DD273324A1 - METHOD FOR THE REALIZATION OF AN ADAPTIVE SCRAP CONTROL - Google Patents

Abstract

The solution relates to sampling devices with constant sampling times and variable dynamics, as they are used in control circuits using microcomputers. It should be designed with regard to the utilization of each available actuator power, time-optimal universal Abtastregler which is independent of Stoer and Fuehrungsverhalten whose control algorithm consists of few clear and easy to calculate links and is able to change the number without the controller Modify control steps so that when using the available actuator power the smallest number of control steps is required. A new approach for the controller design for disturbance behavior and in connection with a new approach for a setpoint filter for interference and Fuehrungsverhalten was found, according to their algorithms according to the respective available actuator performance, the number of control steps for faulty behavior and the number of control steps for Fuehrungsverhalten independent are adjustable from each other.

Description

y (n) = ZT a..y (n ~ i) + 21 c. , R (n ~ i-in.)

i =: l i = 0 '· -

in the

y (n) = the current controller output quantity at the sampling instant t = η * T with m ₀ = 0 (TA sampling time)

y (n-i) = the controller output to the sampling point t = (n-i) T

8i; Ci - regulator co-efficient

R (n-1 -mi) = · the time-optimal (dead-beat) controller component at the point in time t = T (n-1 -mi) s = the order of the controlled system

m, = the number of control steps

is, «

takes place, after which a control algorithm is calculated, which consists of the sum of a number of s evaluated controller output variables and the sum of a number of s + 1 time-offset valued dead-beat components according to the order of the controlled system, wherein the adjustability of the rise time of the disturbance by Change in the number of control steps is achieved.

In one embodiment of the invention, the current controller output variable is compared with a maximum output variable to form the lowest number of control steps until the equation

Ytm.) <YmK <Vlm. - 1)

is satisfied.

In a further embodiment of the invention, the respective current is to realize a temporally optimal sampling controller

maximum controller output resulting from the allowable limit value and the respective load value after the equation

/ ym «/ = /ye.gr. - Yl (0 /

results, provided.

In a further development of the invention, the realization of good leadership behavior is based on time-optimal interference behavior

designed regulator performed a setpoint filter, jng, the design of the filter based on the found

filter equation

in the

w '(n) = the current filter output at the sampling time t = η · T with T as sampling time

w '(n-rrii) = the filter output to sample point t = (n-mi) T

bi; dj = filter coefficients

w (n-nrij) = the filter input at sampling time t = (n-mj) T

s = the order of the controlled system

f = the order of the filter

m, = the number of steps in the optimization of the disturbance behavior

{& Controller optimization)

mi = the number of controller steps in the optimization for leadership behavior

takes place, after which a filter algorithm is calculated, which according to the order (s) of the controlled system consists of the sum of a number of s weighted, time-offset filter outputs and the sum of a number of f-rated time-offset filter inputs, the adjustability of the rise time of the disturbance variable by change the number of control steps for disturbance behavior and the adjustability of the rise time of the reference variable is achieved by changing the number of control steps in leadership behavior.

In one embodiment of this development of the invention, the current controller output variable is compared with a maximum output variable to form the least number of control steps in the case of disturbance behavior and the least number of control steps in leadership behavior, until the equation

MmJ <y _m «x <Mm. -1)

is satisfied.

In a further refinement of the development of the invention, in order to realize a temporally optimal sampling controller for disturbing and guiding behavior, the respectively current maximum output variable, which results from the permissible limiting value and the respective load value according to the equation

/ y _m «/=/ye.gr. - Yl (t) / yields, provided.

Ausführungsbelsplel

The invention will be explained in more detail below in a Ausführabelsplel.

Compared to the previous approach to the suboptimal controller design on leadership behavior of the inventive approach from the controller design on interference behavior, from the course of the controller output variable y for disturbance S from. The actuator power is always assumed to be constant during a starting step m. To simplify the consideration, the actuator size was assumed to be identical to the controller output. The new approach leads to a controller equation for the current controller output

in which mean

y (n -1) = the current controller output at sampling time t = (n -1) T (T & sampling time)

8i; ci = regulator coefficients

R (n-1 -mi) = the time-optimal (dead-beat) controller component at time t = (n -1 -πη)

s = the order of the controlled system

m, = the number of control steps.

