# Digital predicting and differentiating device

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to automatic and computer equipment and can be used to predict stationary and non-stationary random processes, increasing quality and accuracy of control in digital systems for controlling and guiding various objects. The prediction unit includes a first derivative estimation unit. The estimation unit has three subunits for calculating first derivatives. The first subunit, which is meant for the first n-th (current) reference point of the history of the predicted process, consists of one adder. The second subunit, which is meant for the second (n-1)-th history reference point, consists of one adder. The third subunit, which is meant for the third (n-2)-th history reference point, consists of two adders and an inverter unit. Outputs of the subunits are the data outputs of the device.

EFFECT: broader functional capabilities by obtaining first derivative estimates for all equidistant points (nodes) of the two-level history of an input smoothed discrete process.

The invention relates to automation and computer engineering and can be used for forecasting of stationary and nonstationary random processes, improve the quality and accuracy control in digital systems of control and guidance of various (including ballistic objects.

A device for adaptive extrapolation (prediction) (ed. St. USSR №1246775, CL G06F 15/353, 1984)containing the block smoothing unit extrapolation (prediction), containing three myCitadel, the address register y (reference points) the prehistory of the input process, specifying the time of the forecast, and the output of adder evaluating quadratic forecast. The device has limited functionality.

The closest to the technical nature of the claimed device is selected as a prototype device for adaptive extrapolation (prediction) (ed. St. USSR №1572281, CL G06F 15/353, 1988)containing the block smoothing unit extrapolation (prediction), which includes three serially connected myCitadel, the register set a time of the forecast, the output of adder evaluating quadratic prediction and evaluation unit of the first derivative of the three adders. The latter gives the value of the first derivative only in the first n-th of the four reference points (nodes) three-level background predictable process, and at four the ferry for the third (n-2)-th design point (B.P. Demidovich and Maron I.A. Foundations of computational mathematics. M., "FM", 1960, .XV, § 4, str).

The technical problem for the proposed device is to expand its functionality by obtaining estimates of the first derivatives for all three equally-spaced points (nodes) duplex prehistory input smoothed discrete process.

Therefore, in the digital prediction and differentiating device, comprising: a block smoothing (ed. St. USSR №1531808, CL NN 17/04, 1988)containing the adder, the first and second reversible counters, single channel subunit smoothing (ed. St. USSR №748417, CL G06F 15/32, 1980) of the serially connected adder and a register, the sub job ratio deviations, containing a register, a counter and a delay element, a subunit of the actual deviations of the containing block inverters, two Comparators and the element And the sub unit increments containing two elements And the inverter, the sub control dynamic response, contains two pulse shaper element OR the counter, the three elements and the trigger mode, the information output unit smoothing information, a first control and pulsing the input device; node clocking unit forecast containing a delay element, trigger the pulse generator, the element And the shift register; the unit is the forecast, containing the first and second myCitadel, each of which contains a block of register memory, multiplexer, block inverter and adder, sub quadratic forecast, containing a block of inverters, the first and second adders, and the sign of the second term of the first adder through the power of the inverters connected to the output of the second myCitadel, the sign of the first term of the first adder to the information output unit smoothing the output of the first adder connected to the input of the second term of the second adder subunit, the sign of the first term of which is connected to the output of the first myCitadel, the output of the second adder and the subunit is the first information output device, a subunit of the linear prediction containing adder and the unit inverters, the input of the first addend adder connected to the output of the first myCitadel, and the sign of the second term through the power inverter to the output of the first adder subunit quadratic forecast, and the input bus of the second addend adder subunit linear prediction opened with mounting shifts by one bit in the direction of its least significant bits, the output of the adder and the subunit is the second information output device; a control unit dynamics forecast containing the register storing the address of y (reference points) the prehistory of the input process, the entrance of which t is aetsa second managing input device, specifies the time (interval) of the forecast, a comparator, an inverter element And the address counter, the information output of which is wound on the address bus multiplexers both vychitala and to the first input of the comparator, the second input is connected to the output of storage register addresses, the output of the comparator through an inverter connected to the first input element And the second input is connected to the output of the third element And subunit management dynamic characteristic (SOUTH) block smoothing, the output element And is connected to a counter input of the address counter, the bus is reset to "0" which is connected to the output of the second element AND SOUTH block smoothing, to solve the problem in the block prediction introduced the evaluation unit of the first derivatives containing the first subunit of the calculation of the first derivative in the first to n-th (current) point design background predictable process from one adder, in which the first input element connected to the output of the unit inverters of the first myCitadel, the sign of the second term to the output subunit quadratic forecast, and the output of the adder and the subunit is the third information output device, the second subunit of the calculation of the first derivative in the second (n-1)-th point design history from one adder, in which the first term is connected with the information output block smoothing, which is administered the second summand is unit inverters of the second myCitadel, and the output of the adder and the subunit is the fourth information output device and the third subunit of the calculation of the first derivative in the third (n-2)-th point design history of the two adders and block inverters, in which the sign of the first term of the first adder connected to the output of the block smoothing, the sign of the second term to the output multiplexer of the second myCitadel, the output of the first adder through the power inverter is pulled to the entrance of the second term of the second adder subunit, the sign of the first term of which is connected to the output of the second myCitadel, and with a shift input tyres summand on one digit to the side high-order bits of the second adder, the output of which is the output of the subblock and the fifth information output device.

