Abstract This paper introduces an offline testing method for motor parameters in a frequency converter based on traditional electrical machinery testing principles, and describes measures to achieve high accuracy of the tested parameters. Keywords Vector control; motor parameters; offline testing 0 Introduction In the vector control system of an asynchronous motor, the motor parameters are crucial physical quantities. In electrical machinery theory, equivalent circuits are constructed using motor parameters, which serve as the basis for analyzing various operating characteristics of three-phase motors. The performance of vector control in variable frequency speed regulation depends entirely on the accuracy of the motor parameters used. Inaccurate parameters will directly lead to a decrease in vector control performance indicators, and may even cause the frequency converter to malfunction. The basic parameters of a three-phase asynchronous motor include stator resistance, stator leakage inductance, rotor resistance, rotor leakage inductance, and stator-rotor mutual inductance. These parameters can be determined theoretically using technical data from motor design and manufacturing, but this calculation is complex and has a large error compared to reality; alternatively, they can be determined experimentally. The following section specifically describes the experimental testing method for each parameter in a frequency converter. 1. Parameter Measurement Experiment In frequency converters, there are two main methods for testing parameters: online testing and offline testing. Online testing methods mainly include Kalman filtering, model reference adaptive testing, and sliding mode variable structure testing. These methods require high processor speed and have high system hardware requirements. Offline testing methods mainly include frequency response testing and step response testing, but the testing accuracy is not high, and there are problems such as computational complexity and large program computation load, so they are rarely used. This section mainly introduces the offline testing of motor parameters in frequency converters based on traditional motor testing principles, and takes corresponding measures to achieve high accuracy of the test parameters. 1.1 Testing the stator resistance of the motor using the DC volt-ampere method In the frequency converter system, the key to testing the stator resistance using the DC volt-ampere method is how to obtain a low-voltage DC power supply. When the frequency converter is directly connected to the power grid, its DC bus voltage is high. The usual method is to perform voltage chopping control on the DC bus to obtain a series of high-frequency voltage pulses with a very low average value, fixed period, and fixed duty cycle. After filtering by the inductor in the stator winding, a DC current with very small pulsation is obtained. If the duty cycle is D, the DC bus voltage is Udc, and the current is I, then the corresponding stator resistance value is... In order to prevent overcurrent in the inverter during testing, the duty cycle setting should be correctly considered. In actual testing, the duty cycle can be controlled by using a current closed-loop control with a PI regulator, as shown in Figure 1. In Figure 1, I* is the control target, i.e., the given current, and I is the feedback current, i.e., the actual operating current. During testing, the on-state voltage drop of the IGBT switching device in the inverter has a significant impact on the test value. The correctness of the on-state voltage drop compensation directly affects the accuracy of the test resistance. Besides the IGBT's on-state voltage drop, there is also the on-state voltage drop of the freewheeling diode. Furthermore, since the IGBT has a certain delay during turn-on and turn-off, this delay cannot be ignored in order to accurately calculate the output DC voltage. 1.2 Testing Rotor Resistance and Stator/Rotor Leakage Inductance Using a Locked-Rotor Test In practical applications, locking the motor is difficult. Therefore, a single-phase short-circuit test is used instead of a three-phase test. When a single-phase sinusoidal voltage is applied to the motor, no electromagnetic torque is generated, and the electromagnetic phenomena are basically the same as in a three-phase locked-rotor test. In the test, one phase of the motor is open-circuited, and a single-phase sinusoidal alternating current is applied between the other two phases, followed by a certain current. The voltage, current, and input power on the stator are then measured, allowing the calculation of the motor's short-circuit resistance and short-circuit reactance. To prevent overcurrent in the locked-rotor test, a PI closed-loop control can be added by measuring the stator resistance. However, the effects of the IGBT voltage drop and the freewheeling diode voltage drop must also be considered. In inverter operation testing, to prevent shoot-through between the upper and lower bridge arms, a dead time, typically 3–5 μs, must be added between the upper and lower switches of the same bridge arm. Since the erroneous pulses generated by the dead time have a significant impact on system performance, this must be carefully considered. Therefore, for high-precision parameter testing, dead time is a crucial factor that cannot be ignored. Figure 3 shows the relationship between the actual output voltage and the given PWM pulse, considering the effects of dead time and the IGBT turn-on and turn-off times. As can be seen from Figure 3, the dead time affects the actual output voltage. When the load current is greater than zero, the actual phase voltage output is reduced by td + tr - tf compared to the control requirement; when the load current is less than zero, the actual phase voltage output is increased by td + tr - tf compared to the control requirement. In practice, if the PWM pulse is adjusted accordingly based on the sampled motor current, the control deviation introduced by the inverter system can be compensated, thus ensuring that the actual output voltage of the inverter system matches the control requirements. 1.3 Testing Mutual Inductance Between Stator and Rotor Using No-Load Test In the no-load test of a three-phase asynchronous motor, since the motor is in a no-load state, the rotor current is very small, and the slip can be approximated as 0. The equivalent circuit is shown in Figure 4. In the figure, rm is the excitation resistance, and jXm is the mutual inductance reactance. If the rated voltage measured in the test is U1, the no-load current is I1, and the no-load input power is P0, then the no-load impedance Z0, short-circuit resistance Rk, and short-circuit reactance Xk can be calculated. Thus, the reactance and inductance in the equivalent circuit of the asynchronous motor under no-load conditions can be calculated using the calculation method for the locked-rotor test. The motor excitation reactance is... In the test, the control of the switching delay and dead time of the power devices is crucial. The introduction of dead time causes the output voltage to drop and the phase to drift, resulting in a smaller motor current and a larger measured impedance. To improve the measurement accuracy of the no-load test, a combination of current feedforward and current feedback methods can be used in actual testing. The magnitude of the output voltage is determined by detecting the polarity and amplitude of the current. 2. Experimental Verification A 11 kW/380V frequency converter was used to drive a 4 kW motor for parameter testing. The frequency converter's control CPU used a TMS320F2812 chip. The test results were compared with those from conventional motor testing. In conventional motor testing, a Fluke F41B tester was used to measure voltage, current, and power to obtain accurate parameters. The test results and parameters are listed in Tables 1, 2, and 3, respectively. The experimental data shows that the results from the frequency converter test are very close to those from conventional motor testing, with an error generally within 5%. The experiment demonstrates that the accuracy of the parameter testing fully meets the requirements of vector control, which is of particular importance and reference value for accurately determining the parameters of a three-phase asynchronous motor. 3. Conclusion The experimental testing of the parameters of a three-phase asynchronous motor is crucial for the frequency converter to perform vector control. Only by accurately inputting the motor parameters can the vector control function be fully utilized, thus enabling reliable control of the motor. The method described in this article is much more accurate than simply relying on the data on the motor nameplate, and is a more ideal way to obtain motor parameters.