1. A method for constructing a control system of an intelligent valve electric actuator is characterized by comprising the following steps:

step 1, establishing a three-loop control system based on a current loop, a speed loop and a position loop;

step 2, adopting a variable gain coefficient MRAS to perform online identification on the rotational inertia of the system, taking the rotational inertia as the input of an observer to perform online observation on the load torque of the system, and setting the parameters of a system controller according to the observation result;

and 3, monitoring the change of the time constant of the rotor through a temperature detection circuit, and performing temperature compensation aiming at an error caused by temperature to the observation of the load torque of the control system according to the relation between the time constant of the rotor and the electromagnetic torque.

2. The method for constructing the control system of the electric actuator of the intelligent valve according to claim 1, wherein the step 1 of establishing the three-loop control system based on the current loop, the speed loop and the position loop comprises the following specific steps:

step 1.1, taking a current loop as the innermost loop of a control system, tracking a given current in real time, simultaneously inhibiting the fluctuation of the voltage of a direct current bus of the system and the torque pulsation of the system, and weakening the interference of back electromotive force to the system;

step 1.2, the speed loop is used as a control system intermediate loop to realize tracking of a given speed and suppress load disturbance which interferes with system operation;

and step 1.3, taking the position ring as the outermost ring of the control system, carrying out variable speed adjustment on the motor, and simultaneously controlling the position of the valve.

3. The method for constructing a control system of an intelligent valve electric actuator according to claim 2, wherein the position ring is used as the outermost ring of the control system in step 1.3, the motor is adjusted in a variable speed manner, and the valve position is controlled at the same time, specifically as follows:

the position ring controller adopts a sliding mode controller based on a novel approach law, a target position is used as the input of the position ring controller, a reference rotating speed is calculated according to the deviation of the target position and the current position, the input rotating speed of a speed ring is adjusted in real time, the actual rotating speed tracks the reference rotating speed, and the reference rotating speed and the actual rotating speed tend to zero as the current position approaches the target position;

taking the deviation value of the feedback value of the target position and the actual position as a control quantity, obtaining:

wherein x is₁，x₂As input parameters of the switching function, theta_refIs the target position, theta_mIs the actual position;

taking a slip form surface:

S＝cx₁+cx₂ (2)

wherein S is a sliding mode surface, and c is a coefficient in a switching function;

substituting formula (1) into formula (2) to obtain:

and (3) carrying out derivation on the formula (3), establishing a relation between a sliding mode surface and an approach law, and further obtaining:

wherein t is time, epsilon and k are coefficients in an exponential approximation law and are determined according to system requirements;

and obtaining a rotating speed instruction in the speed ring according to the deviation between the target position and the current position, wherein the obtained speed ring instruction is as follows according to an equation (4):

wherein ω is_refFor a given rotational speed input to the speed loop.

4. The method for constructing a control system of an electric actuator of an intelligent valve according to claim 1, wherein the MRAS with the variable gain coefficient in the step 2 is used for identifying the rotational inertia of the system on line, the rotational inertia is used as the input of an observer for observing the load torque of the system on line, and parameters of a system controller are adjusted according to the observation result, specifically as follows:

step 2.1, adopting an MRAS model combined with a variable gain coefficient, referring to a hybrid observation model of an adaptive algorithm and a load torque observer, and carrying out online identification on the rotational inertia of the system;

step 2.2, taking the rotational inertia as the input of a load torque observer, establishing an observation equation of the reduced-order load torque observer, and observing the load torque of the system in real time;

and 2.3, establishing a functional relation between the rotational inertia of the system and the controller parameters, using the identification result of the rotational inertia to adjust the parameters of the speed loop controller in real time, realizing the self-tuning of the PI parameters, and using the observation result of the load torque to perform feedforward compensation on the torque current.

5. The method for constructing the control system of the electric actuator of the intelligent valve according to claim 4, wherein in step 2.1, the MRAS model combined with the variable gain coefficient is adopted, and the rotational inertia of the system is identified on line by referring to a hybrid observation model of an adaptive algorithm and a load torque observer, specifically as follows:

according to the structure of the reference model, constructing a self-adaptive model with variable parameters, and adjusting the variable parameters of the adjustable model by using the errors of the reference model and the self-adaptive model until the output error of the reference model and the self-adaptive model is minimum;

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time;

neglecting the viscous friction term, the discretization difference equation of the kinematic equation is obtained as follows:

in the formula, T_sSampling time for a control algorithm; k. k-1 and k-2 are sampling moments;

within one sampling period, the load torque is constant:

T_L(k-1)＝T_L(k-2) (9)

obtaining a reference model from the equations (7) and (8):

ω_m(k)＝2*ω_m(k-1)+b*ΔT_e(k-1) (10)

wherein b is an identification parameter, Δ T_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

from the reference model, the adaptive model is constructed as follows:

whereinIs an estimate of the actual speed;is a parameter to be identified;ΔT_e(k-1)＝[T_e(k-1)-T_e(k-2)]；moment of inertia, Δ T, identified for the system_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

deviation delta omega of output of reference model and adjustable model_m(k) Comprises the following steps:

according to the discrete time iterative parameter identification mechanism proposed by Landau, the adaptive algorithm is designed as follows:

wherein, beta is self-adaptive gain; b (k) and b (k-1) are respectively the identification parameters of the system at the time k and the time k-1;

β＝β_min+[e(J)]²*β_max (14)

in the formula (14), beta_maxThe value of the adaptive gain coefficient is obtained when the system convergence speed is the fastest under the condition that the rotational inertia identification algorithm converges; beta is a_minThe value is the value when the identification precision is highest under the condition that the rotational inertia identification algorithm is converged; e (J) is the deviation value between the actual value of the moment of inertia and the recognition result.

