SU1043593A1 - Non-linear converter for variable structure systems - Google Patents

Abstract

NONLINEAR CONVERTER FOR SYSTEMS WITH A VARIABLE STRUCTURE, containing a series-connected inertia unit, comparing the element and a switchable element, connected by an output with the input of the inertia unit and with the input of the first. low pass filter, characterized in that, in order to increase accuracy, speed and reduce power consumption of the converter, it contains a second low pass filter connected by an input to the output of the switchable element and the output to a control input of the switchable element made in the form of a three-position relay element displaced along the ordinate axis by a linear zone of static characterization .... g.

Description

with

C71

CP

00

./

The invention relates to self-adjusting systems and can be used for variable structure systems, as well as for transforming object state variables and estimating its phase coordinates. five

Nonlinear transducers are known that contain a two-position relay element (ER) covered by inertial negative feedback (OOS). Depending on the purpose of the jQ device as inputs and. Outputs: Different circuits of the specified circuit are used: in the differentiating device, the output element is a low-pass filter (LF), 5 in the nonlinear filter, the output of the device is the OOS output Cl and C2 3.

In these, as well as in other similar devices, the main means of converting information is a sliding mode generated by an error signal by a switchable element. The accuracy of the information conversion depends on the pass band7 and the OM, i.e. from the degree of approximation of the real sliding mode to the ideal one, which theoretically determines. achievable accuracy. The bandwidth of the ER depends on its own. . nonidealities (hysteresis, retardation, dynamic nonidealities), 30 which determine the maximum possible sliding mode frequencies for each particular construction.

However, at low frequencies, The variable modes are increased by a variable Ze component of the output signal. . at the output of filters installed after the OM, or an increase in their constant time is required. In addition, an increase in the frequency of gliding modes with an increase in the energy consumption of an electronic element due to dynamic losses during switching.

The closest technical solution to the invention is a non-linear transducer for variable structure systems containing inertia bpok connected in series, comparing element and switchable element connected by output to the input of the inertial unit and to the input of the first low-pass filter C3 J .

A disadvantage of the known device is either the need for long time constants when filtering the output signal of the operating element at low frequencies of the sliding mode, or an increase in the dynamic losses with their increase.

The purpose of the invention is to increase the accuracy, performance and reduction of energy consumption of the conversion.

 The goal is achieved by the fact that the converter contains a second low-pass filter connected by the input

to the output of the switchable element, and the output to the control input of the switchable element, made in the form of a three-position relay element with a linear zone shifted along the linear axis of the static, characteristic.

FIG. 1 is a block diagram of a non-linear converter; FIG. 2 and 3 are the same, examples are used and inventions respectively in a non-linear filter and a division unit; in fig. 4 - static characteristics of the switchable element; in fig. 5 is an example of an electrical circuit diagram of a switchable element.

The block diagram (Figs. 1-3) contains a non-linear converter 1, switchable element 2, inertial unit 3, first low-pass filter 4 (LPF), comparing element 5, second low-pass filter b (LPF), a two-stage relay element 7 (OM). .. FIG. 4 denotes the input 5 and output and the signals of the switching element 2, the relay sections a and b of the characteristics., The displaceable linear-zone with the static characteristic, the control signal Uyrip.

The switchable element contains (Fig. 5) an operational amplifier (OU) 8 a circuit of an auto-switched OOS 9, resistors 10-15, a switching element - .16 and an inverter 17.. . .

The circuit shown in FIG. 3, differs from the diagrams (figs. 1 and 2) by the presence of a two-position electronic device 7 connected in series between elements 2 and 3. As the device (fig. 3) performs the operation of dividing (i.e., Z i x / y), the control input RE 7 is the input of signal y,

The switchable element (FIG. 5) is an op-amp 8 covered by an auto-switched (via switching element 16) OOS 9 and resistors 10 and 11. The output of the switching element, which when used by diodes is simultaneously its control input, is connected through a divider ( resistors 13 and 14) with a non-inverting input of the OA 8 and through a resistor 15 with the control input of the switching element 2, and the output of the OA 8 is connected via an inverter 17 s, the output of the switching element 2.

The principle of the switchable element 2.

