FAULT DIAGNOSIS OF PV MODULE DURING MPP TRACKING FOR DISTRIBUTED MPP CONTROLLER

Abstract

In order to, overcome the problem of conventional (P&O) algorithm. In this thesis I have used neural network to track the MPP in short time. This overcomes the problem of large number of central periods required for MPPT. Similarly once the MPP is track conventional (P&O) algorithm is small perturbation steps is used to maintain the operation of the PV submodule close to the MPP in the steady state. The use of small perturbation steps overcomes the problem of steady state power lose cause by using conventional (P&O) with large perturbation steps.