CN113783489A - Phase current double-correction method for single-direct-current bus sensor - Google Patents

Abstract

The invention aims to provide a double correction method for the phase current of a single direct current bus sensor, which is used for reducing reconstruction errors and distortion rate; the zero drift caused by sampling of a single sensor is screened out under the condition that redundant vectors are not inserted, complementary vectors with opposite vectors are screened out, and the zero drift compensation is completed through double sampling of dynamic current; and calculating the current value of the same phase at the same time of the slope prediction and the second sampling by dynamically sampling the same phase current twice, thereby completing the phase error compensation.

Description

Phase current double-correction method for single-direct-current bus sensor

Technical Field

The invention belongs to the technical field of single direct current bus sensor error correction, and particularly relates to a double correction method for phase current of a single direct current bus sensor.

Background

The three-phase full-bridge inverter circuit is widely applied to variable speed alternating current motor driving, uninterruptible power supplies and renewable energy interfaces. The real-time current information is important for the driving control of the permanent magnet synchronous motor, the phase current information is contained in the direct current bus current information, and the phase current reconstruction through the direct current bus sampling is established on the basis. The single-sensor direct-current bus sampling technology has the advantages of low cost, small size, capability of reducing current errors caused by inconsistent parameters of a plurality of current sensors and the like. However, because a set of current sensor and signal processing circuit is adopted, the zero drift error of the current sensor can be enlarged to all phase currents, and further the current reconstruction accuracy is influenced, and the error existing in the single-current sensor direct current bus phase current reconstruction path is shown in fig. 1. The method deeply analyzes the generation mechanism of the error expansion effect of the single current sensor, develops methods such as a zero drift correction strategy and the like, eliminates the influence of the error expansion effect on current reconstruction, realizes self-detection and self-correction of the zero drift error, and is the key for improving the measurement precision of the current detection method of the single current sensor.

For correcting error amplification effect, the current improvement scheme is as follows: the method comprises the following steps of analyzing the characteristics of voltage ripples in direct current, and eliminating the influence of current measurement errors by using a low-pass filter and a more advanced current controller. Using an integral link of a d-axis PI current regulator to output a signal, controlling the d-axis current to be zero or constant, and enabling the output signal to be zero or fixed; when the stator current contains a zero drift error, the output signal of the integral link of the d-axis PI current regulator generates fluctuation in the speed of the rotor, and the error compensation is carried out by direct current offset of the built-up area or readjustment of the input measurement gain of the stator current. And thirdly, eliminating the drift error of the alpha-beta axis by changing the installation position of the single sensor and changing the coordinate transformation of the clark, and estimating the zero drift amount in the sampling result by using a defect filter so as to compensate the direct current drift error in the current sensor. And fourthly, analyzing the current discrete waveform change of the zero drift error, proposing the concept of equivalent current change rate, and calculating the equivalent current change rate and the instantaneous current change rate by a method based on a state observer so as to carry out backtracking prediction on the current correction method.

However, the used prediction current method has large calculation amount, the calculation time caused by the algorithm is increased, and the current correction effect is not ideal; the stability and the anti-interference performance of the circuit structure of the sensor are poor when the mounting position of the sensor is changed, and the measurement error caused by the inconsistency of the sensor cannot be eliminated.

Disclosure of Invention

The invention aims to provide a double correction method for the phase current of a single direct current bus sensor, which is used for reducing reconstruction errors and distortion rate.

