Main Article Content
In hybrid electric vehicle, there are two sources of energy, namely Internal Combustion Engine (ICE) and DC motor. ICE as the prime mover has a smaller capacity than conventional vehicles because of the work is assisted by the DC motor. The DC motor acts to help internal combustion engine reach the torque and the speed as desired. Torque control of hybrid electric vehicle provide of how much torque required by the DC motor to assist the performance of ICE. When the ICE are not able to maintain the speed, the DC motor will help to provide the power. To overcome these problems, neuro-fuzzy predictive methods using inverse models are used. Neuro-fuzzy controller has the advantage of adaptability when the parameters in the system change. HEV itself requires a quick response therefore predictive controller used in order to predict the future value of the torque. Testing results showed that neuro-fuzzy predictive method which combines neurofuzzy controller with inverse models, able to assist ICE follows the reference model. The use of neuro-fuzzy predictive showed better control performance. This is shown from the speed response in 0.25 seconds able to produce a torque of 0.161 N-m, so that the HEV system can follow the desired reference model.