6. The method for constructing the control system of the electric actuator of the intelligent valve according to claim 4, wherein step 2.2 is implemented by taking the moment of inertia as an input of a load torque observer, establishing an observation equation of a reduced-order load torque observer, and observing the load torque of the system in real time, wherein the method specifically comprises the following steps:

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time;

the load torque observer is used to estimate the unknown load torque T_L；

The load torque is considered a constant value over the sampling period:

according to the equations (15) and (16), the following system state equation is obtained:

wherein:

from equation (17), a reduced order load torque observer is obtained as follows:

wherein:

in the formula (I), the compound is shown in the specification,for the estimated state variable, K is the state feedback gain matrix, K₁、k₂Is a coefficient;

the system state error equation of the observer is obtained from the equations (17) and (18):

wherein the content of the first and second substances,is a state error vector;

the characteristic polynomial f (λ) of the system is:

wherein, λ is coefficient, and I is unit matrix;

the desired pole of the state equation is designed on the negative real axis:

λ²-(α+β)+β＝0 (21)

wherein α and β are the poles of the equation of state;

obtaining a feedback gain matrix according to equations (20) and (21):

the discretized difference equation of the load torque observer is obtained by equation (18):

where k, k +1 are sampling times, T_sIn order to be the time of sampling,the estimated values of the rotating speed at the k +1 moment and the k moment respectively,load torque estimated values at the k +1 and k moments respectively; omega_m(k) The mechanical angular speed of the rotor at the moment k;

asynchronous machines with field-oriented vector control, electromagnetic torque T_eWith torque current i_sqThe relationship of (a) to (b) is as follows:

wherein N is_pIs the number of pole pairs, L_mFor stator-rotor mutual inductance, L_rFor self-inductance of rotor, psi_rA rotor flux linkage;

when the motor runs at a constant speed omega_m1When the average value of the corresponding q-axis current is i_sq1Derived from formula (15):

K*i_sq1-T_L-B_m*ω_m1＝0 (25)

when the motor runs at a constant speed omega_m2When the average value of the corresponding q-axis current is i_sq2Derived from formula (15):

K*i_sq2-T_L-B_m*ω_m2＝0 (26)

the load torque including no disturbance torque error is calculated from equations (25) and (26) as follows:

the disturbance torque error caused by the errors of the rotational inertia and the viscous friction coefficient is defined as follows:

adopting MIT self-adaptation rate parameter identification to obtain an identification equation of rotational inertia and an identification equation of viscous friction coefficient as follows:

J(k+1)＝γ_J*T_dis*(ω_m(k)-ω_m(k-1))+J(k) (29)

B(k+1)＝T_dis*γ_B*T_dis*ω_m(k)+B(k) (30)

in the formula, gamma_JThe rotational inertia adjustment rate; gamma ray_BThe viscous friction coefficient adjustment rate.

7. The method for constructing the control system of the electric actuator of the intelligent valve according to claim 4, wherein in step 2.3, a functional relation between the rotational inertia of the system and the parameters of the controller is established, the identification result of the rotational inertia is used for adjusting the parameters of the speed loop controller in real time to realize the self-tuning of the PI parameters, and the observation result of the load torque is used for performing feedforward compensation on the torque current, specifically as follows:

the function relation between the system speed ring controller parameter and the rotary inertia is as follows:

in formula (31), K_TIs a torque constant, T_iIs equivalent of current loopAn inter constant; k of speed loop PI regulator_p、K_iThe control system identifies the rotational inertia and the load torque in real time, wherein the rotational inertia is used for realizing the parameter self-tuning of a speed ring PI, and the load torque is used for performing feedforward compensation on torque current.

8. The method for constructing the control system of the electric actuator of the intelligent valve according to claim 1, wherein the temperature detection circuit monitors the change of the rotor time constant in step 3, and according to the relation between the rotor time constant and the electromagnetic torque, the temperature compensation is performed for the error caused by the observation of the load torque of the control system by the temperature, specifically as follows:

step 3.1, analyzing the functional relation between the rotor time constant and the electromagnetic torque of the electric actuating mechanism;

and 3.2, establishing a functional relation between the rotor time constant and the stator temperature, correcting the electromagnetic torque according to the change of the rotor time constant, and correcting the load torque observation result of the control system.

9. The method for constructing a control system of an intelligent valve electric actuator according to claim 8, wherein the functional relationship between the rotor time constant and the electromagnetic torque of the electric actuator is analyzed in step 3.1, and the method comprises the following steps:

electromagnetic torque T of electric actuator_eComprises the following steps:

in the formula n_pIs the number of pole pairs, L_mIs air gap leakage inductance, i_dFor the excitation current component, #_rIs a flux linkage, L_rIs a rotor inductance;

when the rotor time constant changes, the electromagnetic torque changes:

in the formula T_rIs the rotor time constant, i_qIs a torque current component; t is_e' is the electromagnetic torque at the actual temperature; t is_eIs a theoretical value of electromagnetic torque; l'_m、L'_r、i'_q、ψ'_rRespectively air gap leakage inductance, rotor inductance, exciting current component and flux linkage at actual temperature;

order:equation (7) is simplified to:

wherein i'_dFor the torque current component at actual temperature, T_rIs the rotor time constant, T'_rIs the rotor time constant at the actual temperature;

order to

When a → a is given to a → 0,the electromagnetic torque is proportional to the torque time constant;

when a → ∞ is reached,the electromagnetic torque is inversely proportional to the torque time constant.