The invention is illustrated by drawings, which depict: figure 1 - block diagram of the proposed device; figure 2 - block diagram block smoothing; figure 3 - block diagram single channel subunit smoothing; figure 4 - block diagram of the node clocking block prediction; figure 5 - block diagram block prediction and evaluation unit of the first derivatives.

Known formulas operators forecast obtained analytically using the approximating polynomials by three points (the ordinate) the prehistory of the input discrete random process is CA by the method of least squares (Milne WE The numerical analysis. M., "IL", 1951, str). The operator forecast on approximarely the second order polynomial (quadratic) three-point history is

The operator forecast on approximarely the first-degree polynomial (linear) three-point history is

or its simplified version

where_{p}the first (current) design point (ordinate);

the_{p-1},_{p-2}accordingly, the second and the third reference point (ordinate) duplex prehistory input smoothed discrete sequence. In the numerical analysis is a system of equally spaced points with step h, in real-time h - point spacing (y), i.e. the time (depth) of the projection (H).

Denote Δ_{1}=(2U_{p}-u_{p-1}as birdnest first level of background input discrete sequence, i.e. the difference between twice the current and previous ordinate process, respectively, Δ_{2}=(2U_{p-1}-u_{p-2}- birdnest the second level of the background.

After modifying equations (1) and (3) subject to birdnote get the following expressions for quadratic operators [QTR3] and linear [LN] forecasts that are implemented in the proposed device:

where Z=y_{p}-Δ_{2}and

Known formulas numerical differentiation for equally spaced points, expressed through the function values at these points (Gollard. Handbook of computational methods in statistics. M., "FS", 1982, § 6.6, p.61), in particular for three points have

Below is a table of correspondence numbering reference points (nodes) background numbering of points in the original sources.

The number of design points in prehistory | 3 | 2 | 1 | |

the_{n-i} | the_{n-2} | the_{n-1} | the_{n} | the_{n+1} |

the_{i} | the_{0} | the_{1} | the_{2} | |

f_{i} | f_{-1} | f_{0} | f_{1} |

P the following conversion equations (6), (7) and (8) subject to birdnote get the following expressions for the calculation of the first derivatives:

The implementation of equations (9), (10) and (11) in the proposed device meets the task, and the main typical elements become the adder and the unit inverters, and the multiplication coefficients of the components is appropriate mounting shifts tires last at the input to the adder. Such operations on the block diagram (see figure 5) indicated by the circle.

