CN101719752B - Method and device for detecting rotor position of brushless motor - Google Patents

Abstract

The invention relates to a method and a device for detecting the position of a rotor of a brushless motor, which are characterized in that a rotary transformer and the rotor of a tested brushless motor are arranged coaxially, the rotary transformer outputs two paths of orthogonal high-frequency sine-cosine signals containing information about the position of the rotor, and the information about the position of the rotor is obtained by the methods such as signal conditioning or resolving and the like. The most widely applied integrated chip at present adopting a special Resolver-to-Digital converter (RDC for short) and the like and a peripheral configuration circuit are used for performing resolving, and a special resolving chip of the type is sensitive to signals and has no fault-tolerantfunction, so the reliability is difficult to guarantee. The device of the invention is provided with a simple signal conditioning circuit on the basis of an RDC functional circuit to realize the detection of the position of the rotor of the brushless motor with self-monitoring and fault tolerance functions.

Description

Method and device for detecting rotor position of brushless motor

technical field

The invention relates to a method and a device for detecting the rotor position of a brushless motor, belonging to a method and a device for detecting the rotor position of a brushless motor control system.

Background technique

One of the key technologies of the brushless motor control system is the detection of the rotor position. Only after the actual spatial position (absolute position) of the rotor is detected, the control system can determine the power-on mode, control mode, and frequency and phase of the output current of the inverter. To ensure the normal operation of the brushless motor. Therefore, an accurate rotor position detection device is very important. Once the position detection device fails, the commutation logic of the motor will be confused, the output torque will decrease, the speed will decrease until it stops. In some occasions with high reliability requirements or certain specific occasions, such as: national defense, aerospace, etc., the stalling of the motor may cause personal injury and property loss. This requires that when the position signal fails, the motor can continue to work as much as possible without affecting the function of the entire system, and then troubleshoot when it is shut down. Therefore, it is very necessary to improve the reliability of the brushless motor rotor position signal.

Commonly used sensors for detecting the rotor position of brushless motors include absolute grating encoders, Hall sensors and resolvers. The absolute grating encoder directly converts the angle of the rotating shaft into a digital signal, which is simple and convenient to use, but it is difficult to be widely used due to factors such as environmental adaptability and price; the Hall sensor has a simple structure, but it is difficult to meet the requirements of high-precision angle measurement and is limited. Due to the advantages of reliable structure, good real-time performance, and strong environmental adaptability, resolvers are widely used in high-precision servo systems.

At present, in a large number of documents and patents at home and abroad, the research on the rotor position detection of brushless motors is limited to the conditioning method of the position signal, and the research on the reliability of the rotor position detection of brushless motors is limited to the Hall position sensor. The transformer brushless motor rotor position detection method is difficult to meet the high reliability requirements of national defense and aerospace.

When a resolver is used as the rotor position detection device of a brushless motor, the resolver outputs two high-frequency sinusoidal signals that contain rotor position information and are orthogonal to each other. The rotor position information must be obtained through signal conditioning or calculation methods, which is currently the most widely used. It is widely used integrated chips such as dedicated Resolver-to-Digital Converter (RDC) plus peripheral configuration circuits for solving. This kind of special solving chip is sensitive to signals and has no fault tolerance function, so the reliability Difficult to guarantee.

Contents of the invention

technical problem to be solved

In order to avoid the deficiencies of the prior art, the present invention proposes a method and device for detecting the rotor position of a brushless motor, which has self-monitoring and fault tolerance functions, and greatly improves the reliability of the brushless motor servo system.

Technical solutions

A method for detecting the rotor position of a brushless motor, characterized in that a resolver is installed coaxially with the rotor of the brushless motor to be tested, and the specific steps for detecting the rotor position of the brushless motor are as follows:

Step 1: Convert the two quadrature signals output by the secondary side of the resolver containing the information of the rotor position angle θ to differential to single-ended to obtain a single-ended sinusoidal signal E _s (t, θ) = Esin(ωt)sin(θ) and single-ended cosine signal E _c (t, θ)=Esin(ωt)cos(θ); the two-way signals containing rotor position angle θ information and orthogonal signals are: high-frequency sinusoidal differential signal E _s+ (t, θ) and E _s- (t, θ), high-frequency cosine differential signals E _c+ (t, θ) and E _c- (t, θ); where: E is the effective value of the signal, ω is the signal generated by the sine excitation circuit the angular frequency of the reference signal Refs;