10. The method for constructing a control system of an intelligent valve electric actuator according to claim 8, wherein the step 3.2 establishes a functional relationship between a rotor time constant and a stator temperature, corrects the electromagnetic torque according to the change of the rotor time constant, and corrects the load torque observation result of the control system, specifically as follows:

step 3.2.1, measuring the temperature value of the stator and the rotor of the asynchronous motor, and obtaining the function relation of the stator and the rotor temperature by adopting a curve fitting mode:

t_rotor＝f₁(t_Stator) (35)

Wherein t is_Rotor、t_StatorThe temperature of the rotor and the stator of the asynchronous motor is measured; f. of₁Is a function relation between the temperatures of the stator and the rotor;

step 3.2.2, the rotor resistance changes along with the change of the temperature, the change of the rotor resistance causes the change of the rotor time constant, and the change relation between the rotor time constant and the rotor temperature is set as follows:

T_r＝f₂(t_rotor) (36)

Wherein T is_r、f₂As a function of the rotor time constant and the rotor temperature, t_RotorMeasuring the temperature value of the asynchronous motor rotor;

step 3.2.3, obtaining the relation between the rotor time constant and the stator temperature by the equation (35) and the equation (36):

T_r＝f₂(f₁(t_stator)) (37)

Step 3.2.4, setting the function relation of the stator and rotor temperature as follows:

t_rotor＝f₁(t_Stator)+Δf₁(t_Stator) (38)

In the formula,. DELTA.f₁(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

the rotor time constant is then expressed as:

T_r＝f₂(f₁(t_stator)+Δf₁(t_Stator))＝f₂[f₁(t_Stator)]+Δf₂(t_Stator) (39)

In formula (39): Δ f₂(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

when Δ f₂(t_Stator) At > 0: the theoretical calculation of the rotor time constant is larger, namely the rotor time constant used by the control system is larger than the actual rotor time constant of the motor; when Δ f₂(t_Stator) When < 0: the theoretical calculation of the rotor time constant is smaller, namely the rotor time constant used by the control system is smaller than the actual rotor time constant of the motor;

step 3.2.5, the electromagnetic torque is corrected based on the functional relationship between the rotor time constant and the electromagnetic torque.

Background

With the development of industrial production automation, the application of the electric actuating mechanism is more and more extensive, and the requirements of each control field on the aspects of response speed, control precision, anti-interference performance and the like of the electric actuating mechanism control system are higher and higher. The research and optimization of the electric actuating mechanism control system have important theoretical significance and practical value for improving the control precision and stability of the system.

The control system is used as an important component of an intelligent valve electric actuator, the control target of the control system is to accurately control the position of a valve, the traditional control system usually adopts position single-loop control, the speed of a driving motor cannot be flexibly adjusted in the traditional single-loop control mode, and the dynamic stability of the control system is poor. The problem that the position control can oscillate in a small range due to the fact that the speed cannot be adjusted at will is solved, the position control precision of an actuating mechanism is influenced, the valve blockage phenomenon can be caused in severe cases, and meanwhile the service life of mechanical parts can be greatly shortened. In some application occasions with higher requirements on the performance of a control system, such as a high-pressure and high-flow working environment and a system with high response speed and high control precision of the control system, a common actuating mechanism is difficult to be sufficient due to the insufficient control performance of the actuating mechanism.

The temperature change has certain influence on the performance of the control system of the electric actuating mechanism, and particularly influences the parameter identification of the motor, so that the temperature compensation has important significance for improving the control performance of the system. The current common temperature compensation method is to firstly analyze the error caused by the temperature to the torque measurement of the control system, establish a reasonable temperature compensation system, and design a differential compensation system with a self-calibration function according to different temperature compensation principles and the actual situation of the measurement. The system measures torque and corrects temperature by controlling the connection and disconnection of a servo driving device, but the system cannot meet the operation requirement of an electric actuating mechanism of an intelligent valve.

Disclosure of Invention

The invention aims to provide a method for constructing a control system of an intelligent valve electric actuator, so that the control precision and stability of the intelligent valve electric actuator are improved.

The technical solution for realizing the purpose of the invention is as follows: a control system construction method of an intelligent valve electric actuator comprises the following steps:

step 1, establishing a three-loop control system based on a current loop, a speed loop and a position loop;

step 2, adopting a variable gain coefficient MRAS to perform online identification on the rotational inertia of the system, taking the rotational inertia as the input of an observer to perform online observation on the load torque of the system, and setting the parameters of a system controller according to the observation result;

and 3, monitoring the change of the time constant of the rotor through a temperature detection circuit, and performing temperature compensation aiming at an error caused by temperature to the observation of the load torque of the control system according to the relation between the time constant of the rotor and the electromagnetic torque.

Compared with the prior art, the invention has the remarkable advantages that: (1) the electric actuating mechanism control system is designed into a position + speed + current three-loop control mode, so that the accurate control of the position of the valve of the electric actuating mechanism is realized, and the stability and the control accuracy of the electric actuating mechanism are improved; (2) a speed inner ring is introduced into a traditional control strategy to reduce the speed fluctuation, overshoot and oscillation of the system and improve the positioning precision and control sensitivity of the system; (3) a position ring is additionally arranged outside the speed ring to realize the accurate control of the position of the valve of the electric actuating mechanism, thereby further improving the control accuracy of the control system, avoiding the overshoot phenomenon of the traditional electric actuating mechanism at the start-stop moment and realizing the flexible start-stop function; (4) temperature compensation is introduced into the control system, so that the influence of temperature change on parameter identification in the operation process of the electric actuating mechanism is reduced, the identification precision is improved, and the control performance of the system is improved.