The device comprises (see figure 1) block smoothing 1, block prediction 2 and the evaluation unit of the first derivative 3. Block smoothing 1 (see ed. St. USSR №1531808, CL NN 17/04, 1988) contains (see figure 2) adder 4, sub 5 actual variance containing block of the inverter 6, the two comparator 7.1 and 7.2 and the element And 8, the first reversible counter 9, sub 10 jobs ratio deviations containing the register 11, the counter 12 and the delay element 13, the sub unit 14 increments containing the inverter 15 and the two elements And 16.1 and 16.2, the second reversible counter 17, the subunit 18 control of dynamic characteristics, contains two pulse shaper 19.1 and 19.2, item, OR 20, the counter 21, the three elements 22.1, 22.2, 22.3 and trigger mode 23; information input 24 of the block and smoothing device, the first control is managing 25 and pulsing 26 inputs the device and block smoothing; single channel subunit smoothing 27 (see ed. St. USSR №748417, CL G06F 15/32, 1980), contains (see figure 3) connected in series adder 28 and the register 29; information output 30. Node clocking block 31 forecast contains (see figure 4) delay element 32, the trigger 33, the pulse generator 34, the And gate 35 and the shift register 36. The block prediction 2 (see figure 5) contains the first 37 and second 38 myCitadel, each of which contains a block of register memory 39 from (A) series-connected registers 40, the multiplexer 41, the unit inverters 42 (assuming that the multiplexer does not have inverted outputs) and the adder 43; sub 44 quadratic forecast, containing a block of inverters 45, the first 46 and second 47 adders, the output 48 of the latter is information access subunit; subunit 49 linear prediction, containing a block of inverters 50 and the adder 51, the output 52 of the latter is information access subunit; site 53 dynamics control forecast containing the register 54 storage address (A) ordinate reference points the prehistory of the process, the inlet 55 which is the second managing input device that specifies the time of the forecast H=at (T - cycle operation, And the maximum address of the memory register 40 block 39), the comparator 56, the inverter 57, the And gate 58 and the address counter 59. Unit 3 estimates of the first derivatives contains the first sub 60 calculation is first carried the Noi in the first to n-th point design background process from one adder 61, the output 62 which is the third information output device, the second subunit 63 calculation of the first derivative in the second (n-1)-th point design background from one of the adder 64, the output 65 which is the fourth information output device, the third subunit 66 calculation of the first derivative in the third (n-2)-th point design background, which consists of the first adder 67, the unit inverters 68 and the second adder 69, the output 70 which is the fifth information output device.

The work cycle consists of 3 cycles. Block smoothing 1 operates in two modes: fixed and dynamic (transient), and all operations are performed within one (1) time.

In stationary mode, the unit smoothes the input discrete random process, a deterministic basis (median) which can be constant, linear or non-linear (quadratic) character changes over time. Block smoothing 1 (see figure 2) implements the following modification of the operator's signature exponential smoothing:

where x_{n}and I_{n}- input and output samples;

α=1/K is the smoothing parameter, the parameter adaptation;

Δx_{n}=(x_{n}-u_{n-1}) - the current deviation from the median of the process.

As a criterion of effectiveness (accuracy) of smoothing chosen what about the ratio d between zero and the actual deviations Δx_{
n}. The last form current single increment of either sign output samples in accordance with the signature function in (12):

sign[Δx_{n}/K]=0 for [Δx_{n}-To]<0 (Δx_{n}- zero variance),

sign[Δx_{n}/K]=1 [Δx_{n}-To]>0 (Δx_{n}- actual deviation).

In stationary mode (D=0 - sign mode) block 1 smoothes the input random sequence increments to the level specified by the value d (the real range d=7÷190), which is filled in before you begin with the entrance 25 to the register 11 subunit 10 set the ratio of the variance. The latter is a controlled frequency divider, for example, for d=7 the output direct transfer counter 12 appears every seventh clock pulse from input 26, and, through a delay element 13 overwrites the inverse code d register 11 in the counter 12 (for the next cycle of operation of the divider), and subtracts "1" from the first reversible counter 9 that contains the code of the parameter adaptation K.

