Control device for electric vehicle

A controller for an electric vehicle configured to drive a drive wheel with a motor, the controller comprising:

• a processor;
• and
• a storage device that stores a learned model learned through machine learning, wherein
• the processor is configured to perform vibration suppression control that suppresses vibration caused by torsional resonance of a drive shaft when the electric vehicle is traveling on an uneven road,
• the learned model is a model configured to predict whether the drive wheel is slipping on a road or gripping the road after a specified time by executing a calculation using, as an input, time-series data for a specified period of explanatory variables including a rotation speed ωf the motor, a torque generated by the motor, and a rotation speed ωf the drive wheel,
• the processor, in the vibration suppression control, uses a prediction result obtained with the learned model to determine whether to perform speed-increasing correction for generating a positive torque with the motor or speed-decreasing correction for generating a negative torque with the motor, and
• the processor, in the vibration suppression control, switches between the speed-increasing correction and the speed-decreasing correction based on the determination of whether to perform the speed-increasing correction or the speed-decreasing correction to increase the rotation speed ωf the motor when the drive wheel is slipping on the road and decrease the rotation speed ωf the motor when the drive wheel is gripping the road.