The technical scheme for solving the technical problems of the invention is as follows: a double correction method for phase current of a single direct current bus sensor is characterized by comprising the following steps:

s1: vector control;

s2: updating Ta, Tb and Tc to generate a complete SVPWM vector, wherein Ta is the duty ratio of an A-phase PWM waveform, Tb is the duty ratio of a B-phase PWM waveform, and Tc is the duty ratio of a C-phase PWM waveform, and generating PWM waveforms containing opposite vectors by a random phase-shifting method, and enabling the action time of the vectors to be consistent;

s3: determining sampling time, sampling a direct current bus, and sampling three times in PWM (pulse width modulation) containing opposite vectors, wherein two sampling values i are obtained at the middle point of normal adjacent vectors_{sam_b}、-i_{sam_c}For phase current reconstruction, and the later constructed opposite vector intermediate point is adopted once more to obtain-i_{sam_b}For zero drift correction and phase delay compensation;

s4: carrying out zero drift correction to obtain a drift amount delta i_shDrift and corrected three-phase current value

S5: sampling the three-phase current subjected to zero drift correction and performing phase delay compensation to obtain a three-phase current value subjected to secondary correction;

s6: the phase current reconstruction is performed, and after the reconstruction is completed, the process proceeds to step S1.

The step S4 specifically includes: handle i_{sam_b}+(-i_{sam_b}) Is divided by 2 to obtain a zero point drift amount Δ i_shThen subtracting delta i from the three-phase current sampling values respectively_shTo obtain corrected three-phase current values

The correction of the zero point drift is completed.

The step S5 specifically includes: obtaining the first sampling value of A phase after zero drift correction

Sampling time T_{sam_a}(k-1), second sampled value

Sampling time T_{sam_a}(k) (ii) a The second sampling time of the C phase is T_{sam_c}Namely T'_{sam_a}(k) The error in the phase time,

ΔT_{ph_err}＝T′_{sam_a}(k)-T_{sam_a}(k)

for the correction value

And the sampling time T_{sam_a}(k-1)、T_{sam_a}(k) The calculation of the slope in between is carried out,

in the formula

Obtaining the corrected sampling time of the B phase and the A phase for the corrected predicted current value consistent with the sampling time of the C phaseA consistent predicted current value.

The invention has the beneficial effects that:

compared with the traditional single current sensor phase current reconstruction, the method focuses on the current reconstruction accuracy problem, provides zero drift caused by factors such as drift, inconsistent amplifier internal parameters and temperature change playing a leading role due to the fact that the output accuracy and stability of a voltage reference chip are affected by initial accuracy, temperature, noise and the like, and provides a dynamic current double-sampling high-accuracy current correction strategy for ensuring the consistency of opposite vector slope and sampling time_shThe elimination is about 50%.

In the traditional PMSM control, two sensors are used to simultaneously obtain two-phase current information under an ideal condition, so that phase delay does not exist in sampling, but phase error occurs due to time difference of two-phase sampling in current single-current-sensor phase current reconstruction, and phase compensation is not performed. Providing dynamic two times of in-phase current sampling, calculating the current value of which the slope is estimated and the same time and the same phase as the second time of sampling, and completing the phase time error delta T_{ph_err}And then phase error compensation is completed.

And thirdly, on the basis of the original single current sensor phase current reconstruction, the method completes analysis and compensation of inherent errors caused by the sensor and sampling, improves the detection precision of the phase current, and provides important guarantee for a vector control system and a protection strategy of the motor.

Drawings

FIG. 1 is a schematic diagram of the error in the current reconstruction of the DC bus of the single current sensor of the present invention.

FIG. 2 is a control block diagram of a single DC bus sensor phase current double correction device and a control strategy system thereof.

Fig. 3 is a control block diagram of the legacy device and its control strategy system of the present invention.

FIG. 4 is a flow chart of the dual phase current correction for a single DC bus sensor of the present invention.

FIG. 5 shows the invention V₁₀₀And V₀₁₁The equivalent circuit diagram is used.

Fig. 6 is a schematic diagram of the zero drift correction of the present invention.

Fig. 7 is a schematic diagram of the phase delay correction of the present invention.

