Problem to be solved: to provide a learning system and a learning method capable of suppressing the lowering of learning accuracy of an operation inference learning model caused by the difference in pedal play amount between a vehicle model and a real vehicle.Based on the running condition of the vehicleAn operation Reasoning Learning Model 70 which deduces a vehicle operation andA learning system 10 comprising an autopilot robotOperationAccelerator pedal and brake pedal either or both pedalIncluding pedal operationVehicles andBased on the difference of the pedal play amount with the vehicle model 52 which outputs the simulated running state, theInput pedal manipulated by an inference inference learning model into a vehicle modelFirst pedal play quantity adjustment part 54The second pedal play amount adjustment unit 55 and.We propose a learning system for an operation reasoning learning model which controls an automatic maneuvering robot by applying an adjusted simulated running state to an operation reasoning learning model.Diagram
展开▼