RU2023303C1 - Device for information understanding - Google Patents

Abstract

FIELD: automatics, computer engineering. SUBSTANCE: device is designed for determination of technical condition of industrial articles. It has input amplifier 1, comparators 2,3, unit 4 of setting of levels, AND gates 5, 6, pulse generator 7, integrator 8 and unit 9 of setting of time intervals. Due to insertion of delay elements 10, 11 device digitizes not difference time interval t₁-t₂, where t₁ and t₂ are times of stay of signal accordingly above low level and below high level but these intervals t₁ and t₂ themselves. EFFECT: increased precision of understanding due to reduced relative error of digitization of input signals. 2 dwg

Description

 The invention relates to automation and computer technology and can be used to determine the technical condition of industrial products.

 Known amplitude analyzer [1], performing a similar function. It contains a polytron type cathode ray tube, a sawtooth generator, an integrator, an amplifier and an adder, two inputs of which are connected to a sawtooth voltage generator and amplifier, and the output - with a horizontally-deflecting cathode ray tube - a polytron, and an integrator, connected to tube manifolds.

 The disadvantages of this device are the low accuracy and complexity.

 A device for recognizing information [2] is known, comprising an input amplifier, a weighting unit, an adaptation unit, and an integrator.

 The block of weight coefficients is made on a polytron type cathode ray tube, the horizontal deflecting plates of which are connected to the input amplifier, the collectors to the integrator, and the functional plates to the adaptation block.

 The signal x (t) is amplified by the input amplifier to the range of the polytron scan and is fed to horizontally deflecting plates. The converted signal from one of the collectors is fed to the integrator, by the output value S of which obtained during the integration, it is possible to judge the signal class number x (t). The conversion operator is stored in the form of voltages on the functional plates using memory integrators. The adaptation of the device consists in the selection of such a conversion operator that provides the correct classification of signals. As signs, the values of the signal residence time at different amplitude levels are used. To determine the relative residence time of the signal in the corresponding amplitude interval, it is necessary to divide the signal into intervals and calculate the time of its stay in each of the intervals.

 This device simplifies the design and increases its reliability. However, it does not provide a sufficiently high recognition accuracy.

 Closest to the proposed is a device for recognizing information containing an input amplifier, the input of which is the input of the device, and the output is connected to the first inputs of two comparators, the second inputs of which are connected to the outputs of the level setting unit, a pulse generator, the output of which is connected to the first input of the first element And, the second input of which is connected to the output of the first comparator, and the output to the summing input of the integrator, the control input of which is connected to the output of the unit for setting time intervals, and the output is the output of the device, the second element And [3].

 The disadvantage of this device is the low recognition accuracy associated with increased relative sampling error of the input signals. In this device, part of the signal is directly sampled within the lower and upper threshold levels of operation of the comparators. Due to the smallness of this part of the signal, the relative sampling error is large.

 The aim of the invention is to increase the accuracy of recognition by reducing the relative sampling error of the input signals.

 Figure 1 presents a block diagram of a device for recognizing information; figure 2 is a timing diagram of the operation of the device.

 The device comprises an input amplifier 1, first 2 and second 3 comparators, a unit for setting levels 4, a first 5 and a second 6 elements And, a pulse generator 7, an integrator 8, a unit for setting time intervals 9 and the first 10 and second 11 delay elements.

 The device operates as follows.

The electric signal arriving at the input of the device after amplification by the amplifier 1 is fed to the non-inverting inputs of the comparators 2 and 3. Upper U _p.v. and lower U _n.p. the thresholds for comparators 2 and 3 are set by the unit for setting levels 4, from the output of which constant voltages corresponding to the given thresholds for the operation of comparators are supplied to their inverting inputs.

From the output of the first comparator 2, a sequence of rectangular-shaped positive pulses is taken, the duration of which corresponds to the time the signal spent above the lower threshold level U _bp . At the time of opening of element I 5 by a pulse train, the pulses of the counting frequency passing through it go to the direct input of the integrator 8 The number of counting pulses N ₁ corresponds to the time t _{1 the} signal stays above the lower threshold level U _n.p. .