Adaptive control smoothing parameter, providing a constant output values of the variance of the smoothed process regardless of the degree of variability in the input, as follows. Let (at a certain code To the counter 9) increased the variance of the input discrete process, i.e. increased the number of valid Δx_{
n}(both characters). After comparing them with the option of adapting To output blocks of Comparators 7.1 and 7.2 subunit 5 (playing the role of negative feedback) are logic "1" (mode Comparators: [Δx_{n}>]="1", [Δx_{n}<K]="0")received at the input element And 8. As in the stationary mode the trigger mode 23 is in the state "0" (D=0), with its inverted output to the first input element And 8 subunit 5 also receives a logical "1". High level signals at all inputs of the element And 8 permit the passage of clock pulses from input 26 to a summing input of the first reversible counter 9 (code To increases in the last) and second input elements And 16.1 and 16.2 subunit unit 14 increments. The output signal from one of them (depending on the sign of the deviation) is fed to a summing (or subtractive) the input of the second reversible counter 17 of the smoothing ie signature function (12). The process of growth will reduce the number of valid variance and will continue until, until dynamic equilibrium, i.e. the number of pulses received from the sub 10 on the subtractive input of the counter 9 will be equal to the number of pulses received at its summing input subunit 5, and the output variance of the smoothed discrete sequence are similar the th (d=7: one valid deviation should have seven zero).

Transient (dynamic) mode may be caused by the acceleration curve, the transition from one mode to another and so on, almost abrupt change process. For smoothing discrete input sequence on transient (dynamic) mode (D=1) uses a single channel subunit smoothing 27 (see figure 3), which implements the following statement exponential smoothing:_{n}=1/2(x_{n}+_{n-1}), i.e. with a minimum degree of smoothing and with minimal phase shift (delay) the output samples. Sub 27 works in both modes is triggered by clock pulses from input 26 in register 29, but is used only for transient (dynamic) mode. For a stationary random process, the probability of occurrence of a series, for example, m=8 (eight) deviations from the median (deterministic framework) process is one sign the contract in accordance with the geometric probability distribution is equal to

P(x=m)=(1/2)^{m}=1/256≈0,004,

i.e. so small that it can be considered as the appearance of such series by the beginning of the transition mode. The subunit 18 secures such series and works as follows. As for the stationary mode, the most probable deviation of different signs, then the sign changes in the adder 4 with a "plus" to "minus" and Vice versa trigger formation of the fir pulse 19.1 or 19.2 and through the element OR 20 reset to "0" the counter 21 and the trigger mode 23 (D=0). In dynamic mode (the formers 19 do not work) on the counter 21 (for example, 4-bit) will receive eight consecutive pulses from clock input 26. Output high-order bit of the counter 21 will set the logic "1", a high level signal which will ensure passing through the first element And 22.1 pulsing pulse, which sets the trigger mode 23 1 (D=1). The last signal from the inverted output will block the operation of subunit 5 of the actual variances and, respectively, subunit 14 single increments, and a high level signal direct access will be allowed through the second element And 22.2 overwrite discretes single subunit smoothing 27 in the second reversible counter 17 of the anti-aliasing. At the end of the transitional regime in the adder 4 will inevitably arise deviations of different signs, which will lead to the triggering of the pulse shapers 19 and, respectively, to switch the trigger mode 23 in the state "0" (fixed smoothing mode, D=0).