Step 2: Convert the sinusoidal differential signals Refs+ and Refs- generated by the sinusoidal excitation circuit input to the resolver into a single-ended sinusoidal reference signal -Refs=Esin(ωt);

Step 3: superimpose the single-ended sine signal E _s (t, θ) and the reference signal Refs to obtain the signal Refs+E _s (t, θ), and superimpose the single-ended cosine signal E _c (t, θ) and Refs to obtain the signal Refs+E _c (t, θ);

Step 4: Filter out the high-frequency part in Refs+E _s (t, θ) and Refs+E _c (t, θ) to obtain an analog low-frequency sinusoidal signal E _s (θ)=Esin( θ)+E and low frequency cosine signal E _c (θ)=Ecos(θ)+E;

Step 5: Convert the analog low-frequency sinusoidal signal E _s (θ) = Esin(θ) + E and cosine signal E _c (θ) = Ecos(θ) + E into digital sine signal E _s ^* (θ) and digital cosine signal E _c ^* (θ);

Step 6: Calculate the average U of the digital signals E _s ^* (θ) and E _c ^* (θ);

Step 7: Subtract U from the digital sine signal E _s ^* (θ) to obtain the input quantity Y of the sine signal, and subtract U from the digital cosine signal E _c ^* (θ) to obtain the input quantity X of the cosine signal;

Step 8: Determine the quadrant of the rotor position of the motor according to the sine signal input value Y and the cosine signal input value X:

When X≥0 and Y≥0, the rotor position θ of the motor falls within the range of 0° to 90°;

When X<0 and Y≥0, the motor rotor position θ falls within the range of 90°～180°;

When X<0 and Y<0, the motor rotor position θ falls within the range of 180°～270°;

When X≥0 and Y<0, the motor rotor position θ falls within the range of 270°～360°;

然后根据上一步骤中θ所在的象限确定电机转子位置θ：Step 9: Map X and Y to the first quadrant, and use the CORDIC algorithm to calculate the angle value

Then determine the motor rotor position θ according to the quadrant where θ is located in the previous step:

When 0<θ≤90°, the motor rotor position

When 90°<θ≤180°, the motor rotor position

When 180°<θ≤270°, the motor rotor position

When 270°<θ≤360°, the motor rotor position

When the resolver output signal fails, the specific steps to detect the rotor position of the brushless motor are as follows:

Step a: Calculate the rotation angle of the motor in the two sampling intervals Δθ=ω _r ΔT, where ω _r is the motor speed, the unit is rad/s, the sampling frequency of analog-to-digital conversion is f, and the time interval between two samplings is ΔT = 1/f;

执行步骤4～5；Step b: _{Use the CORDIC inversion algorithm to obtain the sine value E sG *} ⁽ θ _G ) and the cosine value E _cG ^* (θ _G ) of θ G = θ+Δθ, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ and ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ Not all exceed the error limit, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ If it exceeds the error limit, it means that the cosine signal output by the resolver is faulty; the error limit is taken as the average value U of the digital signal

Execute steps 4 to 5;

Step c: If ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ and ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ All exceed the error limit, use the CORDIC inversion algorithm to get the sine value E _sG ^* (θ′ _G ) and the cosine value E _cG ^* (θ′ _G ) of θ′ _G = θ-Δθ, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}^{\&prime;}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G^{\&prime;}\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if ${\mathrm{\&epsiv;}}_{c1}^{\&prime;}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G^{\&prime;}\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ If it exceeds the error limit, it means that the cosine output of the resolver is faulty; the error limit is taken as the average value U of the digital signal

Step d: If the sine signal output by the resolver fails, convert the digital cosine signal E _c ^* (θ) in step 5 to obtain E _s ^* by using the orthogonal relationship between E _c ^* (θ) and E _s ^* (θ) ^* (θ), and replace the digital sine signal E _s ^* (θ) with E _s ^** (θ); if the cosine signal output by the resolver fails, the digital sine signal E _s ^* (θ) in step 5 is used The orthogonal relationship between E _s ^* (θ) and E _c ^* (θ) is converted to E _c ^** (θ), and E _c ^** (θ) is used to replace the digital cosine signal E _c ^* (θ);

Step e: Continue steps 6-9 to obtain the motor rotor position θ.

