This invention is a new method that can in real time predict resistant force expected on the digging/scooping tool (digging bucket or alternative tools) of off-road machineries such as a wheel loader or an excavator. The method features integration of artificial intelligence (AI) and a physics-based prediction approach. It can improve force prediction accuracy by 50% compared to conventional force prediction methods, and ensures high computational efficiency for on-line implementation. This invention contains a process/method for predicting resistance force on a digging/scooping tool, given the planned control command of the digging tool motion trajectory. This invention is important for control of off-road autonomous machineries/vehicles, since on-line control command should be generated based on the resistant force expected to ensure reliable digging control and energy efficiency of vehicle operation. The method integrates neural networks in AI and physics-based modeling to ensure a highly accurate and computationally efficient force prediction. The invention can also be used for off-line force analysis to aid machine design.
Name: Sheikh Ismail