Velocity control strategy based on a cloud model for unmanned agricultural vehicle during obstacle crossing

Objective  To improve the intelligence and safety of remote-operated agricultural vehicles.

Method  A new method for unmanned agricultural vehicles during obstacle crossing was proposed. The collision prediction model of unmanned vehicles in dynamic environment was established, and the real-time collision location was determined. According to the cloud inference principle based on both experience of experts and agricultural operation environment, velocity control strategy was established to realize velocity control.

Result  The algorithm took 0.170 1 s on average to make a prediction, the velocity control results of unmanned vehicles excluded the impact of non-threatening obstacles and accorded with the velocity cloud inference principle.

Conclusion  The established algorithm is able to realize real-time collision prediction, possesses anti-disturbance ability, and satisfies real-time requirement.