CN105656379A - Magnetic resistance characteristic coordinate transformation-based switched reluctance motor position estimation method - Google Patents

Abstract

The invention discloses a magnetic resistance characteristic coordinate transformation-based switched reluctance motor position estimation method. According to the method, on the basis of magnetic resistance data of four special positions, the position of a switched reluctance motor is estimated according to a phase voltage value and a phase current value. A rotor position is divided into two regions; the square of the rotor position angle in the region I and the magnetic resistance value are in a linear function relation; and one second power of the rotor position angle in the region II and the 1/4th power of the magnetic resistance value are in a quadratic function relation. Fitting coefficients a, b, c, d and e are calculated through the magnetic resistance data of the four special positions; the phase voltage value and the phase current value are detected; the current magnetic resistance value is calculated; the region in which the current rotor position is located is judged. If the rotor position is located in the region I, the current position is calculated by the fitting coefficients a and b; if the rotor position is located in the region II, the current position is calculated by the fitting coefficients c, d and e. The position of the switched reluctance motor is estimated in an analysis manner; the method has the advantages of being low in cost, easy to implement, good in applicability and the like.

Description

A kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform

Technical field

The present invention relates to a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform, belong to electric machine without position sensor control field.

Background technology

In switched reluctance machines drive system, rotor position information is most important for the realization of control method and the normal operation of system, and this information is obtained by mechanical position sensor such as photoelectric coding device, rotating transformers usually. But, the existence of mechanical position sensor adds the complicacy of system architecture, it is to increase manufacturing cost, and sensor performance is easily affected by environment. Therefore, research is applicable to the low cost of switched reluctance machines, high precision, high reliability method for controlling position-less sensor be very important.

In order to realize position Sensorless Control, researchist proposes a large amount of position predictor methods. These methods are mainly divided into two classes: non-conduction phase position preestimating method and conducting phase position preestimating method.

First kind method is mainly through giving the non-conduction phase injecting voltage pulse of switched reluctance machines, and under constant frequency low duty ratio voltage drive, winding will produce the detection electric current of low amplitude value, and owing to detection electric current is less, therefore now the counter potential of motor can be ignored. According to voltage balance equation it will be seen that the detection amplitude of electric current and current position phase inductance are inverse ratio, therefore the method can be used for position and estimates. It is characterized in detecting initial position of rotor, continuous print positional information can be detected under middle low speed, but this method is not suitable for high-speed working condition.

2nd class method is mainly based on the flux linkage characteristic of switched reluctance machines, carry out storing or stating with the form such as question blank, neural network by measuring the flux linkage characteristic data obtained, then according to the conducting phase current recorded online and magnetic linkage data, obtain rotor position information. Such method has higher estimate accuracy and the wider rotating speed scope of application, but need a large amount of flux linkage characteristic sampled data, and these data are normally by detailed finite element analysis or what experiment measuring obtained, which increase complicacy and the cost of method, often need to take bigger physics internal memory in realization simultaneously.

Summary of the invention

The relation curve of switched reluctance machines phase magnetic resistance and rotor position is divided into two regions by the present invention, and take different magnetoresistive characteristic coordinate transform modes for different zones, using the rotor position after conversion as independent variable(s), magnetic resistance after conversion, as dependent variable, is expressed as linear function and quadratic function in different zones. On this basis, just can solve rotor position according to magnetic resistance. Technical scheme is as follows:

Step one: in the relation curve of switched reluctance machines phase magnetic resistance and rotor position, by interval [��₀,��₁] it is defined as region I, [��₁,��_a] it is defined as region II. In region I and region II, define four specific position ��₀����₁����_hr����_a. Wherein, ��₀For non-alignment position, ��_aFor aligned position, ��₁And ��_hrCan obtain by formula (1) and (2).

${\mathrm{\θ}}_1={\mathrm{\θ}}_a-\frac{{\mathrm{\β}}_s+{\mathrm{\β}}_r}{2}---\left(1\right)$

${\mathrm{\θ}}_{hr}={\mathrm{\θ}}_a-\frac{{\mathrm{\β}}_r}{2}---\left(2\right)$

Wherein, ��_sAnd ��_rIt is respectively stator poles arc and rotor pole arc.

