CN110912480A - Permanent magnet synchronous motor model-free predictive control method based on extended state observer - Google Patents

Abstract

The invention relates to a model-free predictive control method of a permanent magnet synchronous motor based on an extended state observer, which comprises the following steps: a, estimating unknown and disturbance parts of a system by using a linear extended observer, wherein motor parameters are not involved; and B: and passing the reference current vector and the feedback current vector of the motor through a permanent magnet synchronous motor model-free controller based on an extended state observer, obtaining a reference voltage vector by using a complex vector current regulator, and obtaining six switching signals of the inverter after PWM modulation, thereby realizing the control of the motor. The invention solves the problems of motor noise, steady-state torque, large current ripple, current static error and poor system control performance when motor parameters are inaccurate in the traditional model predictive control scheme.

Description

Permanent magnet synchronous motor model-free predictive control method based on extended state observer

Technical Field

The invention relates to a permanent magnet synchronous motor model-free prediction control method based on an extended state observer.

Background

Model Predictive Control (MPC) has emerged in process control in the late industrial field of the 20 th century, 70 s, and has been widely used in early industrial application industries, such as chemical industry. In the 90 s of the 20 th century, Holtz, a German scholarer, originally applied model predictive control to the field of power electronic transmission. In recent years, along with the improvement of the performance of a digital processor and the reduction of the cost thereof, the large calculation amount is no longer a barrier for limiting the development of the MPC, and the MPC becomes a hot point of research due to the advantages of simple principle, high response speed, easy processing of multi-constraint multivariable problems and the like. MPC has become an important branch of control methodology in the field of power electronics.

MPC, however, relies on controlled object mathematical models and parameter accuracy. In practical application, the variation of the operating environment and the operating condition of the motor can cause the variation of the motor parameters, and the inaccuracy of the parameter measurement can affect the performance of the control algorithm. The noise in the running process of the motor and the static error of the current are even diffused, and the control performance and the stability of the system are influenced.

To solve this problem, some researchers have proposed a model-free predictive control (MFPC) method based on a super-local model, but this method relies on a high sampling rate, and the effect is not ideal when the sampling rate is low, so this method has a high requirement on the performance of the controller.

Disclosure of Invention

The invention aims to solve the problems of motor noise, large steady-state torque, large current ripple, current static error and poor system control performance when motor parameters are inaccurate in the traditional model predictive control scheme. The invention provides a permanent magnet synchronous motor model-free predictive control method (LESO-MFPC) based on a linear extended state observer.

The technical scheme adopted by the invention is as follows:

a permanent magnet synchronous motor model-free predictive control method based on an extended state observer comprises the following steps:

step 1: the whole system adopts a series control structure and a q-axis current instruction

Obtained by a speed outer ring PI regulator;

step 2: according to the rotor position information theta, the q-axis current instruction obtained in the step 1

D-axis current instruction given according to requirements

Coordinate transformation is carried out to obtain a current instruction under a static coordinate system

And step 3: obtaining an instruction voltage vector by utilizing real-time sampling current and given current information in the step 2 according to a model-free algorithm based on a linear extended state observer

And 4, step 4: according to the voltage vector obtained in the step 3

And constructing a driving signal of each switching tube of the inverter by using space vector modulation (SVPWM).

The invention has the following characteristics and advantages:

(1) a model-free control algorithm based on an extended state observer is used, motor parameters are not used, and robustness is strong;

(2) when the parameters are accurate and inaccurate, the dynamic performance and the steady-state performance are good;

(3) the requirement on the sampling frequency is not high, a good control effect can be realized under the lower sampling frequency, and the problem that the traditional model-free control method is limited by a hardware platform is solved.

Drawings

FIG. 1 is a hardware structure diagram of a permanent magnet motor speed regulation control system;

FIG. 2 is a structural block diagram of model-free predictive control of a permanent magnet synchronous motor based on an extended state observer;

FIG. 3 shows the experimental results of the rated load when the permanent magnet motor is operated at 1500rpm with accurate parameters under the dead-beat prediction control of the permanent magnet motor at a sampling rate of 10 kHz;

FIG. 4 is an experimental result of a permanent magnet motor with a rated load when the motor operates at 1500rpm when parameters are accurate and the model-free predictive control of the permanent magnet motor based on a super local model is under a sampling rate of 10 kHz;

FIG. 5 shows the experimental results of the model-free predictive control of the permanent magnet motor based on the extended state observer at a sampling rate of 10kHz and with a rated load when the motor is operated at 1500r/min when the parameters are accurate.

FIG. 6 is an experimental result of model-free predictive control of a super-local model-based permanent magnet motor with inaccurate parameters at a sampling rate of 10 kHz;

FIG. 7 shows the experimental results of the parameter inaccuracy at the sampling rate of 10kHz by the model-free predictive control of the PMSM based on the extended state observer.

Detailed Description

The following examples are presented to enable those skilled in the art to more fully understand the present invention and are not intended to limit the invention in any way.

