**5. Hardware Implementation**

The complexity and cost of the proposed algorithms are promising factors that pave the way for real implementation. Therefore, as a proof-of-concept, the M-VSS-P&O algorithm is implemented on hardware to validate the usefulness of the proposed algorithm. The proposed M-VSS-P&O algorithm is implemented for a low-voltage PV module using Arduino Due, which is equipped with a Cortex-M3 CPU. To analyze and visualize the output voltage and current of the PV module, the output data are exported using GDM Digital Multi Meter (DMM) software through the serial port. DMM software receives the data from the DMM and saves it in an MS Excel sheet. The experimental setup of the MPPT technique is demonstrated in Figure 16a. The setup includes Arduino Due, a voltage and current sensor (MAX471), a boost converter, and a 30-Ω load. The PV module used in the

experimental study is shown in Figure 16b while the STC outputs of the module are shown in the datasheet in Figure S3 in the Supplementary Materials.

**Figure 16.** (**a**) Hardware Implementation and (**b**) PV module.

Figure 17 shows the characteristic curves of the PV module used, as it was tested before using the MPPT algorithm in order to acquire the *I-V* and *P-V* curves. The PV module is tested by recording the current and voltage using DMM software. The maximum power that PV modules can deliver is measured as 15.24 W at a voltage of 15.4 V and a current of 0.98 A as is depicted in Figure 17.

**Figure 17.** Practical *I-V* and *P-V* curves of the PV module.

The response of the conventional P&O is shown in Figure 18 where the PV voltage is plotted in Figure 18a while the PV power is plotted in Figure 18b. As can be inferred from the figure, there is a certain delay in tracking the maximum power in addition to the obvious oscillations. On the other hand, the modified P&O algorithm tracked the maximum power on the *P-V* curve of the solar panel and continued to operate the PV modules at that point as evident in Figure 19. Figure 19a displays the PV voltage while Figure 19b shows the PV power waveforms of the modified algorithm. The figure clearly indicates the improvement in the M-VSS-P&O algorithm as there are no oscillations around the MPP in addition to the response time being reduced. The modified technique tracks the power of the PV module to 14 W, which exhibits an efficiency of 92%. The main reason for the difference in efficiency between the simulation and practical implementation can be attributed to the resolution of sensors.

**Figure 18.** (**a**) PV Voltage and (**b**) PV Power of conventional P&O algorithm.

**Figure 19.** (**a**) PV Voltage and (**b**) PV Power of modified P&O algorithm.
