RU2532673C2 - Method and device of digital processing of signals from pulse sensor of rotor movement in electric motor-encoder - Google Patents

Abstract

FIELD: engines and pumps.

SUBSTANCE: group of inventions relates to the method and device of digital processing of signals from a pulse sensor of rotor movement in an electric motor-encoder, which may be used in an electric drive, in particular, a thrust electric drive of vehicles of different type and purpose. To process information about angular position of the rotor, an observer is used to observe mechanical variables of electric drive condition and a predictor, for functioning of which they previously before start of electric drive operation they set initial values of angular position of the electric motor rotor equal to zero, as well as values of angular speed of rotor rotation and load moment on the electric motor shaft, and an initial correcting signal of the predictor. During each cycle of calculations they produce a measured value of the angular position of the rotor, calculate the preliminary value of the correcting signal of the predictor. Further they produce the value of the correcting signal of the predictor and determine the value of the angular position of the rotor predicted for the start of the next cycle of calculations.

EFFECT: increased accuracy of assessment of angular position and angular speed of electric motor rotor rotation, which makes it possible to considerably improve characteristics of a drive control system.

3 cl, 2 dwg

Description

Technical field

The group of inventions relates to a method and apparatus for digitally processing signals from a pulse sensor for displacing a rotor of an electric motor (encoder), which can be used in an electric drive, in particular in a traction electric drive of vehicles of various types and purposes.

State of the art

Pulse motion sensors (encoders) are widely used in various industries, primarily in automated electric drives. Their advantages include relative simplicity and low cost, stability characteristics, noise immunity and high reliability. The output signal of such sensors is usually transmitted in the form of pulses through two channels (the channel of the main and inverse signals) to protect against interference in the transmission lines. In this case, the signals transmitted through the direct and inverse channels are 90 ° phase-shifted relative to each other, which allows one to determine not only the displacement of the rotor by one discrete, but also the direction of rotation of the rotor. The minimum measurable amount of displacement - one discrete sensor - is determined by the number of pulses per mechanical revolution. The measurement resolution usually ranges from several tens to tens of thousands of discrete per mechanical revolution (for different types of encoders), taking into account the determination of movement along the arrival edges of pulses (which, taking into account two pulse channels, allows quadrupling the number of measured discrete displacements per revolution).

The main problems of using encoders in an automated electric drive are connected precisely with the discreteness of measuring the movement of the rotor. When adjusting the rotation speed from zero to the maximum value, the traditional methods of processing the signals of pulse sensors are measuring the time between pulses arriving, or measuring the number of pulses in a certain time, or averaging (filtering) pulses of a standard area formed by the front of the sensor pulses arrival, as well as a combination of these methods lead to unsatisfactory operation of the drive. This is due, first of all, to an unacceptably large delay in obtaining an estimate of the rotational speed in the region of low rotational speeds (positioning mode). The problem is exacerbated in drives such as traction, in which a wide operating temperature range, a high level of vibration and interference do not allow the use of high-resolution pulsed sensors (for example, photopulse). When using electromagnetic sensors that satisfy the operating conditions in traction drives, the number of pulses per mechanical revolution usually does not exceed several hundreds, and taking into account the multipolarity of such machines, the number of pulses per electric revolution does not exceed several tens.

The prior art method for determining errors in the signals of the position of the rotor using the algorithm for determining the position of the rotor, compensating for errors in the system (see EP 1624563 A3, IPC H02P 25/08, published. 08.02.2006), which is adopted as a prototype.

The above method consists in the fact that compensating signals can be recorded in the control system and used to improve the accuracy of the output signals of the rotor position sensor, improving the output parameters of the electric machine. The correction algorithm can use the prediction / correction model, algorithms for pulse diagnostics, current generation, determination of inductance parameters, etc.

However, the method known from the prior art does not allow solving the above problems, such as a delay in obtaining an estimate of the rotational speed in the region of low rotational speeds of the motor shaft, which in turn leads to unsatisfactory operation of the electric drive as a whole.