Step 2: detection ��₀����₁����_hr����_aThe flux linkage characteristic data �� of four positions₀����₁����_hr����_a, the magnetoresistive characteristic of four positions is obtained by formula (3), wherein, ��_xRepresent ��₀����₁����_hr����_aThe phase magnetic linkage of four positions, N is the phase winding number of turn, and i is phase current.

$R_x=\frac{N^{2}i}{{\mathrm{\ψ}}_x}---\left(3\right)$

The R that will calculate₀��R₁��R_hr��R_aSubstitution formula (4) obtains 5 coefficients a, b, c, d, e.

$\left\{\begin{array}{c}\left[\begin{array}{cc}a& b\end{array}\right]=\left[\begin{array}{cc}{R}_0& R_1\end{array}\right]{\left[\begin{array}{cc}{\mathrm{\θ}}_0^2& {\mathrm{\θ}}_1^2\\ 1& 1\end{array}\right]}^{-1}\\ \left[\begin{array}{ccc}c& d& e\end{array}\right]=\left[\begin{array}{ccc}{R}_1^{\frac{1}{4}}& R_{hr}^{\frac{1}{4}}& R_a^{\frac{1}{4}}\end{array}\right]{\left[\begin{array}{ccc}{\mathrm{\θ}}_1& {\mathrm{\θ}}_{hr}& {\mathrm{\θ}}_a\\ {\mathrm{\θ}}_1^{\frac{1}{2}}& {\mathrm{\θ}}_{hr}^{\frac{1}{2}}& {\mathrm{\θ}}_a^{\frac{1}{2}}\\ 1& 1& 1\end{array}\right]}^{-1}\end{array}\right.---\left(4\right)$

Step 3: detection conducting phase voltage, current value, utilize formula (5) to calculate current phase magnetic resistance value.

$R=\frac{N^{2}i}{\mathrm{\ψ}\left(0\right)+{\∫}_0^t(u-ir)dt}---\left(5\right)$

Wherein, �� (0) is initial magnetic linkage, owing to silicon steel material remanent magnetism is less, usually �� (0) is taken as 0; U, i and r are respectively the phase voltage of switched reluctance machines, phase current and phase resistance.

Step 4: utilize linear interpolation, obtains �� under current electric current₁The magnetic resistance value R at place₁(i)��

Step 5: if R >=R₁I (), shows that rotor position is positioned at region I, utilize linear interpolation to calculate a (i) and b (i). Now magnetic resistance and rotor position angle square are that linear function relation is such as formula, shown in (6), obtaining rotor position angle by formula (7).

R=a (i) ��²+b(i)(6)

$\mathrm{\θ}=\sqrt{\frac{R-b\left(i\right)}{a\left(i\right)}}---\left(7\right)$

Step 6: if R is < R₁I (), shows that rotor position is positioned at region II, utilize linear interpolation to calculate c (i), d (i) and e (i). Now 1/4 power of magnetic resistance and 1/2 power of rotor position angle are that quadratic function relation is such as formula, shown in (8), obtaining rotor position angle by formula (9).

$R^{\frac{1}{4}}=c\left(i\right)\mathrm{\&theta;}+d\left(i\right){\mathrm{\&theta;}}^{\frac{1}{2}}+e\left(i\right)---\left(8\right)$

$\mathrm{\&theta;}={\left(\frac{-d\left(i\right)-\sqrt{d{\left(i\right)}^2-4c\left(i\right)\left(e\right(i)-R^{\frac{1}{4}})}}{2c\left(i\right)}\right)}^2---\left(9\right)$

Step 7: as needed to reduce the systematic error introduced when low current lower rotor part position is estimated further, heterogeneous magnetoresistive characteristic can be adopted to replace single-phase reluctance characteristic to estimate. Phase selection principle is: according to the size of each phase current, and the position that carries out mutually choosing phase current maximum is estimated.