A device for realizing any of the above methods for detecting the rotor position of a brushless motor, characterized in that it includes a resolver, a sinusoidal excitation circuit, a differential signal converter, a precision absolute value adder, an inverter and a low-pass filter; The resolver is installed coaxially with the rotor of the brushless motor under test, and the high-frequency sinusoidal differential signal Refs+ generated by the sinusoidal excitation circuit is input to the primary side of the resolver, and the secondary side of the resolver outputs E _s+ ( t, θ) and E _s- (t, θ), E _c+ (t, θ) and E _c- (t, θ) respectively pass through two differential signal converters to obtain a single-ended sinusoidal signal E _s (t, θ) and single-ended cosine signal E _c (t, θ); the other way, the high-frequency sinusoidal differential signal Refs+ output by the sinusoidal excitation circuit is transformed into a single-ended sinusoidal reference signal-Refs input inverter through a differential signal converter, and the inverter outputs The Refs and the output signals of the two differential signal converters are respectively input to the respective precision absolute value adders, and the output signals Refs+E _s (t, θ) and Refs+E _c (t , θ); the output signals of the two precision absolute value adders are input to their respective low-pass filters, and the low-frequency sinusoidal signal E _s (θ)=Esin(θ)+ which is filtered by the low-pass filter and related to the rotor position E and cosine signal E _c (θ) = Ecos (θ) + E; the output of the low-pass filter is converted into a digital signal by the A/D module and then input to the DSP.

Beneficial effect

A method and device for detecting the rotor position of a brushless motor proposed by the present invention. When a rotary transformer is used as the rotor position detection device of a brushless motor, the rotary transformer outputs two high-frequency sinusoidal signals that contain rotor position information and are orthogonal to each other. The rotor position information can be obtained by means of signal conditioning or solution. The most widely used at present is the use of integrated chips such as a dedicated Resolver-to-Digital Converter (RDC) plus peripheral configuration circuits for resolution. This type of dedicated resolution chip is sensitive to signals and has no fault tolerance. function, so reliability is difficult to guarantee. Based on the RDC solving circuit, the present invention adds a simple signal conditioning circuit to realize the rotor position detection of the brushless motor with self-monitoring and fault-tolerant functions. The specific content includes two parts: detection device, self-monitoring and fault-tolerant method.

The advantages of the device of the present invention are: (1) the hardware circuit is simple, the cost is low, and the reliability of the system is greatly improved; (2) the CORDIC calculation algorithm is efficient and practical, and the motor can be realized through simple iteration and addition and subtraction operations. Fault-tolerant detection of rotor position.

Description of drawings

Figure 1: Circuit schematic diagram of the present invention

Figure 2: Schematic diagram of resolver principle

Fig. 3: the sinusoidal excitation circuit of the embodiment of the present invention

Figure 4: Differential signal converter of an embodiment of the present invention

Figure 5: Precision absolute value adder of an embodiment of the present invention

Figure 6: Inverter of an embodiment of the present invention

Figure 7: Low-pass filter circuit of the embodiment of the present invention

Detailed ways

Now in conjunction with embodiment, accompanying drawing, the present invention will be further described:

Implementation example of the device of the present invention:

The detection device is divided into two parts: a resolver installed coaxially with the rotor of the brushless motor under test and a resolver signal conditioning circuit. The resolver signal conditioning circuit includes a sinusoidal excitation circuit as a resolver signal source and self-monitoring and fault-tolerant algorithm signal conditioning. circuit (contains differential signal converter, precision absolute value adder, inverter, and low-pass filter).

The principle diagram of the resolver is shown in Figure 2. The resolver rotor r1 is coaxially installed with the motor rotor, Ur is the input terminal of the excitation signal, and U1 and U2 are the output terminals of the sine signal and cosine signal respectively.

The sinusoidal excitation circuit adopts an improved Wien bridge oscillator, as shown in Figure 3, U34A and U34B form an oscillation circuit, U36A and U36B form an emitter follower, and U35B, U37A and U37B form a circuit that outputs Refs-. The output sinusoidal excitation signal Refs=(Refs+)−(Refs−) is a sine wave with a frequency of 16 kHz and a peak value of 5V.

The differential signal converter consists of an operational amplifier TL082, as shown in Figure 4, the differential signals x- and x+ are respectively connected to the positive and negative input terminals of the operational amplifier through resistors R3 and R4, R1=R2=R3=R4, after conversion, they can A single-ended signal x is obtained, x=(x-)-(x+).

