//! \brief Computes the scale factor needed to convert from torque created by flux and Iq, from per unit to Nm
//!
_iq USER_computeTorque_Flux_Iq_pu_to_Nm_sf(void)
{
float_t FullScaleFlux = (USER_IQ_FULL_SCALE_VOLTAGE_V/(float_t)USER_EST_FREQ_Hz);
float_t FullScaleCurrent = (USER_IQ_FULL_SCALE_CURRENT_A);
float_t maxFlux = (USER_MOTOR_RATED_FLUX*((USER_MOTOR_TYPE==MOTOR_Type_Induction)?0.05:0.7));
float_t lShift = -ceil(log(FullScaleFlux/maxFlux)/log(2.0));
return(_IQ(FullScaleFlux/(2.0*MATH_PI)*FullScaleCurrent*USER_MOTOR_NUM_POLE_PAIRS*1.5*pow(2.0,lShift)));
} // end of USER_computeTorque_Flux_Iq_pu_to_Nm_sf() function
扭矩计算准确度如何?
Eric Ma:
之前有客户测试在5%左右的误差。
The accuracy of our torque estimation is related to the accuracy of the flux estimate and the Iq estimate. Both of these are dependent on the accuracy of the angle calculation, which is worse at lower speeds and higher current levels.
Also keep in mind that our torque estimate is electromagnetic torque (Tem), not shaft torque. Electromagnetic torque does not take into consideration motor losses such as core losses, friction losses, and rotor windage losses. But a shaft torque sensor does. I suspect the biggest loss is core loss. It might explain why our torque estimate is always on the high side.
这是我们产品线回答。
ERIC