The result of the equation is an algorithm according to which the controller structure is determined only by the order of the controlled system and is independent of the number of control steps m. For a controlled system z. B. 2nd order contains the equation (1) only 5 links

a, · ν (η - 1), β ₂ · ν (η - 2), c ₀ · R (η - 1), c ₍ · R (η - 1 - mi), c ₂ · R (η - 1 - mi), ie m, = 2

Thus only once the dead-beat component R needs to be determined. This is then only used in a time-shifted manner according to the required control steps. In addition, the time-shifted controller output components y (n-1) and y (n-2) are required, which represent themselves as past values of the preceding manipulated variable y, and thus already exist as storage values. The number of required control steps is controlled in each case by the available actuator power. In comparison with the finely tuned control-optimized controller, it can be seen that in the case of noise optimization, the members of the dead-beat components have a variable offset by one cycle. If the actuator branching in the above example (m, - 2) is not sufficient, then at the 3rd term m, only m, + 1 must be increased. Thus, the previously common type of actuator limitation can be avoided. The determination of the actual m, value it takes place according to the equation

'' (M) <- ''., Ιηx < ^y (ii, - i) (2)

by comparing the calculated value with the maximum value y _m «. To determine the respective maximum available actuator value y _mix , which due to the respective load situation yi (t) does not exceed the permissible limiting value y _B ,, _r . must correspond, the respective current maximum output y _m " according to the equation

be determined.

To reduce the larger overshoot, due to the noise optimization, according to the figure, a setpoint filter is switched in front of the controller, the following filter equation is sufficient

· ^Ν 'ί' ^Ί); έ ^!: '·.'. V (H-Ih.) + Σ d..v7 (nm) < ⁴ >

i = l '· j ul J J

in which mean

w '(n-mi) = the filter output at sampling time t «= · (η-mi) T

b ,; dj = filter coefficients

w (nm _f ) = the filter input at the sampling time t = (n-rrijlT

s = the order of the controlled system

f = the order of the filter

m, = the number of control steps when optimizing for disturbance behavior (A controller optimization)

rrif = the number of control steps in the optimization of guidance behavior.

The result here too is an algorithm according to which the filter structure is determined only by the order of the controlled system and is independent of the number of control steps m. For a controlled system z. For example, 2nd order, the equation contains 2 terms for the first term and selectable _mr steps for the second term. It only needs the w 'component to be determined in 2 variants, which are then used according to the required control steps with a time delay.

In order to ensure a time-optimized filter according to the existing actuator output (y _max ), the required control steps are controlled by the available actuator output as with the controller.

It can be seen from the explanations that there are two degrees of freedom, which are characterized by the different parameters tn, andnrtf. You specify the number of setting steps for disturbance variable and command variable changes, and can be set independently of each other

 The invention is preferably intended for implementation with a microcomputer.

Claims (6)

