RU2549149C2 - Digital control method - Google Patents

Abstract

FIELD: physics; control.

SUBSTANCE: invention relates to control of continuous technological processes, particularly inertial objects using computing equipment, and can be used in chemical, petrochemical and other industries. The method includes generating proportional and integral components of control action with adjustment of adjustment factors thereof, outputting a compensating action to task changes, using different discrete intervals for components of the control action.

EFFECT: improved quality of controlling inertial objects.

1 dwg

Description

FIELD OF THE INVENTION

The invention relates to the field of control of continuous technological processes, in particular inertial objects using computing hardware, and can be used in chemical, petrochemical and other industries.

State of the art

A known control method using a controller in which the driving action (task) is changed by a magnitude greater than the required modulating value, alternating values greater than the desired driving influence and smaller than the required driving action or without it, followed by setting the driving action to required value (international application published in accordance with the patent cooperation agreement, number WO 2012/091634 A1, 07/05/2012, G05B 11/00).

The known method as applied to inertial objects does not provide an acceptable quality of control, since it does not provide for effective adjustment of the control action and does not take into account the dynamics of the object.

There is also known a method of forming the PI-law of regulation and diagnostics of an automatic system, in which the input signal proportional to the control error is integrated, the resulting signal is summed with the input signal and the sum is scaled (thus forming a proportional-integral control action), determine the absolute value of the input the signal and when this threshold level is exceeded, the difference between the input signal and the signal proportional to the integration result is integrated, and with the help of Strongly assay times of threshold signal is formed on the emergency state of the system (RU 2150727 C1, 16.06.1999, G05B 11/36) (prototype).

The disadvantage of this method is the low quality control of the inertial object due to the ineffective formation of the control action both when practicing disturbances and when changing the task.

Disclosure of invention

The purpose of the invention is improving the quality of control of inertial objects.

This goal is achieved in that, in contrast to the known technical solution, in the proposed digital control method, the coefficient of the proportional component of the control action is corrected depending on the sign of the control error signal and the sign of its derivative, the coefficient of the integral component of the control action is changed depending on the absolute value of the control error signal , to the control action add a pulse compensating component, determined by the derivative for signal, and discrete intervals of the individual components are selected depending on the dynamic properties of the control object.

Description of drawings

Figure 1 presents the structural diagram of a system that implements the proposed method.

The implementation of the invention

The system that implements the proposed method contains (see Fig. 1) a control object 1, an adjustable variable sensor 2, a task driver 3 and a control controller 4. The controller implements functional blocks: a comparison unit 5, a multiplication unit 6, a proportional component correction unit 7, an integrating unit 8, an integral component correction unit 9, a pulse compensation unit 10, and an adder 11.

A system that implements the proposed method works as follows.

Using controller 4, digital (with a discrete interval h ₀ ) regulation of the variable X of object 1 is carried out by measurement from sensor 2 and the received reference Y from block 3. Regulation error E is calculated by block 5:

E _i = Y _i -X _i,

where i is the number of the digital control cycle with a discrete interval h ₀ .

Using block 6, the proportional component P of the control action U is determined:

P _i = K _n · E _i,

where K _n - tuning factor, corrected using block 7 in the n-th cycle with a discrete interval h ₁ (h ₁ ≥h ₀ ):

, $$K_n=\{\begin{array}{c}{K}^{\mathrm{one}},{\text{if E}}_{\text{n}}\cdot {\dot{E}}_n\ge 0\\ {\mathrm{<figure-callout\; id="2"\; label="K"\; filenames="00000007.png"\; state="\{\{state\}\}">K</figure-callout>}}^2,{\text{if E}}_{\text{n}}\cdot {\dot{E}}_n<0\end{array}$$

,

and custom constants K ¹ > K ² .

Using block 8, the integral component R of the control action U is calculated:

, $$R_i=R_{i-\mathrm{one}}+E_i\cdot \frac{h_0}{T_j}$$

,

where T _j is the tuning constant of the integration time, corrected using block 9 of the jth cycle with discreteness h ₂ (h ₂ ≥h ₀ ):

, $$T_j=\{\begin{array}{c}{T}^{\mathrm{one}},\text{if}\left|{\text{E}}_{\text{j}}\right|<\mathrm{<figure-callout\; id="2"\; label="q\; T"\; filenames="00000007.png"\; state="\{\{state\}\}">q\; </figure-callout>}& \\ T^{}{}^2,\text{if}\left|{\text{E}}_{\text{j}}\right|\ge q\end{array}$$

,

where q, T ¹ , T ² are custom constants, with T ¹ <T ² .

Using block 10 is issued in the n-th cycle with a discrete interval h _{1 the} compensating component of the control action U in the form of pulses of the following duration:

$$S_n=\frac{D\cdot (Y_n-Y{}_{n-\mathrm{one}})}{U_{\mathrm{and}mP}}$$

where U _imp - pulse level as an acceptable value of the compensating effect;

D - tuning factor, depending on the dynamic properties of object 1.

The control action is determined using block 11:

U _i = P _i + R _i + S _n .

An increase in the proportional component with an increase in the absolute value of the error (coincidence of the signs of the error and its derivative) allows one to reduce overshoot and reduce the system oscillation. An increase in the integral component at small error values improves the system astatism, reduces the regulation time. Pulse compensation of changes in the driving signal significantly increases the speed of the control system of the inertial object. The combination of all the components of the digital control action, and with different intervals of discreteness, can significantly improve the quality characteristics of the control of inertial objects.

To implement the method, for example, a BAZIS-21.2TSU control controller or another from the BAZIS ^® series manufactured by Ecoresurs CJSC can be used.

Implementation of the proposed method in commercially available BASIS® series controllers is scheduled for 2013.

Claims (1)

A method of digital control of an inertial object by generating a control error E _i proportional to P _i = K _n * E _i and the integral R _i = R _i-1 + E _i * h ₀ / T _{j of the} control action components, determining the absolute value of the control error, characterized in that it is further adjusted with discreteness h _{1 the} coefficient of the proportional component depending on the signs of the error and its derivative according to the formula

change with discreteness h _{2 the} time constant of the integral component depending on the absolute value of the error by the formula

and to the control action, with a discreteness h _{1, a} pulse compensating component with a pulse width determined in proportion to the derivative of the driving signal Y is added according to the formula S _n = D * (Y _n -Y _n-1 ) / U _imp with a tuning factor D and an acceptable pulse level U _imp

Images