Thus, the objective of this group of inventions is aimed at improving the characteristics of the drive control system.

Disclosure of invention

First of all, it should be noted that the pulse displacement sensor (encoder) measures the displacement by one discrete. It is not difficult to calculate the total number of movements, that is, with the help of a pulse sensor, the movement of the rotor is actually measured. Typically, the movement of the rotor is measured accurate to the initial value of the position of the rotor, which for some drives, for example with an induction motor, does not matter. However, individual sensors have a so-called “zero mark” of revolution, which allows measuring the absolute movement of the rotor. Thus, the pulse sensor is a displacement sensor, wherein displacement is measured discretely in level.

The primary signal processing of the pulse sensor is carried out in the hardware, usually built into a specialized drive controller. Such hardware contains a reversible pulse counter (taking into account the direction of rotation), as well as a protection circuit against interference of the same polarity in the communication line in direct and inverse sensor signals and a protection circuit from the “bounce” of signals arising from the operation of pulse sensor comparators. The pulse counter code at the beginning of the calculation cycle is entered into the control processor.

The technical result of the claimed group of inventions is to ensure correct measurement of the position of the rotor at the time of the start of the current cycle of calculation of the drive control controller and to improve the accuracy of estimating the angular position and angular speed of rotation of the rotor of the electric motor, which can significantly improve the characteristics of the drive control system as a whole, increase the reliability of control, eliminate dangerous transients, improve drive efficiency. These advantages are especially important for traction electric drives.

The claimed technical result is achieved through the use of a digital signal processing method for a pulse displacement sensor of the rotor of an electric motor-encoder using a processor controller that provides the control system of the electric drive with the values of its mechanical variables during the k-th cycle of controller calculations, such as the angular position of the electric motor rotor Θ _k , angular rotor speed n _k and the load moment on the motor shaft M _Lk , and designed to obtain accurate estimates these variables, which consists in the fact that for processing information about the angular position of the rotor, an observer of mechanical variables of the state of the electric drive and a predictor are used, for the operation of which, prior to the operation of the electric drive, the initial values of the estimates of the angular position of the rotor of the electric motor Θ ₀ and the angular velocity of rotation of the rotor p ₀ and the moment of load on the motor shaft M _L0 , and the initial predictor correction signal S ₀ , corresponding to the disconnected state of the electric drive, set the discrete values of the sensor Θd, depending on the parameters of the encoder gear, and enter the duration of the calculation cycle of the controller 7, the cumulative moment of inertia of the motor rotor J and the values of the correction factors L ₁ , L ₂ , L ₃ at the beginning of each calculation cycle of electric drive control system observer mechanical state variables included in the controller receives the desired value of the electromagnetic moment M _k, of the encoder, after primary processing its signal obtained number of pulses in the current calculation interval i, using the timer is performed to measure and record in time the controller buffer t _k from the arrival of the last pulse before the beginning of the calculation cycle, in addition, the buffer stores the measured last cycle calculations rotor angular position value of the motor Θ _{k -1izm} obtained in the last calculation cycle correction signal value predictor S _k-1 and the value of Θ _k estimates the rotor angular position, angular rotor speed and torque n _k elec load rodvigatelya M _Lk, which in this cycle are used as its current value for each computation cycle obtained measured value of the angular position of the rotor from the expression Θ _kizm = i · Θd, calculated provisional value of the correction signal predictor S _'k from the expression S' _k, = t _k · n _k if Θ _kism ≠ Θ _{k-1 ism} , otherwise, from the expression S ' _k = S _k-1 + T · n _k we get the value of the predictor correction signal S _k equal to Θd if S' _k > Θd, and otherwise equal to S ' _k , determine the predicted at the beginning of the next cycle of calculations angular rotor position value from the expression Θ _k Θ * _k = Θk _edited + S _k, and then to the watcher calculated value of angular position of the rotor estimates the angular rotor speed and load torque applied to determine the parameters of electric drive control system according to the expressions:

