Abstract: A method of performing automatic tuning on a deep learning model includes: utilizing an instruction-based learned cost model to estimate a first type of operational performance metrics based on a tuned configuration of layer fusion and tensor tiling; utilizing statistical data gathered during a compilation process of the deep learning model to determine a second type of operational performance metrics based on the tuned configuration of layer fusion and tensor tiling; performing an auto-tuning process to obtain a plurality of optimal configurations based on the first type of operational performance metrics and the second type of operational performance metrics; and configure the deep learning model according to one of the plurality of optimal configurations.

Abstract: A drive device and a drive control method for the electric bicycle are disclosed. The drive device of the electric bicycle is equipped with a power supply/storage unit, a compact motor, an inclination sensor, a brake actuation sensor, a pedal-rotating speed sensor and a drive control module. The drive device performs integrated calculations using the inclination information of the inclination sensor, the brake actuation information of the brake actuation sensor and the pedal-rotating speed information of the pedal-rotating speed sensor, so the drive control modules instructs the compact motor and the power supply/storage unit to quickly enter an auxiliary drive mode for rotating wheels of the electric bicycle, or to enter an energy recovery mode for storing electrical energy generated by the compact motor.

Abstract: A drive device and a drive control method for the electric bicycle are disclosed. The drive device of the electric bicycle is equipped with a power supply/storage unit, a compact motor, an inclination sensor, a brake actuation sensor, a pedal-rotating speed sensor and a drive control module. The drive device performs integrated calculations using the inclination information of the inclination sensor, the brake actuation information of the brake actuation sensor and the pedal-rotating speed information of the pedal-rotating speed sensor, so the drive control modules instructs the compact motor and the power supply/storage unit to quickly enter an auxiliary drive mode for rotating wheels of the electric bicycle, or to enter an energy recovery mode for storing electrical energy generated by the compact motor.

Abstract: The present disclosure related to a control method for a sensorless motor with energy recovery ability, using which duty cycle of a sensorless motor can be changed by the control of complementary switches so as to enable the sensorless motor to switch between a high efficiency driving operation and an energy recovery operation while being activated. Thereby, the conduction loss and the wear and tear to the switches can be minimized while simultaneously enhancing the energy recovery efficiency and thus improving the battery life.

Abstract: The present disclosure related to a control method for a sensorless motor with energy recovery ability, using which duty cycle of a sensorless motor can be changed by the control of complementary switches so as to enable the sensorless motor to switch between a high efficiency driving operation and an energy recovery operation while being activated. Thereby, the conduction loss and the wear and tear to the switches can be minimized while simultaneously enhancing the energy recovery efficiency and thus improving the battery life.