The precision absolute value adder is shown in Fig. 5, after the signal x and signal y pass through the operational amplifiers U1A and U1B and their peripheral circuits, the precision absolute value addition operation can be completed to obtain the signal z, z=|x+y|.

The inverter is very simple and consists of an operational amplifier. The circuit is shown in Figure 6. Input signal x, R2=R3 output signal y, y=-x.

The low-pass filter adopts a first-order active filter. The circuit is shown in Figure 7. The bandwidth is set at 0-2kHz, x is the input signal, and the output signal y is the envelope signal of x.

Through the above hardware circuit, the output signal of the resolver can be adjusted into a low-frequency analog signal that can be recognized by the digital signal processor. The DSP uses TMS320 series chips such as TMS320F2812, and the AD converter inside the DSP samples and converts the two input analog signals. It can be converted into a digital signal, and then the detection of the rotor position of the motor can be completed through the program solidified in the digital signal processor.

Realize the implementation example of the inventive method according to this device:

Step 1: Install the resolver coaxially with the rotor of the brushless motor under test. When the motor rotates, the secondary side of the resolver outputs two orthogonal differential signals, and through the differential converter, a single-ended sinusoidal signal E _s (t, θ )=Esin(ωt)sin(θ) and single-ended cosine signal E _c (t, θ)=Esin(ωt)cos(θ), where: E=5V, ω=16kHz; θ ranges from 0 to 360 degrees , the values in the following steps are the same as in this step;

Step 2: Convert the sinusoidal differential signals Refs+ and Refs- generated by the sinusoidal excitation circuit input to the resolver into a single-ended reference signal -Refs=Esin(ωt);

Step 3: superimpose the signal E _s (t, θ) and Refs to obtain the signal Refs+E _s (t, θ), and superimpose the signal E _c (t, θ) and Refs to obtain the signal Refs+E _c (t, θ );

Step 4: Filter out the high-frequency part in Refs+E _s (t, θ) and Refs+E _c (t, θ) to obtain an analog low-frequency sinusoidal signal E _s (θ)=Esin( θ)+E and low-frequency cosine signal E _c (θ)=Ecos(θ)+E, where the maximum values of E _s (θ) and E _c (θ) are both 10V and the minimum values are 0V.

Step 5: Convert the analog low-frequency sine signal E _s (θ)=Esin(θ)+E and cosine signal E _c (θ)=Ecos(θ)+E into digital sine signal E _s ^* (θ) and digital cosine Signal E _c ^* (θ), the integer of E _s ^* (θ) from 0 to 65535 corresponds to 0 to 10V of E _s (θ), and the integer of E _c ^* (θ) from 0 to 65535 corresponds to the value of E _c (θ) 0~10V;

Step 6: Find the average U=32768 of the digital signal;

Step 7: Subtract U from the digital sine signal E _s ^* (θ) to obtain the sine signal input Y, and the range of Y is -32768 to 32767, and subtract U from the digital cosine signal E _c ^* (θ) to obtain the cosine signal input Quantity X, the range of X is -32768～32767;

Step 8: Determine the quadrant of the rotor position of the motor according to the sine signal input value Y and the cosine signal input value X:

When X≥0 and Y≥0, the motor rotor position θ is in the first quadrant;

When X<0 and Y≥0, the motor rotor position θ is in the second quadrant;

When X<0 and Y<0, the motor rotor position θ is in the third quadrant;

When X≥0 and Y<0, the motor rotor position θ is in the fourth quadrant;

取0～1023的整数代表角度0～360度，然后根据上一步骤中θ所在的象限确定电机转子位置θ，θ取0～1023的整数代表角度0～360度：Step 9: Map X and Y to the first quadrant, and use the CORDIC algorithm to calculate the angle value

Take an integer from 0 to 1023 to represent an angle of 0 to 360 degrees, and then determine the motor rotor position θ according to the quadrant where θ is located in the previous step, and take an integer from 0 to 1023 to represent an angle of 0 to 360 degrees:

When θ is in the first quadrant, the motor rotor position

When θ is in the second quadrant, the motor rotor position

When θ is in the third quadrant, the motor rotor position

When θ is in the fourth quadrant, the motor rotor position

When the resolver output signal fails, the specific steps to detect the rotor position of the brushless motor are as follows:

Step a: Calculate the rotation angle of the motor in the two sampling intervals Δθ=ω _r ΔT, where ω _r is the motor speed, the unit is rad/s, the sampling frequency of analog-to-digital conversion is f=100kHz, and the time interval between two sampling ΔT=1/f=10us;