, $${\Theta }_{k+\mathrm{one}}={\Theta }_k+T{n}_k+\frac{{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="00000016.png"\; state="\{\{state\}\}">T</figure-callout>}}^2}{2J}\left(M_k-M_{Lk}\right)+L_{\mathrm{one}}\left({\Theta }_k-{\Theta }_k^{*}\right)$$

,

, $$n_{k+\mathrm{one}}=n_k+\frac{T}{J}\left(M_k-M_{Lk}\right)+L_2\left({\Theta }_k-{\Theta }_k^{*}\right)$$

,

, $$M_{Lk+\mathrm{one}}=M_{Lk}+L_3\left({\Theta }_k-{\Theta }_k^{*}\right)$$

,

which are written to the controller buffer and in the next cycle of calculations are used to ensure the operation of the observer of mechanical state variables.

To implement the inventive method, in particular, a device for digitally processing signals from a pulse encoder of a rotor of an electric motor-encoder using a processor controller can be used, implementing the method according to claim 1, including a unit for receiving and primary processing of encoder signals, a unit for multiplying the number of encoder pulses by an amount discretes of the sensor, a predictor, which is a set of timer for measuring time from the last pulse in the cycle to the end of the calculation cycle, the preliminary calculation block the actual value of the corrective signal of the predictor, the restriction unit, designed to determine the value of the corrective signal of the predictor and the adder of the values of the measured angular position of the rotor and the corrective signal of the predictor, as well as a processor controller configured to include at least either an observer of the mechanical variables of the state of the electric drive and a controller buffer or at least an observer of the mechanical variables of the state of the electric drive, the controller buffer, and predictably p, in this case, the connections between the observer, the controller buffer and the predictor will be identical in both cases, with the encoder output connected to the input of the encoder signal reception and primary processing unit, the outputs of which are connected to the inputs of the multiplication unit and timer, the timer output is connected to the buffer input controller, the output of which the sensor discrete value is transmitted to the input of the multiplication unit, the output of which is connected to the first input of the adder corresponding to the measured value of the angular position of the rotor, as well as to the inputs of for calculating the preliminary value of the predictor correction signal and the controller buffer, which also has inputs for recording the values of the predictor correction signal calculated in the previous cycle of calculations, estimates of variables obtained in the observer of the mechanical variables of the state of the drive, an input for external programming and outputs, one of which is connected with the input of the block for calculating the preliminary value of the predictor correction signal, the second with one of the inputs of the restriction block, the output of the block the calculation of the preliminary value of the predictor correction signal is connected to another input of the restriction block, the output of which is connected to the second input of the adder corresponding to the predictor correction signal, and the adder output is connected to the first input of the observer of the mechanical variables of the state of the electric drive, to the second input of which the required value is supplied from the electric drive control system electromagnetic moment, and the outputs of which are received for transmission to the control system of the electric drive and for recording to the controller buffer, the values of the estimates of the angular position of the rotor, the angular speed of rotation of the rotor and the load moment, which are used in the next cycle of calculations to ensure the work of the observer of mechanical state variables.

In one embodiment of the device, the predictor and processor controller form a modular design when the predictor is not part of the controller.

Brief Description of the Drawings

A block diagram of a device in which the processor controller includes a controller buffer and an observer of the mechanical variables of the state of the electric drive is shown in FIG.

A block diagram of a device in which a processor controller includes a predictor, a controller buffer, and an observer of mechanical variables of the state of the electric drive is shown in FIG.

The implementation of the invention

A method for digitally processing signals of a pulsed sensor for displacing a rotor of an electric motor (encoder) is illustrated in FIG. 1, which shows a device that implements the method. In this case, an example of the use of an encoder in an electric drive, in particular in a traction electric drive of vehicles of various types and purposes, is given.

The device depicted in figure 1, operates as follows. The electric energy source 1 generates energy that is transmitted by the control system of the traction electric drive 2 to the electric motor 3 of the drive of the propulsion devices (wheels, tracks, etc.) of the vehicle. The control system 2 in the General case includes an adjustable power converter and a controller of the control system (pas 1 and 2 are not indicated by positions), providing the required parameters of electric power at the output of the power converter.