, где Т - период счетных импульсов;
t_зад.1 и t_зад.2 - соответственно времена задержек элементов 10 и 11.From the output of the second comparator 3 is removed a sequence of positive pulses of rectangular shape, the length of which correspond to the time spent above the upper threshold signal U _ae .. shifted second element delay time equal to half the period of counting pulses, this sequence is supplied to one input of AND 6 , to the other input of which counting pulses arrive, shifted by the first delay element by the same amount. The delay times of elements 10 and 11 are the same:
t _ass. 1 = t _ass . ₂ =

where T is the period of the counting pulses;
t _{ass. 1} and t _{ass. 2} - respectively, the delay times of elements 10 and 11.

The shifted counting pulses passing through the open element And 6 are fed to the reverse input of the integrator 8. As a result, the reading N of the latter will correspond to the difference in the signal residence time above the lower t ₁ and upper t ₂ threshold levels
N = f (t ₁ -t ₂ ).

Thus, a nonlinear conversion of the input signal by the expression
y (t) = x (t) Φ (x), where x (t) is the input signal.

y (t) is the output signal;
Ф (х) = 1 for U _a.s. ≅ x ≅ U _a.p. nonlinear transformation operator;
Ф (х) = 0 - in other cases.

 A description of the input signal is generated based on its time spent within a given amplitude level, i.e. the signal is converted into a value proportional to the distribution density of the instantaneous values of its amplitude. The operating time of the integrator 8 is determined by block 1 of the job time intervals 9.

 The output value uniquely corresponds to the shape of the recognized signal and is used in the future as its feature when compared with other analyzed signals.

· 100%
В предлагаемом устройстве дискретизации подвергаются сами интервалы t₁ и t₂, максимальные относительные погрешности которых определяются
δt_1.max=

· 100% и δt_2.max=

· 100%
Максимальное значение относительной погрешности дискретизации временного интервала t_п=t₁-t₂ будет равно наибольшей из погрешностей δ t_1max и δ t_2max. Поскольку t₂>t_п и t₁>t_п, то результирующая погрешность дискретизации в предлагаемом устройстве будет намного меньше, чем в прототипе. Благодаря этому по сравнению с базовым объектом в предлагаемом устройстве повысится точность распознавания информации.Improving the accuracy of the proposed device in comparison with the prototype is as follows. As is known, in the prototype, the time interval t _{p is} sampled, which is formed as the difference in the signal residence time above the lower level t ₁ and below the upper level t ₂ : t _p = t ₁ -t ₂ . The maximum value of the relative sampling error of this interval is
δt _p.max =

· 100%
In the proposed sampling device, the intervals t ₁ and t ₂ themselves are subjected, the maximum relative errors of which are determined
δt _1.max =

100% and δt _2.max =

· 100%
The maximum value of the relative discretization error of the time interval t _p = t ₁ -t ₂ will be equal to the largest of the errors δ t _1max and δ t _2max . Since t ₂ > t _p and t ₁ > t _p , the resulting sampling error in the proposed device will be much less than in the prototype. Due to this, compared with the base object in the proposed device will increase the accuracy of recognition of information.

Claims (1)

 A DEVICE FOR RECOGNITION OF INFORMATION, containing an input amplifier, the input of which is the input of the device, and the output is connected to the first inputs of two comparators, the second inputs of which are connected to the outputs of the level setting unit, a pulse generator, the output of which is connected to the first input of the first element And, the second input of which connected to the output of the first comparator, and the output to the summing input of the integrator, the control input of which is connected to the output of the unit for setting time intervals, and the output is the output of the device, second element I, characterized in that, in order to increase recognition accuracy by reducing the relative sampling error of the input signals, two delay elements are introduced into the device, the inputs of which are connected respectively to the outputs of the second comparator and pulse generator, and the outputs to the inputs of the second element And whose output is connected to the subtracting input of the integrator.

Images