The useful effect of the present invention: 1. method is simple, is easy to realize. After magnetoresistive characteristic is carried out coordinate transform, rotor position angle can be solved by fast resolving; Only need the flux linkage characteristic data of four rotor positions, and only take a small amount of physics internal memory; 2. precision height, strong robustness. The magnetic resistance order of magnitude is relatively big, resolving power height. In addition, by heterogeneous preestimating method, systematic error is reduced; 3. suitability is good. Under Angle-domain imaging, Current cut control and voltage PWM control operating mode, all there is good precision, also it is applicable to different switched reluctance machines topologys.

Accompanying drawing explanation

Fig. 1 is the relation curve figure of switched reluctance machines phase magnetic resistance and rotor position angle under certain electric current.

Fig. 2 is the relation curve figure of certain electric current lower area I switch magnetoresistance motor rotor position angle square with phase magnetic resistance.

Fig. 3 is the relation curve figure of certain 1/2 power at electric current lower area II switch magnetoresistance motor rotor position angle with 1/4 power of phase magnetic resistance.

Embodiment

Below in conjunction with accompanying drawing and specific examples, the technical scheme of the present invention is described in detail. Example motor used is a 1kW three-phase 12/8 pole switching reluctance motor.

Step one: Fig. 1 is the relation curve figure of switched reluctance machines phase magnetic resistance and rotor position angle under certain electric current. By [��₀,��₁] it is defined as region I, [��₁,��_a] it is defined as region II. For the switched reluctance machines that example is given, ��_s����_rAnd ��_aIt is respectively 15 ��, 17 �� and 22.5 ��. By formula (1) and (2), it is possible to draw ��₁And ��_hrIt is respectively 6.5 �� and 14 ��.

Step 2: utilize rotor position fixation method, can obtain this switched reluctance machines at 0 ��, 6.5 ��, the flux linkage characteristic data at 14 �� and 22.5 �� places, and then is obtained the magnetic resistance value R of these four positions by formula (3)₀�㡢R_6.5�㡢R₁₅�㡢R_22.5��. The R that will calculate₀�㡢R_6.5�㡢R₁₅�㡢R_22.5�� substitute into formula (4) obtain 5 coefficients a, b, c, d, e.

Step 3: detection conducting phase voltage, current value, utilize formula (5) to calculate phase magnetic resistance value R (i) now.

Step 4: utilize linear interpolation, calculates R_6.5��(i)��

Step 5: if R (i) >=R_6.5�� (i), shows that rotor position is positioned at region I, and now the relation curve of switch magnetoresistance motor rotor position angle square and phase magnetic resistance is as shown in Figure 2. Utilize linear interpolation to calculate a (i) and b (i), and then utilize formula (7) to obtain rotor position angle.

Step 6: if R (i) is < R_6.5�� (i), shows that rotor position is positioned at region II, and now the relation curve of 1/2 power at switch magnetoresistance motor rotor position angle and 1/4 power of phase magnetic resistance is as shown in Figure 3. Utilize linear interpolation to calculate c (i), d (i) and e (i), and then utilize formula (9) to obtain rotor position angle.

Step 7: as needed to reduce further the rotor position predictor error under low current, adopt three-phase magnetoresistive characteristic to replace single-phase reluctance characteristic to estimate, estimates that to choose strategy mutually as shown in table 1.

Table 1. is estimated and is chosen strategy mutually

Claims (7)

1. the switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform, it is characterized in that: the relation curve of switched reluctance machines phase magnetic resistance and rotor position is divided into two regions, and the magnetoresistive characteristic after linear function and the conversion of quadratic function denotation coordination and the relation between the rotor position angle after coordinate transform is adopted respectively for different zones, and then solving for rotor position, the performing step of the position predictor method that the present invention announces is as follows:

Step one: in the relation curve of switched reluctance machines phase magnetic resistance and rotor position, by interval [��₀,��₁] it is defined as region I, [��₁,��_a] it is defined as region II, in region I and region II, define four specific position ��₀����₁����_hr����_a;

Step 2: the flux linkage characteristic data measuring above-mentioned four specific position places, and calculate corresponding magnetoresistive characteristic data R₀��R₁��R_hr��R_a; The magnetic resistance data of four positions are substituted into the analytical expression of coefficient a, b, c, d, e, calculates a, b, c, d, e five coefficients;