Step b: _{Use the CORDIC inversion algorithm to obtain the sine value E sG *} ⁽ θ _G ) and the cosine value E _cG ^* (θ _G ) of θ G = θ+Δθ, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ and ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ Not all exceed the error limit, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ If the error limit is exceeded, it means that the cosine signal output by the resolver is faulty; the error limit is set to 300, and steps 4 to 5 are performed;

Step c: If ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ and ${\mathrm{\&epsiv;}}_{c1}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ All exceed the error limit, use the CORDIC inversion algorithm to get the sine value E _sG ^* (θ′ _G ) and the cosine value E _cG ^* (θ′ _G ) of θ′ _G = θ-Δθ, if ${\mathrm{\&epsiv;}}_{\mathrm{the\; s}1}^{\&prime;}={\mathrm{E.}}_{\mathrm{sG}}^{*}\left({\mathrm{\&theta;}}_G^{\&prime;}\right)-{\mathrm{E.}}_{\mathrm{the\; s}}^{*}\left(\mathrm{\&theta;}\right)$ If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if ${\mathrm{\&epsiv;}}_{c1}^{\&prime;}={\mathrm{E.}}_{\mathrm{cG}}^{*}\left({\mathrm{\&theta;}}_G^{\&prime;\&prime;}\right)-{\mathrm{E.}}_c^{*}\left(\mathrm{\&theta;}\right)$ If it exceeds the error limit, it means that the cosine output by the resolver is faulty; the error limit is 300;

Step d: If the sine signal output by the resolver fails, convert the digital cosine signal E _c ^* (θ) in step 5 to obtain E _s ^* by using the orthogonal relationship between E _c ^* (θ) and E _s ^* (θ) ^* (θ), and replace the digital sine signal E _s ^* (θ) with E _s ^** (θ); if the cosine signal output by the resolver fails, the digital sine signal E _s ^* (θ) in step 5 is used The orthogonal relationship between E _s ^* (θ) and E _c ^* (θ) is converted to E _c ^** (θ), and E _c ^** (θ) is used to replace the digital cosine signal E _c ^* (θ);

Step e: Continue steps 6-9 to obtain the motor rotor position θ.

Claims (3)

1. A method for detecting the rotor position of a brushless motor, characterized in that the rotor coaxial installation of the rotary transformer and the brushless motor under test, the concrete steps of detecting the rotor position of the brushless motor are as follows: Step 1: Convert the two quadrature signals output by the secondary side of the resolver containing the information of the rotor position angle θ to differential to single-ended to obtain a single-ended sinusoidal signal E _s (t, θ) = Esin(ωt)sin(θ) and single-ended cosine signal E _c (t, θ)=Esin(ωt)cos(θ); the two-way signals containing rotor position angle θ information and orthogonal signals are: high-frequency sinusoidal differential signal E _s+ (t, θ) and E _s- (t, θ), high-frequency cosine differential signals E _c+ (t, θ) and E _c- (t, θ); where: E is the effective value of the signal, ω is the signal generated by the sine excitation circuit the angular frequency of the reference signal Refs; Step 2: Convert the sinusoidal differential signals Refs+ and Refs- generated by the sinusoidal excitation circuit input to the resolver into a single-ended sinusoidal reference signal -Refs=Esin(ωt); Step 3: superimpose the single-ended sine signal E _s (t, θ) and the reference signal Refs to obtain the signal Refs+E _s (t, θ), and superimpose the single-ended cosine signal E _c (t, θ) and Refs to obtain the signal Refs+E _c (t, θ); Step 4: Filter out the high-frequency part in Refs+E _s (t, θ) and Refs+E _c (t, θ) to obtain an analog low-frequency sinusoidal signal E _s (θ)=Esin( θ)+E and low frequency cosine signal E _c (θ)=Ecos(θ)+E; Step 5: Convert the analog low-frequency sinusoidal signal E _s (θ) = Esin(θ) + E and cosine signal E _c (θ) = Ecos(θ) + E into digital sine signal E _s ^* (θ) and digital cosine signal E _c ^* (θ); Step 6: Calculate the average U of the digital signals E _s ^* (θ) and E _c ^* (θ); Step 7: Subtract U from the digital sine signal E _s ^* (θ) to obtain the input quantity Y of the sine signal, and subtract U from the digital cosine signal E _c ^* (θ) to obtain the input quantity X of the cosine signal; Step 8: Determine the quadrant of the rotor position of the motor according to the sine signal input value Y and the cosine signal input value X: When X≥0 and Y≥0, the rotor position θ of the motor falls within the range of 0° to 90°; When X<0 and Y≥0, the motor rotor position θ falls within the range of 90°～180°; When X<0 and Y<0, the motor rotor position θ falls within the range of 180°～270°; When X≥0 and Y<0, the motor rotor position θ falls within the range of 270°～360°; Step 9: Map X and Y to the first quadrant, and use the CORDIC algorithm to calculate the angle value