Detailed Description

The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

First, a phase current double correction device of the single dc bus sensor proposed in this patent and a control scheme thereof are given, as shown in fig. 2. The double correction is mainly completed: (1) the zero drift caused by sampling of a single sensor is screened out under the condition that redundant vectors are not inserted, complementary vectors with opposite vectors are screened out, and the zero drift compensation is completed through double sampling of dynamic current; (2) the current value of the same phase and the same time of the slope prediction and the second sampling is calculated through dynamic two times of in-phase current sampling, phase error compensation is completed, and the original device is used for reconstructing current after obtaining current information through sampling of a direct current bus and then transforming the current information as shown in figure 3. The difference of fig. 2 is that the current information is sampled, then the zero drift correction and the phase delay compensation are performed, and then the current information is reconstructed.

The difference is that after the current information is sampled, because the current information contains errors, if the current information is directly reconstructed, the errors are introduced into a closed loop, so that the current information needs to be corrected when reconstruction is not performed, and two dotted line boxes in fig. 2 represent two correction methods respectively.

FIG. 4 is a flow chart of a single DC bus sensor phase current double correction device and a control strategy control flow chart thereof.

As shown in fig. 3, the present invention comprises the steps of:

s1: vector control;

s2: updating Ta, Tb and Tc to generate a complete SVPWM vector, wherein Ta is the duty ratio of an A-phase PWM waveform, Tb is the duty ratio of a B-phase PWM waveform, and Tc is the duty ratio of a C-phase PWM waveform, and generating PWM waveforms containing opposite vectors by a random phase-shifting method, and enabling the action time of the vectors to be consistent;

s3: determining sampling time, sampling a direct current bus, and sampling three times in PWM (pulse width modulation) containing opposite vectors, wherein two sampling values i are obtained at the middle point of normal adjacent vectors_{sam_b}、-i_{sam_c}For phase current reconstruction, and the later constructed opposite vector intermediate point is adopted once more to obtain-i_{sam_b}For zero drift correction and phase delay compensation;

s4: carrying out zero drift correction to obtain a drift amount delta i_shDrift and corrected three-phase current value

S5: sampling the three-phase current subjected to zero drift correction and performing phase delay compensation to obtain a three-phase current value subjected to secondary correction;

s6: the phase current reconstruction is performed, and after the reconstruction is completed, the process proceeds to step S1.

The step S4 specifically includes: handle i_{sam_b}+(-i_{sam_b}) Is divided by 2 to obtain a zero point drift amount Δ i_shThen subtracting delta i from the three-phase current sampling values respectively_shTo obtain corrected three-phase current values

The correction of the zero point drift is completed.

The step S5 specifically includes: obtaining the first sampling value of A phase after zero drift correction

Sampling time T_{sam_a}(k-1), second sampled value

Sampling time T_{sam_a}(k) (ii) a The second sampling time of the C phase is T_{sam_c}Namely T'_{sam_a}(k) The error in the phase time,

ΔT_{ph_err}＝T′_{sam_a}(k)-T_{sam_a}(k)

for the correction value

And the sampling time T_{sam_a}(k-1)、T_{sam_a}(k) The calculation of the slope in between is carried out,

in the formula

And obtaining the corrected predicted current values consistent with the sampling time of the B phase and the A phase in the same way for the corrected predicted current values consistent with the sampling time of the C phase.

Zero drift error correction principle:

taking the first sector as an example, define the first sampling value as i_{sam_a}The actual value is i_{re_a}The second sampling value is i_{sam_c}The actual value is i_{re_c}Zero point drift amount is Deltai_shTo obtain three-phase current i_{sam_a}、i_{sam_c}And by a three-phase current KCL: i.e. i_a+i_b+i_cI calculated as 0_{KCL_b}As shown in the formula:

the phase B current calculated from the three-phase current KCL has no zero drift error, and the error phase conditions of other sectors are shown in table 1.

TABLE 1 sector and error Current correspondences

As can be seen from the formula (1), the sampled current information has zero drift, and the other phase does not have zero drift, so that i is taken as an example of the first sector and has zero drift error_aAnd-i_cSimilarly, it is summarized in table 1 that the other five sectors contain zero drift errors, so that each sector has two current information.