Step 3: detection conducting phase voltage, current value, substitute into magnetic resistance analytical expression, obtain current phase magnetic resistance value R;

Step 4: utilize linear interpolation, calculates R₁(i);

Step 5: if R >=R₁I (), shows that rotor position is positioned at region I, now rotor position angle square is linear function relation with magnetic resistance, and after can utilizing coordinate transform, the linear function expression formula of magnetoresistive characteristic carries out rotor position and estimates;

Step 6: if R is < R₁I (), shows that rotor position is positioned at region II, now 1/2 power of rotor position angle and 1/4 power of magnetic resistance are quadratic function relation, and after can utilizing coordinate transform, the quadratic function expression formula of magnetoresistive characteristic carries out rotor position and estimates;

Step 7: as needed to reduce the systematic error introduced when low current lower rotor part position is estimated further, heterogeneous magnetoresistive characteristic can be adopted to replace single-phase reluctance characteristic to estimate, phase selection principle is: according to the size of each phase current, and the position that carries out mutually choosing phase current maximum is estimated.

2. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: four specific position �� described in step one₀����₁����_hr����_aMiddle ��₀For non-alignment position, ��_aFor aligned position, ��₁And ��_hrCan by formula ��₁=��_a-(��_s+��_r)/2 and ��_hr=��_a-��_r/ 2 calculate, wherein, and ��_sAnd ��_rIt is respectively stator poles arc and the rotor pole arc of motor.

3. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: the �� described in step 2₀����₁����_hrAnd ��_aThe magnetoresistive characteristic R at place₀��R₁��R_hr��R_aIt is pass through formulaCalculating, wherein, �� is phase magnetic linkage, and N is the phase winding number of turn, and i is phase current.

4. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: coefficient a, b, c, d, e described in step 2 pass through formula $\left\{\begin{array}{c}\left[\begin{array}{cc}a& b\end{array}\right]=\left[\begin{array}{cc}{R}_0& R_1\end{array}\right]{\left[\begin{array}{cc}{\mathrm{\&theta;}}_0^2& {\mathrm{\&theta;}}_1^2\\ 1& 1\end{array}\right]}^{-1}\\ \left[\begin{array}{ccc}c& d& e\end{array}\right]=\left[\begin{array}{ccc}{R}_1^{\frac{1}{4}}& R_{hr}^{\frac{1}{4}}& R_a^{\frac{1}{4}}\end{array}\right]{\left[\begin{array}{ccc}{\mathrm{\&theta;}}_1& {\mathrm{\&theta;}}_{hr}& {\mathrm{\&theta;}}_a\\ {\mathrm{\&theta;}}_1^{\frac{1}{2}}& {\mathrm{\&theta;}}_{hr}^{\frac{1}{2}}& {\mathrm{\&theta;}}_a^{\frac{1}{2}}\\ 1& 1& 1\end{array}\right]}^{-1}\end{array}\right.$ Calculate.

5. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: the phase magnetic resistance value described in step 3, can represent and beWherein, �� (0) is initial magnetic linkage, owing to silicon steel material remanent magnetism is less, usually �� (0) is taken as 0, u, phase voltage, phase current and phase resistance that i and r is respectively switched reluctance machines.

6. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: after region I coordinate transform described in step 5, the linear function expression formula of magnetoresistive characteristic is R=a (i) ��²+ b (i), now rotor position angle is by formulaSolve.

7. a kind of switched reluctance machines position predictor method based on magnetoresistive characteristic coordinate transform according to claim 1, it is characterised in that: after the region II coordinate transform described in step 6, the quadratic function expression formula of magnetoresistive characteristic is $R^{\frac{1}{4}}=c\left(i\right)\mathrm{\&theta;}+d\left(i\right){\mathrm{\&theta;}}^{\frac{1}{2}}+e\left(i\right),$ Now rotor position angle is by formula $\mathrm{\&theta;}={\left(\frac{-d\left(i\right)-\sqrt{d{\left(i\right)}^2-4c\left(i\right)\left(e\right(i)-R^{\frac{1}{4}})}}{2c\left(i\right)}\right)}^2$ Solve.