Motor inverter dead zone compensation is unique and high-performance servo is no problem

A review of inverter dead-time compensation methods

To compensate for the voltage fluctuation caused by TD, researchers have proposed various compensation methods, which can be roughly divided into three categories.

The most common method is to add a pulse train with opposite polarity to the missing pulse train in the interval with the same current polarity to offset its influence. Since one phase of the three-phase current must have opposite polarity to the other two phases, a simple method is to perform double voltage overcompensation on the phase with opposite polarity so that the dead zone effects of the three-phase voltage offset each other and the line voltage waveform is sinusoidal. The literature analyzes the causes and effects of the dead zone in detail, and gives hardware circuit compensation methods for the dead zone based on analog modulation and digital modulation respectively. According to the switching state of the full-bridge circuit, a mathematical model of an inverter with dead zone compensation is proposed. The characteristic of this model is that it is composed of a simple hysteresis loop structure. According to this model, dead zone compensation can be achieved by a calculation formula.

The second method is to achieve dead zone compensation based on the principle of invalid devices. At any time, only one of the two power devices in each bridge arm of the inverter is valid. When the upper bridge arm device is turned off, regardless of whether the lower bridge arm device is turned on or not, the output voltage is the negative terminal voltage of the DC bus. At this time, the lower bridge arm device is called "invalid". The method of dead zone compensation is to maintain the drive signal of the valid device unchanged and change the drive signal of the invalid device to meet the requirements of setting the dead zone. Since the on and off of the "invalid" device does not affect the output voltage state, there is no need for a drive signal. Only the drive signal is sent to the valid device. In this way, there is no need to add a dead zone, and there is no problem of dead zone compensation. However, this method will cause distortion due to errors at the current zero crossing point. Therefore, when using this method, attention should be paid to the processing of the current zero crossing area. Some scholars have further proposed improved methods. Adding a hysteresis loop at the current zero crossing point and using normal switch dead zone protection during the hysteresis loop time can reduce distortion. Due to the interference in current sampling and the complexity of current changes, two drive signals should be given in the area near the current zero crossing point and dead zones and dead zone compensation should be added. By utilizing the PWM off time to realize the switch dead zone protection during commutation, the influence of the switch dead zone can be eliminated.

The third method is current predictive control. Establish a more accurate motor system model, analyze the distortion of the current waveform, and realize the correction of the current waveform through current predictive control. Predict the dead zone problem of current control, and compensate for the voltage waveform distortion and current zero-point clamping phenomenon by estimating the back electromotive force. Establish the matrix equation of the asynchronous motor model, and compensate its spatial voltage vector based on the prediction of the stator phase current in the SVPWM algorithm. Design the observer using the PMSM model in the DQ rotating coordinate system, observe the voltage loss of the Q-axis, and convert it into the dead time TC to be compensated through the formula to realize online compensation of the dead zone. Time delay control is used to estimate the interference voltage caused by the dead zone, and it is fed back to the voltage reference to compensate for the dead zone. The current prediction method is cumbersome to calculate, and the compensation effect is directly related to the accuracy of the motor model and the time-varying parameter values, so it is not easy to obtain satisfactory results. Influence of inverter dead zone

From the basic principle of PWM dead time generation, it can be known that the voltage deviation of the inverter output voltage caused by the deviation pulse in the winding current cycle t1 can be equivalent to a square wave. For the convenience of analysis, it is assumed that the voltage deviation pulses are equally spaced in time, and the height of the equivalent square wave is:

As the current polarity changes, the direction of the error voltage pulse also changes, and as the carrier frequency increases, the number of error voltage pulses also increases. Although the dead time is very short, only a few microseconds, the error voltage accumulates within a cycle and will also have a greater impact on the fundamental amplitude of the output voltage. The qualitative relationship between the error voltage and the ideal voltage and the actual output voltage is shown in Figure 2.

Performing Fourier analysis on the deviation square wave in Figure 2 yields:

Where, ω1 is the fundamental angular frequency of the current; ψ is the phase difference between the desired voltage and the motor current.

Therefore, ignoring the high-frequency noise caused by the power switch, the output voltage of the inverter is:

Among them, ma modulation is the ratio of the modulated sine wave amplitude to the triangle wave carrier amplitude.

