A comparative study is done in regards to the performance of the popular Perturb and Observe algorithm and the Genetic Assisted-Radial Basis Function-Neural Network (GA-RBF-NN) algorithm, both incorporating the Interl...A comparative study is done in regards to the performance of the popular Perturb and Observe algorithm and the Genetic Assisted-Radial Basis Function-Neural Network (GA-RBF-NN) algorithm, both incorporating the Interleaved Boost converter. The Perturb and Observe method (P&O) is inarguably the most commonly used algorithm as its advantages pertaining to its ease in implementation and simplicity enable to track the Maximum Power Point (MPP). However, it is absolutely unreliable when subjected to rapidly fluctuating irradiation and temperature levels. More importantly, the system has the tendency to swing back and forth about the Maximum Power Point without reaching stability. At this juncture, the implementation of the Genetic-Assisted Radial Basis Function (GA-RBF) algorithm helps the system achieve MPP at a shorter time when compared to the Perturb and Observe technique. The ever reliable and robust Levenberg-Marquardt algorithm is included along with the MPPT controller that minimizes the Mean Square Error (MSE) and aids in faster training of the neural network. This PV system drives a brushless DC motor (BLDC), employing rotor position sensors.展开更多
文摘A comparative study is done in regards to the performance of the popular Perturb and Observe algorithm and the Genetic Assisted-Radial Basis Function-Neural Network (GA-RBF-NN) algorithm, both incorporating the Interleaved Boost converter. The Perturb and Observe method (P&O) is inarguably the most commonly used algorithm as its advantages pertaining to its ease in implementation and simplicity enable to track the Maximum Power Point (MPP). However, it is absolutely unreliable when subjected to rapidly fluctuating irradiation and temperature levels. More importantly, the system has the tendency to swing back and forth about the Maximum Power Point without reaching stability. At this juncture, the implementation of the Genetic-Assisted Radial Basis Function (GA-RBF) algorithm helps the system achieve MPP at a shorter time when compared to the Perturb and Observe technique. The ever reliable and robust Levenberg-Marquardt algorithm is included along with the MPPT controller that minimizes the Mean Square Error (MSE) and aids in faster training of the neural network. This PV system drives a brushless DC motor (BLDC), employing rotor position sensors.