The control system 2 receives the values of the estimates of the variables of the mechanical state of the electric drive (the angular position of the rotor of the electric motor углов _{k the} angular speed of rotation of the rotor n _k and the moment of load on the shaft of the electric motor M _Lk ) from the processor controller 4 having a given calculation cycle duration. For convenience of explanation, we denote the cycle number of the controller calculations by the index k.

The processor controller 4 includes a buffer of the controller 5, which provides recording, storage and output of the values of various constants, variables and coefficients necessary for the operation of the controller 4.

The buffer 5 has a number of inputs and outputs, in particular an input for external programming. This input is used to enter into the processor controller 4 before calculating the duration of the calculation cycle T, the total moment of inertia of the rotor of the electric motor J, the displacement sensor discrete Θd (the angular value corresponding to the rotation of the gear by one division), the correction factors L ₁ , L ₂ , L ₃ , setting the rate of convergence of estimates of the mechanical variables of the rotor and / or other greatness. Alternatively, the parameters J, L ₁ , L ₂ , L ₃ can be calculated inside the processor controller 4 according to known relations (given below) and written to the buffer 5.

The processor controller 4 is programmed in such a way that each time it is loaded (before the drive starts to work), initialization (zeroing) occurs in buffer 5 of the initial values of the angular position of the rotor Θ ₀ , the angular speed of rotation of the rotor n ₀ and the load moment on the motor shaft M _L0 , and also a predictor correction signal S ₀ .

In the processor controller 4, the observer 6 of the mechanical variables of the state of the electric drive is also implemented by software or hardware and software.

The functioning of the state observer is based on the use of a process model. Since the angular position of the rotor is measured by a pulse sensor, the following model of the mechanical movement of the rotor is taken as the base:

, $$\frac{d\Theta }{dt}=n$$

,

, $$\frac{dn}{dt}=\frac{\mathrm{one}}{J}(M-M_L)$$

,

where Θ is the angular position of the rotor, n is the angular velocity, M is the electromagnetic moment, M _L is the load moment, J is the total moment of inertia of the rotor. In this case, the electromagnetic moment M is known (measured). The value of the electromagnetic moment in this calculation cycle of the controller M _k comes from the output of the drive control system 2 to the input of the observer 6.

The magnitude of the load moment M _L is not known beforehand, however, for the traction drive model it is quasi-constant (the magnitude of the load moment can vary, but slowly compared to the rate of obtaining estimates of state variables):

. $$\frac{d{M}_t}{dt}=0$$

.

Since estimates of state variables are formed in the control processor controller 4, which operates in discrete time, it is necessary to use a difference model. The difference model of mechanical motion is constructed by analytical integration of the equations of mechanical processes on the time discretization interval. The values of the modeled variables in the state observer are corrected by measuring the angular position Θ * (when analyzing the processes of estimation by the discreteness of position measurements, the pulse sensor can be neglected). With a constant input action - the electromagnetic torque, the initial and final values of the estimates of state variables in the cycle are connected by a difference relation (equations of the state observer):

, $${\Theta }_{k+\mathrm{one}}={\Theta }_k+T{n}_k+\frac{{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="00000016.png"\; state="\{\{state\}\}">T</figure-callout>}}^2}{2J}(M_k-M_{Lk})+L_{\mathrm{one}}({\Theta }_k-{\Theta }_k^{*})$$

,

, $$n_{k+\mathrm{one}}=n_k+\frac{T}{J}(M_k-M_{Lk})+L_2({\Theta }_k-{\Theta }_k^{*})$$

