-link voltage of ZSI are caused by semiconductor switching and massive inrush current due to the shoot-through condition [29]. In addition, the inductor current increase during shoot-through condition with an average value is 2.1 A in this experiment, which is slightly different from its design current of 1.92 A. Briefly, the value in the experiment is higher than both calculations and simulation process.

D. Testing of ZSI Output Voltage and Current
When ZSI is loaded by 120 Ω, ZSI can keep the voltage constant at 60 Vrms and the current of 0.239 A, respectively.
Both variables produced a period of about 20 ms that means they can form 50 Hz frequency in the load side. Figures 14(a) and 14(b) show the output voltage and current waveforms, respectively. The resulting waveform by ZSI is also 120° apart, proving that it produces a three-phase waveform as described in Figure  15. A passive filter on the output side can produce the sine waveform. The total harmonic distortion (THD) at the output voltage is also analyzed, as shown in Figure 16. An enormous RMS voltage value is found at the fundamental frequency of 50 Hz, while the other RMS voltage values are spread over the nth order frequency. The THD value of the output voltage waveform is 7.43%. This value is relatively low, so it can be concluded that the simple boost control technique works appropriately to reduce the harmonic output. However, the THD value still must be minimized more until the requirement of standard THD is below 5%. The conversion ratio test aims to determine the voltage increase factor in the ZSI circuit. The conversion ratio test is conducted by supplying a constant input voltage of Vin = 52 V, and the shoot-through duty ratio is increased gradually in the range of 0 -20%, with the final voltage output target is 60 V. The higher shoot-through duty ratio, the larger output voltage that will be obtained. Based on the test, the final output voltage is reached in a shoot-through duty ratio of 17%, slightly different from the simulation test as shown in Figure 17. The conversion ratio at this shoot-through duty ratio is about 2.05 times.

F. Testing of ZSI Efficiency
The efficiency testing is carried out to determine the effect of inverter performance. In this test, the efficiency is determined by comparing the input power with the inverter output power. Inverter efficiency testing increases the shootthrough duty ratio at a different load. Figure 18 shows the ZSI efficiency from the experiment, with an average efficiency of 80%. ZSI has the highest efficiency of 84.78% at 50 W (50% of rating load) and has the lowest efficiency of 75.14% at 13 W (13% of rating load). In general, the load addition causes the accession of ZSI efficiency until a specific load, then decreases until full load. The losses factor causes this condition due to the voltage drop for shoot-through duty ratio addition. Voltage drops that occur in ZSI are caused by a non-ideal factor on inductors, capacitors, and semiconductor components [30]. In addition, the voltage drops also occur due to the switching losses, core inductor losses, fast switching diode loss, and charge-discharge capacitor loss. In [31] explained that the efficiency of ZSI experienced a massive drop at low loads. Moreover, by increasing voltage gain, the efficiency tends to decrease. The two previous problem makes the large power circulation in the ZSI topology. The current flow becomes more extensive, increasing the power losses in the inductor, capacitor, and diode components. Figure 18 shows that the efficiency drop occurs before reaching the rating load. This characteristic is appropriate to [31], wherein the highest efficiency is achieved before nominal loads.
Based on [32], the SBC method has a maximum shootthrough duty ratio of 1 -M, which means only 8% of the modulation index of 0.92. Meanwhile, the shoot-through design of ZSI in this experiment is 17%. Hence, it is clear that the shoot-through value used is higher than the allowable operating value on the SBC. This condition possible to make a higher shoot-through value and affects the efficiency decrease. Therefore, designing ZSI using SBC should consider the shootthrough limits allowed on SBC and choose the appropriate modulation index.

G. Result Comparison
Compared to the VSI and CSI topology connected to the buck or boost converter, ZSI can increase and decrease voltage only on a single circuit. In addition, ZSI only has one stage of power conversion, no need for dead time control, and is more economical because it uses fewer switching devices, as shown in Table IV Compared to similar experiments on ZSI, in [33] designed a ZSI with conventional PWM for a 6 W power rating that used the identical capacitor (C1 and C2) and inductor (L1 and L2) of 470 uF and 10.1 mH, respectively. Using the same capacitor and 1.5 mH less of the inductor, ZSI with SBC in this study can apply on a rating of 100 W. Furthermore, in terms of the control mechanism, this study only used one microcontroller while [33] added another device that only functioned as a bridge between the main microcontroller and the voltage sensing. Hence, this results in higher experimental costs. Table V shows several ZSI topologies with various switching methods. By employing the smaller shoot-through duty ratio in the comparable value of the modulation index, the ZSI with the SBC method can generate a boosting factor with a higher value. This result is more advantageous than the other method producing a boosting factor with a higher shootthrough duty ratio. For the THD aspect, ZSI based on SBC has an output voltage with THD of 3.16% and 7.43% in simulation and experiment, respectively. Surprisingly, simulation yields better THD than 7-levels ZSI controlled by maximum boost control (MBC), which has a THD of 10.13% [34].
The THD comparison of ZSI using SBC also yield a better result compared to ZSI using other methods such as Trapezoidal PWM (TPWM), Sinusoidal PWM (SPWM), and Space Vector PWM (SVPWM) [34]. Moreover, experimental research with similar consideration for line-to-line voltage of 60.78 Vrms and modulation index of 0.9 in [35] has THD value of 28.39% and 37.38% of both switching control applied. The output is obtained from an input voltage of 60 Vdc, which means the boosting factor is much smaller than ZSI using the SBC method.
The experimental result from [37] shows that the ZSI with SBC has an anomaly result because the output waveform of the voltage is distorted due to the auxiliary power supply. On the contrary, the present result with the same method yields a better result in which the output waveform of the voltage is smoother and undamaged. Hence, this study can rectify the voltage quality from the previous experimental study.
IV. CONCLUSION This paper proposes a laboratory-scale MHP system with ZSI based on SBC. This research has been carried out both from simulations and experiments. A rewinding induction generator is used for this MHP system and functions as a ZSI voltage supply through a rectifier circuit. The waveform characteristic of ZSI is similar to the theoretical result, and its value is close to the desired design. The output voltage 60 V is reached in a shoot-through duty ratio of 17%, and this condition has a conversion ratio of about 2.05 times. SBC strategy can reduce harmonic output significantly until 7.43%. The efficiency of ZSI has a value of 84.78% at 50% of rating load 100 W and an average value of 80%. In general, the efficiency of ZSI rises with the increase in total load was given. Compared to the previous study, the existing experiment with the same method yields an economical design using the high rating power with the same component. Also, it has a smoother and undamaged output waveform of the voltage. This study is proven the limitation of ZSI design that should be considered. Next, changing the controller method and designing a filter with an appropriate value to reach THD less than 5% could be future research work.