1. A method for implementing an adaptive sampling controller with a constant sampling time and adjustable start-up time for control circuits with disturbing or disturbance and leadership behavior, characterized in that the controller design of the noise-optimized controller based on the found controller equation y ( ¹ ) = Z_ ^ ₁ .y (ii-i) + 2, C ₁ .R (nlr, i.), in the y (m) = the current controller output at sampling time t = η · Tmitm ₀ = O (T Ä sampling time) y (n-i) = the controller output at sampling time t = (n-i) T ai C | = Controller coefficients R (n - 1 - mi) = the time-optimal (dead beat) controller component at time t = T (n-1-rrij) s ~ the order of the controlled system m ₈ = the number of control steps is performed, after which a control algorithm is calculated, which according to the order (s) of the controlled system from the sum of a number of s evaluated controller output variables (y (n - i)) and the sum of a number of + 1 time-offset valued dead-beat components (R (n - 1 - rrij)), wherein the adjustability of the rise time (t _a ) of the disturbance is achieved by changing the number of control steps (m ₈ ). 2. The method according to claim 1, characterized in that the formation of the least number of control steps (m,), the current controller output variable (y (n)) with a maximum output (Ymax) is compared until the equation is satisfied. 3. The method according to claim 1 and 2, characterized in that for the realization of a temporally optimal sampling controller, the respective current maximum controller output variable (y _max ), which consists of the permissible limiting value (yeegr.) And the respective load value (yi. (T)) according to the equation / Υιι-χ / = / yeegr. - YL (t) / results is provided. 4. A method for realizing a good leadership behavior of the time-optimal noise behavior designed controller according to claim 1 by means of a setpoint filter, characterized in that the design of the filter based on the found filter equation _t S f v / '(n). · = 2Γ V'.v' (nm, _L ) + ΣΓ d .iv (n-rii,) in the w '(n) = the actual filter output at sampling time t = η -TmitTals j Sampling time I w '(n-nrii) = the filter output at sampling time t = (n-nri |) T bj; dj = filter coefficients w (n-rT) j) = the filter input at sampling time t = (n-mj) T s = the order of the controlled system f = the order of the filter m _s = the number of control steps in the optimization of disturbance behavior (ύ: controller optimization) • m _f = the number of (3 steps in the optimization to guide behavior takes place, according to which a filter algorithm is calculated, which, according to the order (s) of the controlled system, is the sum of a number of filter outputs (w '(n-mi)) evaluated in a time-offset manner and the sum of a number of filter inputs (w (n - mj)), wherein the tunability of the start-up time (t _a ) of the disturbance variable by changing the number of control steps (m _s ) at Störverhalten and the adjustability of the rise time (t _a ) of the reference variable by changing the number of control steps (m ₍ ) at (-Ü'nrungsverhalten is achieved. 5. Method according to claim 1 and 4, characterized in that the current controller output variable (y (n)) is used to form the lowest number of control steps (m ") in the case of disturbance behavior and the least number of control steps (m ₍ ) in the case of leadership behavior. is compared with a maximum output (Vmax) until the equation V (m,> <Ymax <Y (m, -1) is satisfied. 6. The method according to claim 1,4 and 5, characterized gekonnzeichnet that for the realization of a temporally optimal sampling controller for disturbance and leadership behavior, the respective current maximum Reglerausgangsgröße6 (y _max ) resulting from the permissible limiting value (veegr.) And the respective load value (y _L (t)) according to the equation / ymax / = / Vgr. ~ Yl (0 / results is provided.
For this 1 page drawing Field of application of the invention The invention relates to the realization of sampling regulators with constant sampling times and variable dynamics. Sense controllers with constant scan line are of great importance in the realization of control loops using microcomputers, since these micro rake controllers must process a specific calculation algorithm (controller equation) cyclically within a constant time interval T. Characteristic of the known technical solutions Every control loop is required to correct faults in the shortest possible time and respond optimally to a change in the reference variable. The shortest rise time is also dependent on the available Stellgiedleistung available at the time. In order to realize a time-optimal controller for small and large signal behavior with regard to the available actuator power, it is necessary to change the number of setting steps m with respect to the respective available actuator power. For this purpose, only controllers with finite response time, whose control algorithms are not only independent of the setup time, but also require a minimum of computational effort. A controller with finite response time, which corresponds to these conditions, is described in patent application DD-WP G 05 F / 2916163. However, this regulator has the disadvantage that it is only a time-optimal controller for the maximum performance of the available actuator power. For control loops which require a good disturbance behavior or a good disturbing and guiding behavior, this controller does not work in a time-optimized manner. Object of the invention The aim of the invention is to design a time-optimized sampling controller with a constant sampling frequency, which has a controller structure which ensures time-optimal behavior in the event of changes in the control and disturbance variables. Explanation of the essence of the invention The invention has for its object to find a method for the design of a sampling controller with a constant sampling frequency, the Ragelalgorithmus consists of a few members, which is without controller structure changes in a position to change the number of control steps so that when using the available actuator power, the smallest number of control steps for off and / or adjustment, regardless of the control and disturbance variable behavior, is required. This object is achieved in that the controller design based on the found controller equation