,

, $$M_{Lk+\mathrm{one}}=M_{Lk}+L_3({\Theta }_k-{\Theta }_k^{*})$$

,

where T is the duration of the calculation cycle, ak is the cycle number. Moreover, k + 1 indicates that the next cycle of calculations is being performed. Accordingly, as can be seen from the above expressions, in the k + 1 cycle of calculations, the values of estimates of the variables of the mechanical state of the electric drive (Θ _k , n _k , M _Lk , etc.), which are defined in the previous cycle of calculations, are used. The values of the correction factors L ₁ , L ₂ , L _{3 are} determined by the characteristic equation obtained for the rate of change of the estimation error (mode):

. $$-{\mathrm{<figure-callout\; id="3"\; label="\lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\lambda </figure-callout>}}^3+{\lambda }^2(3+L_{\mathrm{one}})-\lambda \left(3+2L_{\mathrm{one}}+\frac{L_3{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="00000016.png"\; state="\{\{state\}\}">T</figure-callout>}}^2}{2J}-L_{2}T\right)+\left(\mathrm{one}+L_{\mathrm{one}}-L_{2}T-\frac{L_3{\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="00000016.png"\; state="\{\{state\}\}">T</figure-callout>}}^2}{2J}\right)=0$$

.

For example, if the desired motion is characterized by mode values λ ₀ = λ ₁ = λ ₂ = λ ₃ , i.e. the characteristic equation should have the form (λ ₀ -λ) ³ = 0, then the required values of the correction factors are calculated by the formulas:

L ₁ = 3 · (λ ₀ -1) ³ ;

; $${\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="00000016.png"\; state="\{\{state\}\}">L</figure-callout>}}_2=(9\cdot {\lambda }_0-3<figure-callout\; id="0"\; label="\cdot \; \lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\cdot \; </figure-callout>{\lambda }_{}^{}{}_0^2-{\mathrm{<figure-callout\; id="0"\; label="\lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\lambda </figure-callout>}}_0^3-5)/(2\cdot T)$$

;

. $${\mathrm{<figure-callout\; id="3"\; label="L"\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">L</figure-callout>}}_3=(-3<figure-callout\; id="0"\; label="\cdot \; \lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\cdot \; </figure-callout>{\lambda }_{}{}_0+3\cdot {\mathrm{<figure-callout\; id="0"\; label="\lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\lambda </figure-callout>}}_0^2-{\mathrm{<figure-callout\; id="0"\; label="\lambda "\; filenames="00000015.png,00000016.png"\; state="\{\{state\}\}">\lambda </figure-callout>}}_0^3+\mathrm{one})\cdot J/(T^2)$$

.

The value of λ ₀ must be specified (in the list of parameters). The constancy of the values of L ₁ , L ₂ , L ₃ (which is a consequence of the linearity of the equations of mechanical motion) allows you to set their values directly as parameters of the processor controller.

The choice of the roots of the characteristic equation λ ₁ , λ ₂ , λ ₃ determines the rate of convergence of the estimates of the angular position 0, the angular velocity n and the load moment M _L to their actual values in the drive. For λ ₀ = 0, finite-step convergence takes place (in this case, a maximum of three steps of computation in the processor); for -mod (λ) <1 (the roots can be chosen by complex conjugates), the convergence is asymptotic.

The observer is a system with two inputs: by the electromagnetic moment M (control) and by measuring the position of the rotor Θ. For λ ₀ ≠ 0, the input signal Θ will obviously be filtered, and the more so, the closer the root value to unity. However, filtering the signal of the displacement sensor does not mean dynamic delay in a closed control loop (for example, when controlling angular velocity). Indeed, the input signal M is not filtered by the observer, that is, the observer does not impose restrictions on the choice of regulation parameters, which is a consequence of the independence of the problems of regulation and estimation when using the observer.

The state observer’s equations contain a parameter - moment of inertia J. In this connection, it should be noted that passport data, experimental determination of the moment of inertia or automatic identification obtained under the assumption of constancy (quasi-static change) of the load moment:

$$J_{k+\mathrm{one}}=\frac{T\cdot (M_{k-\mathrm{one}}-M_{k-2})}{n_k-2\cdot n_{k-\mathrm{one}}+n_{k-2}}.$$

In the latter case, verification of the conditions of identifiability is necessary.

