A new microcontroller-based MPPT algorithm for photovoltaic applications

In this work, a new and improved microcontroller-based maximum power point tacking (MPPT) algorithm for photovoltaic (PV) applications is proposed and implemented. The algorithm incorporates new ideas for overcoming the challenges associated with rapidly changing insolation levels and the effect of partial shading. It is simple with ( i.e. no complex computations) and can be implemented on commercial micro-controllers. The proposed algorithm enhances the steady-state and dynamic responses by introducing an improved adaptive step-size for updating the PV module's reference variable ( i.e. voltage, current or duty cycle). This new adaptive step-size approach exploits the first derivative of power as a function of duty cycle ( dP/dD ) and the sign of the second derivative for dividing the panel's operating range into four different regions. Consequently, the PV module's operating point can be tracked more precisely, thereby leading to more accurate step-size update compared to traditional step-size update. The instability issue, under rapidly changing insolation levels, is addressed by incorporating a current measurement I null at the end of each sampling period. The measured I null is used to estimate power dP 2 caused by the insolation change during the sampling period, which is compared to power change dP 1 caused by MPPT update. The algorithm also considers the issue of partial shading, where multiple peaks appear in the power function of the PV modules. The proposed algorithm exploits the relation between the maximum power current I MP and the global maximum power P m ( i.e. P m = Ü I mp ) to estimate the global maximum. Based on this relation, periodic interrupt routines are invoked to estimate the expected global maximum power P m of the present operating current I pv using ( P m = Ü I pv ). This power P m is then compared with the actual output power P o to ensure that the module is operating at the global maximum. The proposed MPPT system is modeled in SIMULINK with the theoretical models of a PV module and a buck converter. Simulation results are presented to validate the algorithm performance under different irradiation schemes, and are then compared to the results of several conventional algorithms ( e.g. P&O, adaptive ICM). In addition, a hardware prototype is implemented where the experimental results are presented and compared to a conventional algorithm