Operations forecasting and calculation of the first derivatives are performed for two bars, respectively, the 2nd and 3rd. They are formed by a series of two pulsing pulses from node clocking 31 (see figure 4). Pulsing the pulse input 26 clears the trigger 33, and writes "1" in the least significant bit of the shift register 36. The same pulsing pulse zaderzhan the th delay element 32 sets "1" trigger 33,
thus allowing the passage of pulses from generator 34 through the element And 35 in the shift register 36, the tyres least significant bits ("a" and "b") and you receive the above series. In the 2nd step of recording the output current (first) design points in_{p}in the first register 40 of block 39 of register memory of the first vicites 37. At the same time overwrites (shift) all the preceding y in adjacent registers 40 (i.e. a background of the input process). The address input of multiplexer 41 receives address code (A), the ordinate history from the address counter 59, which is equal to the code address recorded with the second control input 55 to the register 54 storage addresses before you begin device, and specifying the time (interval) prediction N=AT. In accordance with this address, the y output of multiplexer 41 (as the second design point y_{p-1}) through the power of the inverter 42 is fed to the input of the second term of the adder 43, at the entrance of the first term of which is a double ordinate of the previous design point y_{p}. At the output of the adder 43 of the first vicites 37 is set birdnest 1 level of background input discrete sequence. In the 3rd quantum operations similar to those described above, but for the second 38 myCitadel, the output of which sets birdnest 2-g the level of background.
All the adders in the device - Raman. Upon completion of the 3rd beat of the output subunit 44 in accordance with the formula (4) code set evaluation quadratic (non-linear) prediction for nonstationary discrete input sequence, the output 52 49 sub in accordance with formula (5) code of estimating linear prediction for stationary or slowly changing input discrete sequence, and outputs 62, 65 and 70 unit 3 estimates of the first derivatives in accordance with formula (9), (10) and (11) - codes of these estimates.

Node dynamics control forecast 53 is intended to exclude a prediction operation on a dynamic (transient) mode (D=1) operation of the device, by resetting to "0" of the address counter 59 pulsing signal (U_{0}from subunit 18 to control the dynamic characteristics of the block smoothing. Zero address counter 59 to address the tyres multiplexers 41 both vychitala will provide during the transition mode, the calculation and setting outputs 48 and 52 of both subunits forecast 44 and 49 of the code of the current increments y_{n}the minimum of the smoothed input process. With the transition of the device in the stationary mode (D=0) the trigger mode 23 sub 18 will permit the passage of pulsing pulse from input 26 through the open element And 58 at the counting input (U_{a}) of the address counter 59. With the growth of the addresses in the past,the outputs of both subunits forecast 48 and 52 are mounted codes (from_{
n+1}predicted input process using information from two background only for new stationary mode. Growth code address h in the counter 59 (h=aT, a=1,2,3,...), i.e. the recovery of a specified prediction time H will continue with each cycle until, until it becomes equal to the storage register 54 h=H. the Comparator 56 (mode: [H=h]→"1", [H≠h]→"0") in this case through the key 57 and the element And 58 will close the counting input of the address counter 59.

The proposed device has one myCitadel smaller than the prototype, and the introduction to the operation unit calculating the first derivatives in the first (n)-th second (n-1)-th and in the third (n-2)-th reference points background (i.e. separated by time) significantly extends the functionality and scope of the device, it is possible to analyse the nature of (trend) changes parameters of the predictable process (increase - decrease, acceleration - deceleration, and so on), to assess its intensity and other

