Multiresolution controller based on WaveNets for nonlinear systems

The design of robust and free-model controllers has been an open problem for many years. Although many approaches have been reported in the literature, the PID controller is the most used for its simplicity and speed response under some conditions. Thus, in many approaches, PID-like controllers are...

Descripción completa

Detalles Bibliográficos
Autores: García-Castro, Oscar Federico, Vega-Navarrete, Mario Alejandro, Ramos-Velasco, Luis Enrique, García-Rodríguez, Rodolfo, Escamilla-Hernández, Enrique
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2023
País:México
Institución:UNIVERSIDAD AUTÓNOMA DEL ESTADO DE HIDALGO
Repositorio:PÄDI Boletín Científico de Ciencias Básicas e Ingeniería del ICBI
Idioma:español
OAI Identifier:oai:repository.uaeh.edu.mx:article/11391
Acceso en línea:https://repository.uaeh.edu.mx/revistas/index.php/icbi/article/view/11391
Access Level:acceso abierto
Palabra clave:Multi-resolution analysis
Intelligent control
Wavelet function
Quanser helicopter
Neural network
Análisis multiresolución
Control inteligente
Función wavelet
Helicóptero Quanser
Redes neuronales
Descripción
Sumario:The design of robust and free-model controllers has been an open problem for many years. Although many approaches have been reported in the literature, the PID controller is the most used for its simplicity and speed response under some conditions. Thus, in many approaches, PID-like controllers are a recurrent controller structure. In recent years, due to the systems carrying more complex tasks subject to different environmental conditions, as in aircraft or robots, artificial intelligence tools such as neural networks have been used to self-tune feedback gain controllers. Based on the wavelet transformation, this paper proposes a multi-resolution proportional controller (PMR) scheme to control non-linear systems, avoiding system information. The PMR decomposes the tracking error to obtain different information on scale and frequency, which allows for compensation for various uncertainties in the system. The wavelet~neural network (WaveNet) information approximates the dynamic input-output model and is used for self-tuning feedback gains. Numerical simulation results for a 2-degree-of-freedom Quanser helicopter are presented, under different conditions, to verify the performance of the PMR controller.