Then determine the motor rotor position θ according to the quadrant where θ is located in the previous step: When 0<θ≤90°, the motor rotor position

When 90°<θ≤180°, the motor rotor position When 180°<θ≤270°, the motor rotor position

When 270°<θ≤360°, the motor rotor position 2. The method for detecting the rotor position of a brushless motor according to claim 1, wherein the specific steps for detecting the rotor position of a brushless motor are as follows: Step a: Calculate the rotation angle of the motor in the two sampling intervals Δθ=ω _r ΔT, where ω _r is the motor speed, the unit is rad/s, the sampling frequency of analog-to-digital conversion is f, and the time interval between two samplings is ΔT = 1/f; Step b: _{Use the CORDIC inversion algorithm to obtain the sine value E sG *} ⁽ θ _G ) and the cosine value E _cG ^* (θ _G ) of θ G = θ+Δθ, if

and

Not all exceed the error limit, if If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if If the error limit is exceeded, it means that the cosine signal output by the resolver is faulty; the error limit is taken as the average value U of the digital signal

Execute steps 4 to 5; Step c: If

and

All exceed the error limit, use the CORDIC inversion algorithm to get the sine value E _sG ^* (θ′ _G ) and the cosine value E _cG ^* (θ′ _G ) of θ′ _G = θ-Δθ, if

If the error limit is exceeded, the sinusoidal signal output by the resolver is faulty; if If it exceeds the error limit, it means that the cosine output of the resolver is faulty; the error limit is taken as the average value U of the digital signal

Step d: If the sine signal output by the resolver fails, convert the digital cosine signal E _c ^* (θ) in step 5 to obtain E _s ^* by using the orthogonal relationship between E _c ^* (θ) and E _s ^* (θ) ^* (θ), and replace the digital sine signal E _s ^* (θ) with E _s ^** (θ); if the cosine signal output by the resolver fails, the digital sine signal E _s ^* (θ) in step 5 is used The orthogonal relationship between E _s ^* (θ) and E _c ^* (θ) is converted to E _c ^** (θ), and E _c ^** (θ) is used to replace the digital cosine signal E _c ^* (θ); Step e: Continue steps 6-9 to obtain the motor rotor position θ. 3. A device for realizing any method for detecting the rotor position of a brushless motor according to claims 1 and 2, characterized in that it includes a rotary transformer, a sinusoidal excitation circuit, a differential signal converter, a precision absolute value adder, an inverter Phase device and low-pass filter; the resolver is installed coaxially with the rotor of the brushless motor under test, and the high-frequency sinusoidal differential signals Refs+ and Refs- generated by the sinusoidal excitation circuit are input to the primary side of the resolver and output from the secondary side of the resolver E _s+ (t, θ) and E _s- (t, θ), E _c+ (t, θ) and E _c- (t, θ) containing the motor rotor position angle θ are respectively obtained by two differential signal converters Single-ended sine signal E _s (t, θ) and single-ended cosine signal E _c (t, θ); on the other hand, the high-frequency sinusoidal differential signals Refs+ and Refs- output by the sinusoidal excitation circuit are transformed into single-ended positive The string reference signal-Refs is input to the inverter, and the Refs output by the inverter and the output signals of the two differential signal converters are respectively input to the respective precision absolute value adders, and the output signal Refs+ is obtained by superposition of the precision absolute value adders E _s (t, θ) and Refs+E _c (t, θ); the output signals of the two precision absolute value adders are input to their respective low-pass filters, and the filtered output of the low-pass filters is related to the rotor position Low-frequency sine signal E _s (θ) = Esin (θ) + E and cosine signal E _c (θ) = Ecos (θ) + E; the output of the low-pass filter is converted into a digital signal by the A/D module and then input to the DSP .

Images