For better understanding of the objects, technical solutions and advantages of the present invention, the present invention will be described in detail with reference to the accompanying drawings in conjunction with the following embodiments, wherein the d-axis current command is generated in step 2

The examples are provided for the purpose of illustration only and are not intended to limit the scope of the invention:

fig. 1 is a hardware circuit structure diagram of the present invention, which includes a three-phase voltage source, a three-phase diode rectifier bridge, a dc-side capacitor, a permanent magnet motor, a voltage and current sampling circuit, a DSP (Digital Signal Processing) controller, and a driving circuit. The voltage and current sampling circuit respectively collects direct current side voltage and phase current of the permanent magnet motor a and b by using the voltage Hall sensor and the current Hall sensor, and sampling signals enter the DSP controller after passing through the signal conditioning circuit and are converted into digital signals. The DSP controller completes the operation of the method provided by the invention, outputs six paths of switching pulses, and then obtains final driving signals of six switching tubes of the inverter after passing through the driving circuit.

Fig. 2 is a control schematic block diagram of the present invention, and the control method is implemented on the DSP controller of fig. 1 in sequence according to the following steps:

step 1: according to a q-axis current instruction obtained by an outer ring rotating speed PI regulator

Is particularly shown as

(k_pAnd k_iProportional gain and integral gain in PI regulator, s is complex parameter in pull-type conversion, omega^*And ω is the rotational speed command and the actual rotational speed, respectively).

Step 2: according to the q-axis current instruction obtained in the step 1

And rotor position information theta obtained by the photoelectric encoder

Converting the abc- αβ coordinates to obtain a current command in a static coordinate system

And step 3: the model-free controller based on the permanent magnet synchronous motor complex vector super local model comprises the following components:

where F is the quantity containing system configuration information including parts unknown to the system and possible interference.

Discretization can obtain

The linear extended state observer was introduced as follows:

the observer discrete model is as follows:

in the formula, β₀₁＝T_scβ₁，β₀₂＝T_scβ₂Gain coefficients adjusted for the observer.

Superscripts k and k +1 denote the values at times k and k +1, respectively, subscript s denotes the stator variable, ^ denotes the estimated value, and z denotes the state variable. T is_scDenotes the sampling period, u denotes the voltage, i denotes the current, e_rrTo estimate the error.

Obtaining an estimated value of the next moment F by the extended state observer, replacing the current of the next moment with the current instruction obtained in the step 2, and calculating according to the super-local discrete model to obtain a voltage instruction of the next moment

And 4, step 4: according to the voltage vector obtained in the step 3

And constructing a driving signal of each switching tube of the inverter by using space vector modulation (SVPWM).

The effectiveness of the proposed method can be obtained by comparing the experimental results shown in fig. 4 and 5 with those shown in fig. 6 and 7. Fig. 4 is an experimental result of the model-free predictive control of the permanent magnet motor based on the super-local model at a sampling rate of 10kHz with accurate parameters and the motor running at 1500rpm with a rated load, and fig. 5 is an experimental result of the method of the present invention at a sampling rate of 10kHz under the same conditions. In fig. 4 and 5, waveforms sequentially include rotation speed, electromagnetic torque, stator flux linkage amplitude and motor stator terminal a-phase current from top to bottom. From the comparison of fig. 3, fig. 4 and fig. 5, it can be found that, under the condition of accurate parameters, the method of the present invention has ten steady-state effects when the sampling rate is equal to that of the conventional schemeAre divided into two parts. FIGS. 6 and 7 show the inaccuracy of the parameters of the motor at a sampling rate of 10kHz (1R,2 psi)_f2L), fig. 6 corresponds to the experimental result of model-free predictive control of the permanent magnet motor based on the super local model, and fig. 7 corresponds to the experimental result of the method of the present invention. It can be seen from fig. 6 and 7 that in a dynamic process, the method described in the present invention enables good tracking for a given current, whereas the conventional scheme presents current biasing; compared with the traditional scheme, the method has the advantages that the current is smoother while the method has similar rapid dynamic performance.

Those skilled in the art will appreciate that the above embodiments are merely exemplary embodiments and that various changes, substitutions, and alterations can be made without departing from the spirit and scope of the application.

Claims (3)

1. A permanent magnet synchronous motor model-free predictive control method based on an extended state observer is characterized by comprising the following steps:

a, estimating unknown and disturbance parts of a system by using a linear extended observer, wherein motor parameters are not involved;

and B: and passing the reference current vector and the feedback current vector of the motor through a permanent magnet synchronous motor model-free controller based on an extended state observer, obtaining a reference voltage vector by using a complex vector current regulator, and obtaining six switching signals of the inverter after PWM modulation, thereby realizing the control of the motor.

2. The method of claim 1, wherein step a comprises:

the linear extended state observer was introduced as follows:

the observer discrete model is as follows:

in the formula, β₀₁＝T_scβ₁，β₀₂＝T_scβ₂A gain factor adjusted for an observer;

obtaining an estimated value of the next moment F by the extended state observer, replacing the current of the next moment with the current instruction obtained in the step 2, and calculating according to the super-local discrete model to obtain a voltage instruction of the next moment

3. The method of claim 1, wherein step B comprises:

the model-free controller based on the permanent magnet synchronous motor complex vector super-local model is established as follows:

where F is the quantity containing system structure information, including parts unknown to the system and possible interference;

discretization can obtain

Based on estimated unknown part of the system

And sampling the current

Given current

To obtain a given voltage vector of

And then, carrying out PWM modulation to obtain six switching signals of the inverter, thereby realizing the control of the motor.

Images