Drawings

Fig. 1 is a schematic block diagram of a three-loop control system in the method for constructing a control system of an intelligent valve electric actuator according to the present invention.

Fig. 2 is a schematic block diagram of the overall control system in the present invention.

FIG. 3 is a functional block diagram of a parameter identification model in the present invention.

Detailed Description

The invention discloses a method for constructing a control system of an intelligent valve electric actuating mechanism, which comprises the following steps:

step 1, establishing a three-loop control system based on a current loop, a speed loop and a position loop;

step 2, adopting a variable gain coefficient MRAS to perform online identification on the rotational inertia of the system, taking the rotational inertia as the input of an observer to perform online observation on the load torque of the system, and setting the parameters of a system controller according to the observation result;

and 3, monitoring the change of the time constant of the rotor through a temperature detection circuit, and performing temperature compensation aiming at an error caused by temperature to the observation of the load torque of the control system according to the relation between the time constant of the rotor and the electromagnetic torque.

Further, the establishment of the three-loop control system based on the current loop, the speed loop and the position loop in step 1 is as follows:

step 1.1, taking a current loop as the innermost loop of a control system, tracking a given current in real time, simultaneously inhibiting the fluctuation of the voltage of a direct current bus of the system and the torque pulsation of the system, and weakening the interference of back electromotive force to the system;

step 1.2, the speed loop is used as a control system intermediate loop to realize tracking of a given speed and suppress load disturbance which interferes with system operation;

and step 1.3, taking the position ring as the outermost ring of the control system, carrying out variable speed adjustment on the motor, and simultaneously controlling the position of the valve.

Further, in step 1.3, the position ring is used as the outermost ring of the control system to perform variable speed adjustment on the motor and control the position of the valve, specifically as follows:

the position ring controller adopts a sliding mode controller based on a novel approach law, a target position is used as the input of the position ring controller, a reference rotating speed is calculated according to the deviation of the target position and the current position, the input rotating speed of a speed ring is adjusted in real time, the actual rotating speed tracks the reference rotating speed, and the reference rotating speed and the actual rotating speed tend to zero as the current position approaches the target position;

taking the deviation value of the feedback value of the target position and the actual position as a control quantity, obtaining:

wherein x is₁，x₂As input parameters of the switching function, theta_refIs the target position, theta_mIs the actual position;

taking a slip form surface:

S＝cx₁+cx₂ (2)

wherein S is a sliding mode surface, and c is a coefficient in a switching function;

substituting formula (1) into formula (2) to obtain:

and (3) carrying out derivation on the formula (3), establishing a relation between a sliding mode surface and an approach law, and further obtaining:

wherein t is time, epsilon and k are coefficients in an exponential approximation law and are determined according to system requirements;

and obtaining a rotating speed instruction in the speed ring according to the deviation between the target position and the current position, wherein the obtained speed ring instruction is as follows according to an equation (4):

wherein ω is_refFor a given rotational speed input to the speed loop.

Further, the step 2 of performing online identification on the system moment of inertia by using the variable gain coefficient MRAS, performing online observation on the system load torque by using the moment of inertia as the input of the observer, and setting the system controller parameters according to the observation result, specifically as follows:

step 2.1, adopting an MRAS model combined with a variable gain coefficient, referring to a hybrid observation model of an adaptive algorithm and a load torque observer, and carrying out online identification on the rotational inertia of the system;

step 2.2, taking the rotational inertia as the input of a load torque observer, establishing an observation equation of the reduced-order load torque observer, and observing the load torque of the system in real time;

and 2.3, establishing a functional relation between the rotational inertia of the system and the controller parameters, using the identification result of the rotational inertia to adjust the parameters of the speed loop controller in real time, realizing the self-tuning of the PI parameters, and using the observation result of the load torque to perform feedforward compensation on the torque current.

Further, in step 2.1, the moment of inertia of the system is identified on line by using the variable gain coefficient-combined MRAS model, referring to a hybrid observation model of an adaptive algorithm and a load torque observer, which specifically includes:

according to the structure of the reference model, constructing a self-adaptive model with variable parameters, and adjusting the variable parameters of the adjustable model by using the errors of the reference model and the self-adaptive model until the output error of the reference model and the self-adaptive model is minimum;

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time;

neglecting the viscous friction term, the discretization difference equation of the kinematic equation is obtained as follows:

in the formula, T_sSampling time for a control algorithm; k. k-1 and k-2 are sampling moments;

within one sampling period, the load torque is constant:

T_L(k-1)＝T_L(k-2) (9)

obtaining a reference model from the equations (7) and (8):

ω_m(k)＝2*ω_m(k-1)+b*ΔT_e(k-1) (10)

wherein b is an identification parameter, Δ T_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

from the reference model, the adaptive model is constructed as follows:

whereinIs an estimate of the actual speed;is a parameter to be identified;ΔT_e(k-1)＝[T_e(k-1)-T_e(k-2)]；moment of inertia, Δ T, identified for the system_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

deviation delta omega of output of reference model and adjustable model_m(k) Comprises the following steps:

according to the discrete time iterative parameter identification mechanism proposed by Landau, the adaptive algorithm is designed as follows:

wherein, beta is self-adaptive gain; b (k) and b (k-1) are respectively the identification parameters of the system at the time k and the time k-1;

β＝β_min+[e(J)]²*β_max (14)

in the formula (14), beta_maxThe value of the adaptive gain coefficient is obtained when the system convergence speed is the fastest under the condition that the rotational inertia identification algorithm converges; beta is a_minThe value is the value when the identification precision is highest under the condition that the rotational inertia identification algorithm is converged; e (J) is the deviation value between the actual value of the moment of inertia and the recognition result.