Slope derivation:

the action of the voltage vector is realized in the form of switching on and off of an open tube of the inverter, and the discrete switching state causes the current chopping effect to generate a T_sSeven segments of current waveforms occur within the cycle. The determination of the current change rate must be based on the theoretical circuit equivalent of the voltage vector. The correspondence relationship between the voltage vector and the phase voltage in each sector is shown in table 2.

TABLE 2 voltage vector to phase voltage correspondence

Is the voltage vector and the phase voltage, U_aN、U_bN、U_cNRepresenting the phase voltage, the voltage on the DC bus being U_dc. The three-phase voltage output by the inverter is U_aN、U_bN、U_cNWhich are respectively applied to planar coordinate systems spatially differing from each other by 120 degrees.

The eight voltage vectors have their thevenin equivalent circuits corresponding to them, the phase current change is derived from the motor voltage equation, since the resistance voltage component is lower than the inductance voltage component, neglecting the resistance component, the current change rate can be obtained by thevenin equivalent circuit of the eight voltage vectors, and V is used₁₀₀And V₀₁₁For example, the equivalent circuit is shown in FIG. 5, E_{be_a}、E_{be_b}、E_{be_c}Is A,B. C back-emf. Calculate i_aThe current change rate is calculated as

In other cases, the rate of change of each phase current when other vectors are applied is shown in table 3.

TABLE 3 correspondence between voltage vectors and phase current change rates

Resistance components and back electromotive force are ignored at low rotating speed, time change is microsecond level in the current correction process, and the correction influence is small. Through the above derivation of the rate of change of current, three sets of vectors (V)₁(100) And V₄(110))、(V₂(110) And V₅(001) And (V)₃(010) And V₆(101) Are vectors of opposite vectors, corresponding to the same opposite current information, and opposite rates of change of current, e.g., V₃(010) Corresponding current information i_b，V₆(101) Corresponding current information-i_b. It is proposed on the basis of this to correct the "zero drift" by means of an opposite vector. As shown in the zero drift correction diagram of fig. 6.

Determining the zero drift amount Deltai_sh：

Due to the combination with V₆(101) In the action time period, the current is constantly changed due to the chopping effect caused by the discrete action of the voltage vector, so that the sampling values have equal amplitude in order to ensure the correction precision, i.e. i_{re_b}＝|-i_{re_b}If the current slope is the same as the sampling time, as shown in the formula

In the formula, i_{max_b}The phase B current is at V₃(010) Maximum value in the period of action, T_sFor PWM period, T_samIs the first sampling time, T'_samIs the second sampling time.

Satisfying the formula, the formula can be obtained

Wherein the first sampling value is i_{sam_b}The second sampling value is i'_{sam_b}Thus, the drift amount Δ i can be obtained_sh，

The three-phase current after correcting the zero drift is as follows:

in the formula

There is no zero drift amount for the corrected three-phase current values.

Phase delay principle:

in an actual motor control strategy, the switching frequency, i.e., the PWM frequency, is above 10kHz, while the electrical frequency of the motor is much lower than the PWM frequency, which can be considered approximately as the current change to linearity within one PWM waveform, and the sampling value at the next time is predicted by the sampling values of the first two times, as shown in fig. 7.