Digital prediction and differentiating device, comprising: a block smoothing containing the adder, the first and second reversible counters, single channel subunit smoothing of the series-connected adder and a register, the sub job ratio deviations, containing a register, a counter and a delay element, sub the Lok valid variance, containing a block of inverters, two Comparators and the element And the sub unit increments containing two elements And the inverter, the sub control dynamic response, contains two pulse shaper element OR the counter, the three elements and the trigger mode, the information output unit smoothing information, a first control and pulsing the input device; node clocking unit forecast containing a delay element, a trigger pulse generator, the element And the shift register; block prediction, containing the first and second myCitadel, each of which contains a block of register memory, multiplexer, block inverter and adder the subunit quadratic forecast, containing a block of inverters, the first and second adders, and the sign of the second term of the first adder through the power of the inverters connected to the output of the second myCitadel, the sign of the first term of the first adder to the information output unit smoothing the output of the first adder connected to the input of the second term of the second adder subunit, the sign of the first term of which is connected to the output of the first myCitadel, the output of the second adder and the subunit is the first information output device, a subunit of the linear forecast containing the adder and the unit inverters, the input of the first addend adder connected to output the control of the first myCitadel, and the sign of the second term through the power inverter to the output of the first adder subunit quadratic forecast, and the input bus of the second addend adder subunit linear prediction opened with mounting shifts by one bit in the direction of its least significant bits, the output of the adder and the subunit is the second information output device; a control unit dynamics forecast containing the register storing the address of y (reference points) the prehistory of the input process, the input of which is the second managing input device that specifies the time (interval) of the forecast, a comparator, an inverter element And the address counter, the information output of which is wound on the address bus multiplexers both vychitala and to the first input of the comparator, the second input is connected to the output of storage register addresses, the output of the comparator through an inverter connected to the first input element And the second input is connected to the output of the third element And subunit management dynamic characteristic (SOUTH) block smoothing, the output element And is connected to a counter input of the address counter, the bus is reset to "0" which is connected to the output of the second element AND SOUTH block smoothing, characterized in that the block prediction introduced the evaluation unit of the first derivatives containing the first subunit of the calculation of the first derivative in the first to n-th current) point design background predictable process from one adder, in which the first input element connected to the output of the unit inverters of the first myCitadel, the sign of the second term to the output subunit quadratic forecast, and the output of the adder and the subunit is the third information output device, the second subunit of the calculation of the first derivative in the second (n-1)-th point design history from one adder, in which the first term is connected with the information output unit smoothing the input of the second term - with power inverters second myCitadel, and the output of the adder and the subunit is the fourth information output device and the third subunit of the calculation of the first derivative in the third (n-2)th design point prehistory of two adders and block inverters, in which the sign of the first term of the first adder connected to the output of the block smoothing, the sign of the second term to the output multiplexer of the second myCitadel, the output of the first adder through the power inverter is pulled to the entrance of the second term of the second adder subunit, the sign of the first term of which is connected to the output of the second myCitadel, and with a shift input tyres summand on one digit toward the high-order bits of the second adder, the output of which is the output of the subblock and the fifth information output device.

**Same patents:**

FIELD: electricity.

SUBSTANCE: device comprises an amplifier - limiter, an amplitude modulator with transformer separation of an "input-output" circuit, a summator, an integrator, a relay element, a unit of a logical function "Equivalence", a frequency divider, a separation transformer, a rectifier (demodulator).

EFFECT: improved accuracy of device operation and reduction of weight and dimensional indices.

4 dwg

FIELD: information technology.

SUBSTANCE: device has five comparators, five closing and five opening switches.

EFFECT: possibility of changing coefficients of realised functions without changing gain of operational amplifiers.

1 dwg, 1 tbl

FIELD: information technology.

SUBSTANCE: multizone frequency-pulse-duration alternating voltage controller has a terminal, a setting signal source, first and second adders, an integrator, first, second and third relay elements, first, second, third, fourth, fifth and sixth comparators, first, second and third D flip-flops, first, second and third power switches, a load distributed on phases A, B, C, terminals for connecting a three-phase voltage source A, B, C, and output terminals.

EFFECT: high reliability of operation of the voltage controller.

6 dwg

FIELD: information technology.

SUBSTANCE: device has a signal source, an input four-terminal network, an adder, an integrator, a relay element, a switch element, a four-terminal feedback network, a nonlinear link with a dead zone, two univibrators, a count pulse generator, an adding counter, a memory register, a digital comparator, a binary signal source, a delay element and logic elements AND, OR and NAND.

EFFECT: high accuracy of operation of the sensor.

4 dwg

FIELD: information technology.

SUBSTANCE: multi-zone analogue-discrete current sensor has a first and a second operating four-terminal network, a group of switch elements equal to the number n of relay elements, a pulse generator, a univibrator, a first and a second NOT logic element, a first and a second three-input AND logic element, a delay element, a first and a second binary counter, a first and a second memory register, an arithmetic-logic unit, which compares codes, counts the number of '0' and '1' signals, and also performs the inhibit function if the number of '0' and '1' signals is not equal.