The observer of the state of variables of mechanical motion allows, at a relatively slow rate of convergence, to filter the measured signal of the displacement sensor quite well, including smoothing the signal discrete in level. However, there remains the problem of beats of the frequency of occurrence of pulses and the frequency of the calculation cycle: the number of pulses per cycle of the calculation varies from cycle to cycle of calculations even at a constant rotation speed. This problem is especially acute when using pulsed sensors with low resolution, with a relatively small number of pulses per revolution, which is typical, for example, for traction drives.

To eliminate beats caused by the asynchronous arrival of the encoder pulses relative to the calculation cycle in the processor controller 4, it is proposed to use the predicted values of the angular position as measurements. You can get the predicted values using the predictor, which allows you to adjust the measurement of the value of the position of the rotor at the beginning of the current cycle of calculations. The operation of the predictor is described in detail below.

To determine the values of the mechanical variables of the electric drive, a gear wheel 7 is installed on one shaft with an electric motor 3, next to which is a pulse encoder 8. The encoder, in particular, can be a pulse electromagnetic sensor with a quadrature output signal and one or two (direct and inverse) outputs, whose signals are 90 ° phase-shifted relative to each other, which makes it possible to determine not only the displacement of the rotor by one discrete, but also the direction of rotation rotor. The pulse frequency of the encoder 8 is directly proportional to the rate of change of the angle of rotation of the gear 7, i.e. ultimately the rotational speed of the rotor of the electric motor 3.

The pulses of the encoder 8 are received at the input of the unit 9 for receiving and primary processing of the encoder signals with resetting data to one full revolution of the gear wheel. Such a block can include a doubling circuit (for an encoder with two outputs - four) of pulses, which generates a short counting pulse on each edge of the input pulses of the encoder, as well as a reversible pulse counter (for an encoder with two outputs taking into account the direction of rotation), various protection schemes ( from interference of the same polarity, from “bounce” of signals, etc.), filtering of signals, etc. In any case, a signal is generated at one of the outputs of block 9, which is proportional to the number of pulses in the current calculation interval i.

The reversible pulse counter, which is included in block 9, is reset when the signal i is saved (transmitted for further calculations), proportional to the number of pulses in the current calculation interval, i.e. after the arrival of the last pulse in this cycle of calculations. In this case, the signal from the second output of block 9 is fed to the predictor 10, namely, to the input of the timer 11, which is reset at the beginning of each calculation cycle k (synchronized with the processor controller 4 in the case when it is not included in it). Thus, the timer 11 measures and writes to the controller buffer 5 the time t _k from the arrival of the last pulse of the previous calculation cycle to the beginning of the current cycle (the duration of operations in block 9 and timer 11 is negligible compared to the duration of the calculation cycle).

From the output of block 9, signal i is supplied to the first input of block 12 multiplying the number of encoder pulses by the sensor discrete, the second input of which from the buffer of controller 5 receives a signal proportional to Θd, and the signal of the measured value of the rotor angular position is generated at its output

_{Изм kism} = i · Θd.

The signal _{изм kism is} written to buffer 5 for use in the next controller calculation cycle. In addition, it enters the predictor 10, at the first input of the adder 13, and also at the input of the block 14 for calculating the preliminary value of the predictor correction signal S ' _k .

In block 14 from the buffer 5 also receives the values of T, Θ _{k-1 ism} , t _k , n _k , and S _k-1 , necessary to determine the preliminary value of the correction signal of the predictor.

An estimate of the angular velocity of rotation n _k in the current cycle is formed by the observer. Assume for now that in this estimate there are no variations associated with the discreteness of the sensor and beats of frequencies.

Then it is natural to assume that at the end of the cycle the true value of the angular position of the rotor Θ * _k will be equal to the sum of the readings of the pulse counter Θ _ISM , plus the product of the angular speed of rotation of the rotor n _k by the time from the arrival of the last pulse of the previous calculation cycle to the beginning of the current cycle t _k . If there were no pulses in this cycle, then the indicated time is equal to the cycle time t _k = T.

