Neural network training for serial multisensor of autonomous vehicle system

<span>This study aims to find the best artificial neural network weight values be applied autonomous vehicle system with ultrasonic multisensor. The implementation of in required long time process due its training process. Therefore, this research is using offline before implementing online by embedding obtain outputs faster according desired targets. Simulink were used train offline. Eight sensors are on all sides and arranged a serial multisensory configuration as inputs network. With eight inputs, one sixteen-depth hidden layer, five outputs, it was trained back-propagation algorithm By 100000 iterations, output target almost same, indicating convergency minimum errors. result weights networks. These can implemented fixed-weight training.</span>