Further, step 2.2, the rotational inertia is used as an input of the load torque observer, an observation equation of the reduced-order load torque observer is established, and the load torque of the system is observed in real time, which specifically includes the following steps:

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time;

the load torque observer is used to estimate the unknown load torque T_L；

The load torque is considered a constant value over the sampling period:

according to the equations (15) and (16), the following system state equation is obtained:

wherein:

u＝T_e，y＝ω_m

C＝[1 0]

from equation (17), a reduced order load torque observer is obtained as follows:

wherein:in the formula (I), the compound is shown in the specification,for the estimated state variable, K is the state feedback gain matrix, K₁、k₂Is a coefficient;

the system state error equation of the observer is obtained from the equations (17) and (18):

wherein the content of the first and second substances,is a state error vector;

the characteristic polynomial f (λ) of the system is:

wherein, λ is coefficient, and I is unit matrix;

the desired pole of the state equation is designed on the negative real axis:

λ²-(α+β)+β＝0 (21)

wherein α and β are the poles of the equation of state;

obtaining a feedback gain matrix according to equations (20) and (21):

the discretized difference equation of the load torque observer is obtained by equation (18):

where k, k +1 are sampling times, T_sIn order to be the time of sampling,the estimated values of the rotating speed at the k +1 moment and the k moment respectively,load torque estimated values at the k +1 and k moments respectively; omega_m(k) The mechanical angular speed of the rotor at the moment k;

asynchronous machines with field-oriented vector control, electromagnetic torque T_eWith torque current i_sqThe relationship of (a) to (b) is as follows:

wherein N is_pIs the number of pole pairs, L_mFor stator-rotor mutual inductance, L_rFor self-inductance of rotor, psi_rA rotor flux linkage;

when the motor runs at a constant speed omega_m1When the average value of the corresponding q-axis current is i_sq1Derived from formula (15):

K*i_sq1-T_L-B_m*ω_m1＝0 (25)

when the motor runs at a constant speed omega_m2When the average value of the corresponding q-axis current is i_sq2Derived from formula (15):

K*i_sq2-T_L-B_m*ω_m2＝0 (26)

the load torque including no disturbance torque error is calculated from equations (25) and (26) as follows:

the disturbance torque error caused by the errors of the rotational inertia and the viscous friction coefficient is defined as follows:

adopting MIT self-adaptation rate parameter identification to obtain an identification equation of rotational inertia and an identification equation of viscous friction coefficient as follows:

J(k+1)＝γ_J*T_dis*(ω_m(k)-ω_m(k-1))+J(k) (29)

B(k+1)＝T_dis*γ_B*T_dis*ω_m(k)+B(k) (30)

in the formula, gamma_JThe rotational inertia adjustment rate; gamma ray_BThe viscous friction coefficient adjustment rate.

Further, step 2.3, establishing a functional relationship between the system rotational inertia and the controller parameter, using the identification result of the rotational inertia to adjust the speed loop controller parameter in real time, implementing the self-tuning of the PI parameter, and using the observation result of the load torque to perform feed-forward compensation on the torque current, specifically as follows:

the function relation between the system speed ring controller parameter and the rotary inertia is as follows:

in formula (31), K_TIs a torque constant, T_iIs the equivalent time constant of the current loop; k of speed loop PI regulator_p、K_iThe control system identifies the rotational inertia and the load torque in real time, wherein the rotational inertia is used for realizing the parameter self-tuning of a speed ring PI, and the load torque is used for performing feedforward compensation on torque current.

Further, step 3, monitoring the change of the rotor time constant through a temperature detection circuit, and performing temperature compensation aiming at an error caused by temperature to the observation of the load torque of the control system according to the relation between the rotor time constant and the electromagnetic torque, specifically as follows:

step 3.1, analyzing the functional relation between the rotor time constant and the electromagnetic torque of the electric actuating mechanism;

and 3.2, establishing a functional relation between the rotor time constant and the stator temperature, correcting the electromagnetic torque according to the change of the rotor time constant, and correcting the load torque observation result of the control system.

Further, the functional relationship between the rotor time constant and the electromagnetic torque of the electric actuator is analyzed in step 3.1, specifically as follows:

electromagnetic torque T of electric actuator_eComprises the following steps:

in the formula n_pIs the number of pole pairs, L_mIs air gap leakage inductance, i_dFor the excitation current component, #_rIs a flux linkage, L_rIs a rotor inductance;

when the rotor time constant changes, the electromagnetic torque changes:

in the formula T_rIs the rotor time constant, i_qIs a torque current component; t is_e' is the electromagnetic torque at the actual temperature; t is_eIs a theoretical value of electromagnetic torque; l'_m、L'_r、i'_q、ψ'_rRespectively air gap leakage inductance, rotor inductance, exciting current component and flux linkage at actual temperature;

order:equation (7) is simplified to:

wherein i'_dFor the torque current component at actual temperature, T_rIs the rotor time constant, T'_rIs the rotor time constant at the actual temperature;

order to

When a → a is given to a → 0,the electromagnetic torque is proportional to the torque time constant;

when a → ∞ is reached,the electromagnetic torque is inversely proportional to the torque time constant.

