| Summary: | DC-DC converters are extensively applied in lots of electronics devices to provide a stable voltage output. The control of converters and parameter estimation are two subjects of many researches. For simplicit, PID controllers become popular in various application. Based on the mathematical definition, the application of PID controller greatly influences the overall performance of converters. As the improvement of Artificial intelligence (AI) technology, many evolutionary algorithms are introduced to deal with parameters optimization of PID controllers. This paper selects a DC-DC buck converter as a study target, applying Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to design the parameters of the PID controller. The results are evaluated by performance indexes like overshoot and steady-state time. On the other hand, parameter estimation plays a crucial part in fault prediction. BP Neural Network (BPNN) is applied to estimate passive components of buck converter by measuring output voltage and current ripples. With the aim of accuracy enhancement of the network, GA is combined with BPNN to search the optimal initialization of the network.
|
|---|