A sign of the appearance of a sensor pulse in the previous cycle is the inequality of the previous and current values of Θ _ISM in the calculation cycles (using this attribute is preferable to other signs).

Therefore, in block 14 of calculating the preliminary value of the predictor correction signal S ' _k , the validity of the condition is checked:

Θ _kism ≠ Θ _k-1ism .

If the condition is satisfied (in the previous cycle of calculations there was at least one sensor pulse), then

_{S 'k = t k · n} k,

because the shaft of the electric motor 3 should rotate by an angle t _k · n _k during the time t _k when the signal of the pulse counter of block 9, proportional to the number of pulses in the current calculation interval i, did not change, but the shaft rotated at a speed of n _k .

If the condition Θk _MOD ≠ Θ _k-1izm not satisfied (in the previous computation cycle was encoder pulses), then

S ' _k = S _k-1 + _Tnk ,

because during the cycle T, the angular position of the rotor should change by T · n _k , and the total change, taking into account the previous one, corresponds to the predictor correction signal in the previous calculation cycle S _k-1 , increased by T · n _k .

Finally, due to the inevitable errors in the calculations and measurements, it is natural to limit the value of the predictor correction signal to the value of the angular discrete of the sensor Θd (otherwise, a sensor pulse would appear in the interval (t _k , T)). This is done in restriction block 15, in which the condition is checked.

If the condition is satisfied (the value S ' _k obtained in block 14 exceeds the sensor discrete), then the value of the predictor correction signal S _k = Θd is set. If the condition S ' _k > Θd is not fulfilled, then the value of the predictor correction signal S _k = S' _{k is set} .

From the output of the restriction block 15, the predictor correction signal S _k goes to the second input of the adder 13, at the output of which the value of the rotor angular position Θ * _k predicted at the beginning of the next cycle of calculations is obtained from the expression Θ * _k = k _ISM + S _k .

In the value _k * there are no components related to the discreteness of measurements and beatings of frequencies, therefore, if we use the pulse counter readings thus adjusted to measure the position of the rotor at the end of the calculation cycle and use it as measurements in state observer 6, then the estimates the rotation speed n _k and the load moment M _Lk, these components will be absent. This justifies the assumption made earlier.

A signal proportional to the value of Θ _k * comes from the output of block 15 to the second input of the observer 6, in which the values of the estimates of the angular position, angular velocity of rotation of the rotor and the load moment (Θ _{k + 1} , n _k-1 and M _{Lk + 1} ) are calculated . These estimates from the output of the observer 6 go to the input of the control system 2, in which they are used to determine the control parameters of the electric drive during the next cycle of calculations. They are also recorded from the output of the observer 6 in the buffer of the controller 5 and in the next cycle of calculations are extracted from it to ensure the operation of the observer and are used for digital processing of encoder signals in accordance with the method described above.

Claims (3)