Further, the step 3.2 of establishing a functional relationship between the rotor time constant and the stator temperature, correcting the electromagnetic torque according to the change of the rotor time constant, and correcting the observation result of the load torque of the control system specifically includes:

step 3.2.1, measuring the temperature value of the stator and the rotor of the asynchronous motor, and obtaining the function relation of the stator and the rotor temperature by adopting a curve fitting mode:

t_rotor＝f₁(t_Stator) (35)

Wherein t is_Rotor、t_StatorThe temperature of the rotor and the stator of the asynchronous motor is measured; f. of₁Is a function relation between the temperatures of the stator and the rotor;

step 3.2.2, the rotor resistance changes along with the change of the temperature, the change of the rotor resistance causes the change of the rotor time constant, and the change relation between the rotor time constant and the rotor temperature is set as follows:

T_r＝f₂(t_rotor) (36)

Wherein T is_r、f₂As a function of the rotor time constant and the rotor temperature, t_RotorMeasuring the temperature value of the asynchronous motor rotor;

step 3.2.3, obtaining the relation between the rotor time constant and the stator temperature by the equation (35) and the equation (36):

T_r＝f₂(f₁(t_stator)) (37)

Step 3.2.4, setting the function relation of the stator and rotor temperature as follows:

t_rotor＝f₁(t_Stator)+Δf₁(t_Stator) (38)

In the formula,. DELTA.f₁(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

the rotor time constant is then expressed as:

T_r＝f₂(f₁(t_stator)+Δf₁(t_Stator))＝f₂[f₁(t_Stator)]+Δf₂(t_Stator) (39)

In formula (39): Δ f₂(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

when Δ f₂(t_Stator) At > 0: the theoretical calculation of the rotor time constant is larger, namely the rotor time constant used by the control system is larger than the actual rotor time constant of the motor; when Δ f₂(t_Stator) When < 0: the theoretical calculation of the rotor time constant is smaller, namely the rotor time constant used by the control system is smaller than the actual rotor time constant of the motor;

step 3.2.5, the electromagnetic torque is corrected based on the functional relationship between the rotor time constant and the electromagnetic torque.

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

Examples

With reference to fig. 1, fig. 2, and fig. 3, the method for constructing a control system of an intelligent valve electric actuator according to the present invention includes the following steps:

step 1: the three-loop control system of the current loop, the speed loop and the position loop is designed to improve the position control precision and the operation stability of the electric actuating mechanism control system, and meanwhile, the flexible start and stop of the electric actuating mechanism are realized, and the three-loop control system is specifically as follows by combining a figure 1:

step 1.1, a current loop is used as the innermost loop of a control system, given current is tracked in real time, meanwhile, the voltage fluctuation of a direct-current bus of the system and the torque ripple of the system are restrained, and the interference of back electromotive force to the system is weakened;

step 1.2, the speed loop is used as a control system intermediate loop to realize the tracking of a given speed and suppress the load disturbance which interferes with the operation of the system;

step 1.3, the position ring is used as the outermost ring of the control system, the target position is used as the input of the regulator, a reference rotating speed suitable for the current error is calculated according to the error between the target position and the current position, the input rotating speed of the speed ring is regulated in real time, the actual rotating speed tracks the reference rotating speed to carry out variable speed regulation on the motor, meanwhile, the valve position is accurately controlled, and no overshoot is ensured, and the method specifically comprises the following steps:

the position ring controller adopts a sliding mode controller based on a novel approach law, a target position is used as the input of the position ring regulator, a reference rotating speed suitable for the current error is calculated according to the error between the target position and the current position, the input rotating speed of the speed ring is regulated in real time, the actual rotating speed tracks the reference rotating speed, the reference rotating speed and the actual rotating speed tend to zero along with the approach of the current position to the target position, the reference rotating speed is self-adaptively regulated according to the current position, the closer the reference rotating speed is to the target position, the smaller the reference rotating speed is, and the flexible start-stop function is realized;

taking the deviation amount of the feedback value of the given position and the actual position of the system as a control amount, the following can be obtained:

wherein x is₁，x₂As input parameters of the switching function, theta_refIs the target position, theta_mIs the actual position;

taking a slip form surface:

S＝cx₁+cx₂ (2)

wherein S is a sliding mode surface, and c is a coefficient in a switching function;

the formula (1) can be substituted for the formula (2):

the formula (3) is derived, and the relation between the formula (3) and the approach law is established, so that the following result is obtained:

wherein t is time, epsilon and k are coefficients in an exponential approximation law and are determined according to system requirements;

the core content of the position controller is to obtain a rotating speed instruction in a speed ring according to the deviation between the position and the position instruction, and according to the formula (4), the speed ring instruction can be obtained as follows:

wherein ω is_refFor a given rotational speed input to the speed loop.

