Bended Dipole Antenna Parameter Estimation by Using Artificial Neural Network

This paper investigates the feasibility of using artificial neural networks (ANN) to establish the design parameters of the bended dipole antenna. The feedforward backpropagation based on the Scaled Conjugate Gradient (SCG) algorithm was applied to develop the ANN model. The proposed ANN model includes two hidden layers, and each hidden layer contains ten neurons. A total of 66 datasets were used for training, testing, and validating the proposed ANN model to identify bending positions and angles of a dipole antenna with a working frequency range of 1.6–2.2 GHz. Simulation and experimental validations were conducted to test the proposed method. Both simulation and experimental results showed that the proposed method has the potential to provide dipole antenna design parameters in a fast and cost-effectively manner.