It can be seen from the above formula that due to the existence of the inverter dead time, not only the fundamental wave of the inverter output voltage changes, but also the output voltage contains higher-order harmonics such as the 3rd, 5th, and 7th orders.

The switching dead zone causes the inverter output voltage waveform to be distorted, resulting in the output current waveform distortion, that is, the current crossover distortion.

The longer the dead time is, the greater the inverter output fundamental voltage loss is, and the greater the voltage waveform distortion is; the more the load fundamental current amplitude decreases, the more serious the current waveform distortion is.

For a certain dead time, when the load power factor becomes smaller, the inverter output fundamental voltage amplitude will increase, the voltage waveform distortion rate will decrease, the fundamental current amplitude will decrease, and the current waveform distortion rate will increase.

When the output voltage is low, the space voltage vector amplitude is small, the relative conduction time of the three bridge arms becomes shorter, and the influence of the dead time becomes larger.

The dead zone not only affects the output voltage amplitude, but also its phase; the dead zone makes the PWM waveform no longer symmetrical about the center, so the amplitude of the space voltage vector deviates and the phase also changes. Position-based dynamic dead zone compensation method

A common feature of various dead zone compensation methods is to compensate the voltage signal according to the current waveform. Therefore, it is necessary to detect the actual current value, determine the positive and negative current of each phase, and use its zero crossing point as the switching moment of the compensation voltage signal. The current detection link consists of a current sensor, a low-pass filter and an A/D converter. In order to reduce noise, digital filtering is also required in the program. The detected current has errors caused by device accuracy and interference, and there is a phase delay. Therefore, it is difficult to accurately compensate for the dead zone effect using the actual detected current signal, and even greater current distortion may be caused due to incorrect compensation near the zero crossing point.

Nowadays, the torque control of PMSM is mostly achieved through vector control. In order to accurately control the motor current, the current loop response frequency is very high, which can reach more than 1kHz, and the actual current can accurately track the current command signal. In high-precision AC servo systems, high-resolution position sensors are required to achieve high-precision position servo control, generally reaching 16 or 17 bits, while high-speed and high-precision A/D devices are relatively expensive, and their resolution is generally 10 or 12 bits. Since the current vector is related to the rotor position, if the position signal is used to determine the positive and negative current, and the voltage dead zone compensation signal is applied, the compensation accuracy can be higher than the actual current signal accuracy, and it is not affected by interference signals.

It can be seen from the pmsm vector diagram that the current vector of the field-oriented control is 90° (electrical angle) with the rotor pole and rotates synchronously with the rotor. The position of the rotor pole can be determined by a high-resolution encoder. After the rotor field-oriented control, the electrical angle of the current changing with time and the spatial rotation angle of the magnetic pole changing in space coincide, and then the spatial position of the current vector can be obtained. According to the spatial position of the current vector, the zero-crossing point of each phase current can be determined.

The phase relationship between the magnetic pole position angle and the current is fixed. After analysis, we perform voltage compensation according to the following position change law:

When the angle is 0 < θ < π, ia > 0, the a phase compensates the forward voltage; otherwise, it compensates the reverse voltage.

When the angle is 2π/3 <θ <5π/3, ib>0, the b phase compensates the forward voltage; otherwise, it compensates the reverse voltage.

When the angle is -2π/3 <θ <π/3, ic>0, the c phase compensates the forward voltage; otherwise, it compensates the reverse voltage.

The amplitude calculation formula of the compensation voltage is:

Where factor is the adjustment coefficient, which is generally taken as 0.7.

Figures 4 and 5 are comparisons of the experimental results without and with dead zone compensation. It can be seen from the current waveform that the current without dead zone compensation is distorted at the zero crossing point.

After adding the dead zone compensation method proposed above, the actual current shown in Figure 5 tracks the given current and obtains a good sinusoidal waveform.

The switching dead zone effect of the inverter has a great influence on the performance of the AC servo system, so it is necessary to correct and compensate the switching dead zone. Based on the analysis of various dead zone compensation methods, this paper proposes a dynamic compensation method based on position detection signals. This method uses a high-resolution encoder to improve the accuracy of current direction judgment, and experiments have shown that it has a good compensation effect.