Step 2: the method comprises the following steps of adopting a model reference adaptive algorithm combined with a variable gain coefficient and a mixed observation model of a load torque observer to identify the rotational inertia of a system on line, then taking an identification result as the input of the load torque observer, establishing an observation equation of a reduced-order load torque observer, observing the load torque of the system in real time, finally setting parameters of a system controller in real time according to the identification result of system parameters, and constructing a parameter adaptive system, wherein the method specifically comprises the following steps:

step 2.1: the method comprises the following steps of adopting an MRAS model combined with a variable gain coefficient, referring to a self-adaptive algorithm and a mixed observation model of a load torque observer, and carrying out online identification on the rotational inertia of the system, wherein the method specifically comprises the following steps:

according to the structure of the reference model, constructing a self-adaptive model with variable parameters, and adjusting the variable parameters of the adjustable model according to a certain self-adaptive algorithm by using the errors of the reference model and the self-adaptive model until the output error of the reference model and the self-adaptive model is minimum;

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time;

considering that the sampling rate of the control algorithm is high enough, the viscous friction term can be ignored, and the discretization difference equation of the kinematic equation is obtained as follows:

in the formula, T_sSampling time for a control algorithm; k. k-1 and k-2 are sampling moments;

since the sampling time is sufficiently short, the load torque can be considered constant within one sampling period:

T_L(k-1)＝T_L(k-2) (9)

obtaining a reference model from the equations (7) and (8):

ω_m(k)＝2*ω_m(k-1)+b*ΔT_e(k-1) (10)

wherein b is an identification parameter, Δ T_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

from the reference model, the adaptive model can be designed as follows:

whereinIs an estimate of the actual speed;is a parameter to be identified;ΔT_e(k-1)＝[T_e(k-1)-T_e(k-2)]；moment of inertia, Δ T, identified for the system_e(k-1) is the difference between the electromagnetic torque at the time k-1 and the electromagnetic torque at the time k-2;

the deviation of the reference model and the adjustable model output is:

according to the discrete time iterative parameter identification mechanism proposed by Landau, the adaptive algorithm is designed as follows:

wherein, beta is self-adaptive gain; b (k) and b (k-1) are respectively the identification parameters of the system at the time k and the time k-1;

β＝β_min+[e(J)]²*β_max (14)

in the formula (14), beta_maxThe value of the adaptive gain coefficient is obtained when the system convergence speed is the fastest under the condition that the rotational inertia identification algorithm converges; beta is a_minThe value is the value when the identification precision is highest under the condition that the rotational inertia identification algorithm is converged; e (J) is the deviation value between the actual value of the moment of inertia and the recognition result.

Step 2.2: the method comprises the following steps of establishing an observation equation of a reduced-order load torque observer by taking the moment of inertia as the input of the observer, and observing the load torque of a system in real time, wherein the method specifically comprises the following steps:

the mechanical kinematic equation of the asynchronous motor is as follows:

wherein, ω is_mIs the rotor mechanical angular velocity; j is the load torque inertia of the motor; b is_mIs a viscous friction coefficient; t is_eIs an electromagnetic torque; t is_LIs the load torque; t is time; the load torque observer is used to estimate the unknown load torque T_L；

Considering that the sampling rate of the control algorithm is sufficiently high, the load torque can be considered as a constant value during the sampling period:

according to the equations (15) and (16), the following system state equation is obtained:

wherein:

u＝T_e，y＝ω_m

C＝[1 0]

from equation (17), a reduced order load torque observer is obtained as follows:

wherein:

in the formula (I), the compound is shown in the specification,for the estimated state variable, K is the state feedback gain matrix, K₁、k₂Is a coefficient;

the system state error equation of the observer is obtained from the equations (17) and (18):

wherein the content of the first and second substances,is a state error vector;

the characteristic polynomial f (λ) of the system is:

where λ is the coefficient and I is the identity matrix.

The desired pole of the state equation is designed on the negative real axis:

λ²-(α+β)+β＝0 (21)

wherein α and β are the poles of the equation of state;

obtaining a feedback gain matrix according to equations (20) and (21):

the discretized difference equation of the load torque observer is obtained by equation (18):

where k, k +1 are sampling times, T_sIn order to be the time of sampling,the rotation at the time k +1 and kThe speed of the vehicle is estimated by the speed estimation value,load torque estimated values at the k +1 and k moments respectively; omega_m(k) The mechanical angular speed of the rotor at the moment k;

asynchronous machines with field-oriented vector control, electromagnetic torque T_eWith torque current i_sqThe relationship of (a) to (b) is as follows:

wherein N is_pIs the number of pole pairs, L_mFor stator-rotor mutual inductance, L_rFor self-inductance of rotor, psi_rA rotor flux linkage;

when the motor runs at a constant speed omega_m1When the average value of the corresponding q-axis current is i_sq1Derived from formula (15):

K*i_sq1-T_L-B_m*ω_m1＝0 (25)

when the motor runs at a constant speed omega_m2When the average value of the corresponding q-axis current is i_sq2Derived from formula (15):

K*i_sq2-T_L-B_m*ω_m2＝0 (26)

the load torque including no disturbance torque error is calculated from equations (25) and (26) as follows:

the disturbance torque error caused by the errors of the rotational inertia and the viscous friction coefficient is defined as follows:

adopting MIT self-adaptation rate parameter identification to obtain an identification equation of rotational inertia and an identification equation of viscous friction coefficient as follows:

J(k+1)＝γ_J*T_dis*(ω_m(k)-ω_m(k-1))+J(k) (29)

B(k+1)＝T_dis*γ_B*T_dis*ω_m(k)+B(k) (30)

in the formula, gamma_JThe rotational inertia adjustment rate; gamma ray_BThe viscous friction coefficient adjustment rate.