With a power amplifier 8, covered by an auto-mutable OOS 9 lm, a relay signal is generated if (Fig. 4), where 5 is a constant value, and a line signal if ISK S. A signal is formed by a relay if .. switching element 16 passes the signal (one of the diodes is open), i.e. the OOS circuit is broken and switchable element 2 becomes a voltage comparator. The trigger level I (S value) depends on the ratio of resistors 10 and 12 and on the magnitude of the signal at the output of element 16, moreover, the trigger level; does not depend on the magnitude of the control voltage, jpp, since it is fed simultaneously through the appropriate voltage divider (resistors 13 and 14) to the non-inverting input of the OU 8. The magnitude of the voltage Uynp determines the displacement of the static static characteristic from within the dead zone defined by tioi positioning the relay sections ot and b. x characteristics and f (S, and Resistor 15) is used to form a small positive pattern that contributes to the accelerated transition from section c to sections d and b of the static characteristic. The converter for converting the input signal S of the switchable element 2 into its output signal is: O, with S: 5, control Kj, with IS I d, with S 5, is the dead zone of the relay characteristic of element 2 (Fig. 4)} UQ - amplitude of the relay signal determined by the saturation levels of link 9; static coefficient of transfer of the linear characteristic / coefficient of transmission over the control path (signal gain Uynp). Nonlinear Converter works as follows. If S is 0, i.e. as a switchable element, a two-position re is used, and the use of link b is not required; the device (Fig. 1) is a differentiating device known as C1 and Cz. Principle dy-. This is based on the fact that, if the conditions of the sliding regime are 1Im § 0 {1im S O, S, the value of S is close to zero. The output signal of the integrator 3 is close in magnitude to the signal X, i.e. . Consequently, the constant component (0.8 - in the terminology of the known ZZ relay signal U is close. To the sigial X. In an ideal sliding mode, Fri Z.-X (with), g "and S -Si. If the switchable element 2 has a static characteristic (Fig. 4), then the sliding mode can occur if conditions are met: a) Urn S O, S S o; S. 5) j,. ; where S and S are the positions of the relay sections “and b are the characteristics of U.“ f (S, u ,, sp). In other words, if the corresponding conditions are fulfilled, a sliding mode can arise both at section c and at section b of the relay characteristic. For example, at some moment t, at which the output signals of links 3, 4 and b (Fig. 1) are equal to zero, a signal of the form X -, et begins to be input to the converter, where t is a constant value (f UQ), When when the signal x reaches the border S. static characteristic and f (S, U., rtp), in the contour of their elements 2-3-5 (figure 5) a sliding mode arises, with a slope § -0-, S - X - ( v and, therefore, and - "B (Fig. 4b). A signal appears at the output of the device. Where D. (p) is the variable component of the signal, the amplitude of which depends on the constant time T of the link 4, on the frequency of change U and from its amplitude A (Fig. 4b). As the signal increases, the amplitude A of the signal oscillations U decreases due to the shift of the section from the static characteristic to the value (Fig. 46). At А &} Q U & K8 the mode stops and the signal S begins to decrease (as,), i.e. the operating point of the circuit 2-3-5 is determined by the linear characteristic, in which case the variable component of the signal disappears and, which has the form -, - Ki:; and "With an appropriate choice of K and T values, it is easy to achieve that i X J. At the same time, the output signal z has Tp T 1, i.e. in the -stable signal there is no variable component, and in transient modes it is less than in known circuits. In this case, in connection with a decrease in the level of the signal ripple, the time constant of the low-pass filter 4 can be reduced. This is due to the fact that the amplitude (corresponding to the double amplitude of the signal U) is always less than the double amplitude of the signal of the two-position electronic device, i.e. A 2U. It should be noted that the frequency of oscillations of the signal U depends little on the magnitude of U0J, since the first is mainly determined by the dynamic capabilities of the OM and it cannot be significantly increased in order to reduce the signal pulsations -z due to an increase in the UV saturation of the OM (signal U If during the design of the circuit it turns out that the time constants T and C (links 4 and 6) are close in magnitude, you can use the output sig in this case, Z, as the Uypp signal, i.e. required, and the control input of the switchable element is not osredstvenno coupled to yield tron devices, all above arguments fully refer to the diagrams in Figures 2 and 3, with the following refinements.

.J

9Ve.2 When the occurrence of a glide sim is in a nonlinear filter (Fig. 2), Ug ST or Uj. -S, (xo; oJt. |. Consequently, UU, i.e., the low-frequency signal x is little distorted by the filter. In the division block: (Fig. -3), the mathematical description, which is generally given in the well known SzL, always exists Sliding mode, formed by RE 7, in transient modes - switchable element 2 .. In steady-state mode, signal 1Г to ensure switching of RE 7 should be close to zero, therefore, to ensure that signal S is close to zero, the K-factor 4 must be greater than 1. From the description given, it follows that the nonlinear converter for ATP provides compared with the known higher accuracy of conversion (due to the reduction of the ripple signal U), increased fast-response (due to the decrease of the inertia of the first low-pass filter), lower power consumption (due to the reduction of dynamic losses in the switchable element). In addition, the proposed The device is a source of interference of lower intensity (due to a decrease in the amplitude of the signal and) in comparison with the known ones. Switch item when used in. These devices have a lower ripple output voltage.

“) Uyfip O

S)

FIG.

us.S i /

Claims (1)

Nonlinear converter FOR SYSTEMS WITH VARIABLE STRUCTURE, containing inertial block connected in series ··, comparing the element and the switched element connected by the output to the input of the inertial block and to the input of the first. low-pass filter, characterized in that, in order to improve the accuracy, speed and reduce the power consumption of the converter, it contains a second low-pass filter connected to the output of the switched element by the input and the output to the control input of the switched element, made in the form of a three-position relay / element with the linear area of the static characteristic displaced along the ordinate axis. SU 1043593