1. A method for digitally processing the signals of a pulse encoder of a rotor of an electric motor-encoder using a processor controller, which ensures that the drive control system receives the values of its mechanical variables during the k-th cycle of the controller’s calculations, such as the angular position of the electric motor rotor Θ _k , the angular rotational speed of the rotor n _k and the load moment on the motor shaft M _Lk , and intended to obtain accurate estimates of these variables, which consists in the fact that for processing information In order to determine the rotor angular position, an observer of the mechanical variables of the state of the electric drive and a predictor are used, for the operation of which, prior to the operation of the electric drive, the initial values of the estimates of the angular position of the rotor of the electric motor Θ ₀ , the angular speed of rotation of the rotor n ₀ and the moment of load on the motor shaft M _{L0 are} set to zero, and correcting the initial predictor signal S ₀ corresponding to the state of the disconnected drive, define discrete unit-size sensor Θd, dependent pairs meters gear encoder, and introduced into the buffer controller cycle time controller T computing total moment of inertia of the rotor of the motor J and the values of the correction factors L _1, L _2, L _3, at the beginning of each computation cycle of the drive control system of the observer mechanical state variable belongs into the controller, receives the required value of the electromagnetic moment M _k , from the encoder after the primary processing of its signal, the number of pulses for the current calculation interval i, s with the help of a timer, the time controller t _k is measured and written to the buffer from the arrival of the last pulse to the beginning of the current cycle of calculations, in addition, the value of the angular position of the motor rotor _{виг k-1ism} measured in the last cycle of calculations is saved in the buffer, and the correction value obtained in the last cycle of calculations the predictor signal S _k-1 , and the values of the estimates of the angular position of the rotor Θ _k , the angular velocity of rotation of the rotor n _k and the load moment of the electric motor M _Lk , which in this cycle are used as its current values, during each calculation cycle, the measured value of the angular position of the rotor is obtained from the expression _{изм kism} = i · Θd, the preliminary value of the predictor correction signal S'k is calculated from the expression S ' _k = t _k · n _k if Θ _kism ≠ Θ _{k- 1ism} , otherwise, from the expression S ' _k = S _k-1 + T · n _k , we obtain the value of the predictor correction signal S _k equal to Θd if S' _k > Θd, otherwise equal to S ' _k , determine the predicted at the beginning of the next cycle of calculations the value of the angular position of the rotor Θ * _k from the expression Θ * _k = Θ _kizm + S _k , and d Then, in the observer, the values of the estimates of the angular position of the rotor, the angular speed of rotation of the rotor and the load moment are used to determine the parameters of the drive control system according to the expressions:

,

,

,
which are written to the controller buffer and in the next cycle of calculations are used to ensure the operation of the observer of mechanical state variables. 2. A device for digitally processing signals of a pulsed encoder for moving a rotor of an electric motor-encoder using a processor controller, the method according to claim 1, including a unit for receiving and processing the encoder signals, a unit for multiplying the number of encoder pulses by the sensor discrete, a predictor representing a set of timer for measuring time from the last pulse in the cycle to the end of the calculation cycle, the block for calculating the preliminary value of the predictor correction signal, the ogre block the value intended to determine the value of the correction signal of the predictor and the adder of the values of the measured angular position of the rotor and the correction signal of the predictor, as well as a processor controller configured to include at least either an observer of mechanical variables of the state of the electric drive and a buffer of the controller, or at least an observer of mechanical variables of the state of the drive, the controller buffer and the predictor, while the connection between the observer, the controller buffer and the forecast in one case and the other will be identical, and the encoder output is connected to the input of the encoder signal reception and primary processing unit, the outputs of which are connected to the inputs of the multiplication unit and timer, the timer output is connected to the input of the controller buffer, from the output of which the sensor discrete value is transmitted to the input multiplication unit, the output of which is connected to the first input of the adder corresponding to the measured value of the angular position of the rotor, as well as to the inputs of the block for calculating the preliminary value of the correction signal prog the controller’s noser and buffer, which also has inputs for recording the values of the predictor correction signal calculated in the previous cycle of calculations in the restriction block, estimates of variables obtained in the observer of the mechanical variables of the drive’s state, an external programming input and outputs, one of which is connected to the input of the block of calculating the preliminary value of the predictor correction signal, the second one with one of the inputs of the restriction block, the output of the block of calculating the preliminary value of the correcting its predictor signal is connected to another input of the limiting unit, the output of which is connected to the second input of the adder corresponding to the correcting signal of the predictor, and the output of the adder is connected to the first input of the observer of the mechanical variables of the state of the electric drive, to the second input of which the required value of the electromagnetic moment is supplied from the electric drive control system, and the outputs of which are obtained for transmission to the control system of the electric drive and for writing to the controller buffer the values of the estimates of the angular rotor position, angular velocity of rotation of the rotor and the load moment, which in the next cycle of calculations are used to provide the observer with mechanical state variables. 3. The device according to claim 2, in which the predictor and processor controller form a modular design in the case when the predictor is not part of the controller.

Images