Phase delay compensation amount calculation:

a-phase first sampling value after zero drift correction

Sampling time T_{sam_a}(k-1), second sampled value

Sampling time T_{sam_a}(k) (ii) a The second sampling time of the C phase is T_{sam_c}Namely T'_{sam_a}(k) The error in the phase time,

ΔT_{ph_err}＝T′_{sam_a}(k)-T_{sam_a}(k)

phase delay compensation:

correction value

And the sampling time T_{sam_a}(k-1)、T_{sam_a}(k) The calculation of the slope in between is carried out,

in the formula

The corrected predicted current value consistent with the sampling time of the phase C can be obtained according to the corrected predicted current value consistent with the sampling time of the phase B and the phase A, and the method specifically comprises the following steps:

the invention has the beneficial effects that:

compared with the traditional single current sensor phase current reconstruction, the method focuses on the current reconstruction accuracy problem, provides zero drift caused by factors such as drift, inconsistent amplifier internal parameters and temperature change playing a leading role due to the fact that the output accuracy and stability of a voltage reference chip are affected by initial accuracy, temperature, noise and the like, and provides a dynamic current double-sampling high-accuracy current correction strategy for ensuring the consistency of opposite vector slope and sampling time_shThe elimination is about 50%.

In the traditional PMSM control, two sensors are used to simultaneously obtain two-phase current information under an ideal condition, so that phase delay does not exist in sampling, but phase error occurs due to time difference of two-phase sampling in current single-current-sensor phase current reconstruction, and phase compensation is not performed. Providing dynamic two times of in-phase current sampling, calculating the current value of which the slope is estimated and the same time and the same phase as the second time of sampling, and completing the phase time error delta T_{ph_err}And then phase error compensation is completed.

And thirdly, on the basis of the original single current sensor phase current reconstruction, the method completes analysis and compensation of inherent errors caused by the sensor and sampling, improves the detection precision of the phase current, and provides important guarantee for a vector control system and a protection strategy of the motor.

Claims (3)

1. A double correction method for phase current of a single direct current bus sensor is characterized by comprising the following steps:

s1: vector control;

s2: updating Ta, Tb and Tc to generate a complete SVPWM vector, wherein Ta is the duty ratio of an A-phase PWM waveform, Tb is the duty ratio of a B-phase PWM waveform, and Tc is the duty ratio of a C-phase PWM waveform, and generating PWM waveforms containing opposite vectors by a random phase-shifting method, and enabling the action time of the vectors to be consistent;

s3: determining sampling time, sampling a direct current bus, and sampling three times in PWM (pulse width modulation) containing opposite vectors, wherein two sampling values i are obtained at the middle point of normal adjacent vectors_{sam_b}、-i_{sam_c}For phase current reconstruction, and the later constructed opposite vector intermediate point is adopted once more to obtain-i_{sam_b}For zero drift correction and phase delay compensation;

s4: carrying out zero drift correction to obtain a drift amount delta i_shDrift and corrected three-phase current value

S5: sampling the three-phase current subjected to zero drift correction and performing phase delay compensation to obtain a three-phase current value subjected to secondary correction;

s6: the phase current reconstruction is performed, and after the reconstruction is completed, the process proceeds to step S1.

2. The method for double correction of phase current of a single direct current bus sensor according to claim 1, wherein: the step S4 specifically includes: handle i_{sam_b}+(-i_{sam_b}) Is divided by 2 to obtain a zero point drift amount Δ i_shThen subtracting delta i from the three-phase current sampling values respectively_shTo obtain corrected three-phase current values

The correction of the zero point drift is completed.

3. The method for double correction of phase current of a single direct current bus sensor according to claim 1, wherein: the step S5 specifically includes: obtaining the first sampling value of A phase after zero drift correction

Sampling time T_{sam_a}(k-1), second sampled value

Sampling time T_{sam_a}(k) (ii) a The second sampling time of the C phase is T_{sam_c}Namely T'_{sam_a}(k) The error in the phase time,

ΔT_{ph_err}＝T′_{sam_a}(k)-T_{sam_a}(k)

for the correction value

And the sampling time T_{sam_a}(k-1)、T_{sam_a}(k) The calculation of the slope in between is carried out,

in the formula

And obtaining the corrected predicted current values consistent with the sampling time of the B phase and the A phase in the same way for the corrected predicted current values consistent with the sampling time of the C phase.

Images