EFFECT: high accuracy of operation of a multi-zone scanning transducer.

5 dwg

FIELD: radio engineering.

SUBSTANCE: signal mixer comprises a multiplying Hilbert cell, the first and the second output transistors of the channel "Y", the first and second dividing separators, from the first to the third current-stabilising dipoles, a load circuit, the first and second additional transistors.

EFFECT: reduced voltage of supply compared to a prototype by at least 1 volt when using a single-phase signal along a channel.

8 dwg

FIELD: radio engineering.

SUBSTANCE: analogue mixer of two signals with an output cascode comprises an input multiplying Gilbert cell, the first and second output transistors, the first and second load resistors, the first and second additional resistors, the first and second current-stabilising dipoles.

EFFECT: higher ratio of mixer amplification with low-voltage power supply.

3 cl, 12 dwg

FIELD: radio engineering.

SUBSTANCE: analogue mixer of two signals comprises an input multiplying Gilbert cell, the first and second additional transistors, the first and second load resistors, the first and second additional resistors, the first and second dipoles with low differential resistance by AC, the first and second circuits of static mode setting, the first and second additional current-stabilising dipoles.

EFFECT: higher ratio of mixer conversion K_{y} for its main f_{x}+f_{y}, f_{x}-f_{y} output harmonics with low-voltage power supply.

3 cl, 12 dwg

**FIELD:** computational engineering.

**SUBSTANCE:** additional groups of m fifth I elements, a delay element, a trigger, a group of elements, a group of sixth registers were introduced, wherein the output of each of them being connected to the first input of the fifth I elements group, the second input of which is connected to the startup unit, the delay element input is connected startup unit, its output being connected to the first trigger input and the second input of which is connected to the counter overflow output and the output being connected to the second I element input.

**EFFECT:** enhanced the functional capabilities of the device with reference to fast determination of optimum knapsack filling with various items in such a the total value of the knapsack is as large as possible, its total weight being fixed.

1 dwg, 1 dwg

FIELD: information technology.

SUBSTANCE: device has five analogue multipliers, four analogue subtractors and three analogue adders.

EFFECT: performing logic and arithmetic operations with discrete and analogue values of zeroes and ones.

1 dwg, 3 tbl

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to digital signal processing and information measuring equipment and can be used to linearise the transform function of units and systems, as well as interpolation and extrapolation of measurement results. The apparatus has an analogue-to-digital converter, an integrator, a unit for calculating sequences of quotients and differences and a unit for calculating parameters of the approximating unit, connected in series, as well as a unit for setting integration intervals, the output of which is connected to the control input of the integrator, and a threshold circuit whose input is connected to the output of the unit for calculating sequences of quotients and differences, and the output is connected to the control input of the unit for calculating sequences of quotients and differences.

EFFECT: high accuracy of approximation.

3 cl, 5 dwg

FIELD: information technology.

SUBSTANCE: apparatus has an input realisation storage unit, a unit for determining quasi-stationarity areas, a mask forming unit, an element-by-element multiplier, an approximation unit, an estimate storage unit, an estimate averaging unit, a useful component estimate storage unit, a current row counter, delay units, an averaging signal generator, a current column counter and a clock pulse generator.

EFFECT: two dimensional estimation of the useful component in conditions with insufficient prior information on statistical characteristics of additive noise and the useful component function.

2 cl, 3 dwg

FIELD: information technology.

SUBSTANCE: invention relates to radio engineering and can be used to filter information processes transmitted using frequency-modulated signals. The method involves prediction of the estimate of the modulating signals on the i-th (i=1, 2,…) extrapolation interval with duration h; determining on that interval the derivative of the reference radio signal dependant on said predicted estimate, synchronously with time sampling of said derivative and the input signal with frequency F>1/h, determining products of their readings; generating a signal on the i-th extrapolation interval which is proportional to the sum of these products; improving the predicted estimate of the modulating signal using said sum; frequency-modulation, using the predicted estimate of the modulating signal, of a pulsed signal which controls sampling of the input signal and the derivative of the reference radio signal.