On-line identification of the moment of inertia based on a disturbance load torque observer requires calculation of real-time load torque according to equation (27), i.e., two different steady-state rotational speeds are required for calculation. However, in many applications, the position and speed commands are issued by an upper computer (PLC or motion control card), and thus it is difficult to independently calculate the load torque. A hybrid observation model combining a model reference adaptive algorithm and a load torque observer is provided, and the model combines the advantages of the two algorithms to obtain ideal results. Firstly, a model reference adaptive algorithm is adopted to identify the rotational inertia in real time, and then the rotational inertia is used as the input of a load torque observer, so that the load torque can be observed in real time.

Step 2.3, establishing a functional relation between the rotational inertia of the system and the load torque and the controller parameters, using the identification result of the rotational inertia to adjust the parameters of the speed loop controller in real time, realizing the self-tuning of the PI parameters, and using the observation result of the load torque to perform feedforward compensation on the torque current, thereby improving the stability and the control precision of the control system, which are specifically as follows:

the function relation between the system speed ring controller parameter and the rotary inertia is as follows:

in formula (31), K_TIs a torque constant, T_iIs the equivalent time constant of the current loop; k of speed loop PI regulator_p、K_iAre all in direct proportion to the rotational inertia J of the system, and the control system identifies the rotational inertia and the load torque in real time, wherein the rotational inertia is used for realizing the speed loop PIAnd self-tuning parameters, wherein the load torque is used for performing feedforward compensation on the torque current.

And step 3: the temperature compensation scheme is designed for temperature compensation according to the relation between the rotor time constant and the torque current and the error caused by the observation of the temperature to the load torque of the control system by monitoring the change of the rotor time constant through the temperature detection circuit, and specifically comprises the following steps:

step 3.1, analyzing the functional relation between the rotor time constant and the electromagnetic torque of the electric actuating mechanism, which is concretely as follows:

electromagnetic torque T of electric actuator_eComprises the following steps:

in the formula n_pIs the number of pole pairs, L_mIs air gap leakage inductance, i_dFor the excitation current component, #_rIs a flux linkage, L_rIs a rotor inductance;

when the rotor time constant changes, the electromagnetic torque changes:

in the formula T_rIs the rotor time constant, i_qIs a torque current component; t is_e' is the electromagnetic torque at the actual temperature; t is_eIs a theoretical value of electromagnetic torque; l'_m、L'_r、i'_q、ψ'_rRespectively air gap leakage inductance, rotor inductance, exciting current component and flux linkage at actual temperature;

order:equation (7) is simplified to:

wherein i'_dFor the torque current component at actual temperature, T_rIs the rotor time constant, T'_rIs the rotor time constant at the actual temperature;

order to

When a → a is given to a → 0,the electromagnetic torque is proportional to the torque time constant;

when a → ∞ is reached,the electromagnetic torque is inversely proportional to the torque time constant.

Step 3.2: the method comprises the following steps of establishing a functional relation between a rotor time constant and a stator temperature, correcting electromagnetic torque according to the change of the rotor time constant, correcting a load torque observation result of a control system, improving the identification precision of the control system, and improving the control performance of the control system, wherein the functional relation comprises the following specific steps:

step 3.2.1, measuring the temperature value of the stator and the rotor of the asynchronous motor, and obtaining the function relation of the stator and the rotor temperature by adopting a curve fitting mode:

t_rotor＝f₁(t_Stator) (35)

Wherein t is_Rotor、t_StatorThe temperature of the rotor and the stator of the asynchronous motor is measured; f. of₁Is a function relation between the temperatures of the stator and the rotor;

step 3.2.2, the rotor resistance changes along with the change of the temperature, the change of the rotor resistance causes the change of the rotor time constant, and the change relation between the rotor time constant and the rotor temperature is set as follows:

T_r＝f₂(t_rotor) (36)

Wherein T is_r、f₂As a function of the rotor time constant and the rotor temperature, t_RotorFor measuring obtained asynchronous motor rotorThe temperature value of the seed;

step 3.2.3, obtaining the relation between the rotor time constant and the stator temperature by the equation (35) and the equation (36):

T_r＝f₂(f₁(t_stator)) (37)

Step 3.2.4, setting the function relation of the stator and rotor temperature as follows:

t_rotor＝f₁(t_Stator)+Δf₁(t_Stator) (38)

In the formula,. DELTA.f₁(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

the rotor time constant is then expressed as:

T_r＝f₂(f₁(t_stator)+Δf₁(t_Stator))＝f₂[f₁(t_Stator)]+Δf₂(t_Stator) (39)

In formula (39): Δ f₂(t_Stator) Is the deviation of the stator and rotor temperature functional relationship;

when Δ f₂(t_Stator) At > 0: the theoretical calculation of the rotor time constant is larger, namely the rotor time constant used by the control system is larger than the actual rotor time constant of the motor; when Δ f₂(t_Stator) When < 0: the theoretical calculation of the rotor time constant is smaller, namely the rotor time constant used by the control system is smaller than the actual rotor time constant of the motor;

and 3.2.5, obtaining the change of the rotor time constant according to the stator temperature change measured by the temperature sensor and the functional relation between the rotor time constant and the stator temperature, and finally correcting the electromagnetic torque according to the functional relation between the rotor time constant and the electromagnetic torque to improve the control performance of the control system.

The invention designs the electric actuator control system into a position + speed + current three-loop control mode, avoids the contradiction between control sensitivity and positioning accuracy in single-loop control adopted by the traditional electric actuator, and realizes accurate control on the position of the electric actuator valve, thereby improving the stability and control accuracy of the electric actuator.