EFFECT: high processing accuracy owing to adaptive clustering of observations when filtering frequency-modulated signals.

4 dwg

FIELD: information technology.

SUBSTANCE: filter comprises difference units, correction units, adder units, delay circuits, matrix function units, shaping and readout unit of apriori data, regularisation parameter unit.

EFFECT: accuracy of estimating information process parameters in measuring systems.

7 dwg

FIELD: electricity.

SUBSTANCE: electric element having ports and linear electric properties characterised in matrix, which is impedance matrix, admittance matrix or dissipation matrix of electric element and connecting voltage applied to ports with current passing through these ports. Electric element has inactivity determined by means of parameters perturbation up to perturbated set of parameters provided that this perturbated set of parameters corresponds to function in Boolean values.

EFFECT: increase of forecast accuracy for technically relevant linear electric properties of electric elements.

16 cl

FIELD: information technology.

SUBSTANCE: digital predicting and differentiating device includes a unit for estimating first derivatives, having a subunit for calculating a first derivative at a second (n-1)-th reference point of process history consisting of three adders, the output of which is the third data output of the device, and a subunit for calculating a first derivative at the third (n-2)-th reference point of the process history consisting of three adders and a block of inverters, the output of the subunit being the fourth data output of the device.

EFFECT: broader functional capabilities of prediction devices by obtaining estimations of first derivatives using numerical differentiation formulae for prehistory nodes of an input smoothed discrete sequence.

5 dwg, 2 tbl

FIELD: information technology.

SUBSTANCE: device has an image storage unit, a pixel storage unit, a turning unit, a vocabulary forming unit, a vocabulary storage unit, a processing unit, a priority calculating unit, turning unit, an adaptive form determination unit, a similarity search unit, a pixel averaging unit, an image filling unit and a clock pulse generator.

EFFECT: reconstruction of image pixel values under incomplete a priori information conditions.

6 dwg

FIELD: information technology.

SUBSTANCE: method of detecting and eliminating pulse noise when processing images involves comparing values of the original digital image with different threshold values. A set of penalties is then created for values of the original digital images exceeding the threshold values. The resultant penalty values are formed by adding separate penalty values for each reading. Readings whose resultant penalty values exceed the calculated threshold level are considered abnormal. Further, two-dimensional arrays of penalties are formed for each threshold level. Two-dimensional nonstationarity regions are determined and localised by a two-dimensional programmed detector with subsequent zeroing of the penalty values. The detected pulse noise values are eliminated by replacing them values of a first-order approximating surface on the localised regions.

EFFECT: detection and elimination of pulse noise values when processing digital images in conditions of non-parametric expected uncertainty of statistical characteristics of pulse noise and the image.

2 cl, 4 dwg

FIELD: information technology.

SUBSTANCE: device has a smoothing unit consisting of an adder, inverters, comparators, counters, AND logic elements, a deviation ratio setting unit and a dynamic characteristic control unit, and a prediction unit comprising three subtractors, two prediction subunits and a register.

EFFECT: high accuracy of prediction and simplification of the device.

7 dwg, 2 tbl

FIELD: information technology.

SUBSTANCE: apparatus for processing two-dimensional signal when reconstructing images has an image storage unit, a pixel storage unit, a directory creating unit, a directory storage unit, a processing unit, a priority calculation unit, an adaptive form determining unit, a resemblance search unit, a pixel averaging unit, an image filling unit and a clock-pulse generator.

EFFECT: reconstruction of image pixel values with incomplete prior information.

5 dwg

FIELD: computer science.

SUBSTANCE: device has sum forming blocks, matrix functions forming block, difference forming block, delay lines, apriori data output block.

EFFECT: higher precision.

6 dwg