CC BY
10ИНТЕЛЛЕКТУАЛЬНЫЕ ИНФОРМАЦИОННЫЕ СИСТЕМЫ
УДК 621.396.4
Method of adaptive intellectual control of technical condition of radio-electronic systems
P.A. Budko, A.M. Vinogradenko, A.V. Mezhenov, V.E. Zaremba
Abstract. At present, monitoring of technical state of complicated technical objects under different attacks and destabilizing factors, aging and technological dispersion of parameters is a crucial problem. Requirements to the quality, security, and reliability of complicated technical systems are consistently increased. In this chapter, we propose new method for adaptive control of technical states of radio-electronic systems. This approach is based on the interval complex estimation of parameters, use of knowledge base of critical and regular states, and also inner connections between the controlled parameters considering false negative result and false positive result. Multidimensional presentation of technical state of controlled systems is possible using the accurate monitoring of technical states of radio-electronic systems, increased accuracy and reliability of state identification, and extended possibilities of control and diagnostic equipment.
Keywords: Parameter estimation, Control, Radio-electronic system, Technical state, Knowledge base, Complex parameter evaluation.
1 Introduction
Radio-Electronic Equipment (REE) is a part of modern technological systems in different industries of manufacturing electronics and electrical equipment. The importance and relative volume of REE in technical systems increases continuously. That demands a creation of effective control and diagnostic equipment applied at the test and exploitation stages. Nondestructive control is especially important for uninhabited objects in remote territories, where execution of the control and diagnostic functions of REE by staff is impossible. At the same time, volume of increased Measuring Information (MI) during the control of the remote and territorial distributed objects requires the volume reduction and increased reliability of the transmitted messages.
New approaches of contactless diagnostics (for example, computer vision) are successfully developed. These approaches are applied in Information and Measuring Systems (IMS), telemetry systems, and Autonomous Automated Control Systems (ACS). They are implemented on the basis of contactless ways of MI exchange, digital signal processing, etc. Functions of telemetry systems are the following [1, 2]. First, MI entering from control objects is collected and processed by one or several peripheral control elements. Second, after MI delivering through communication channels its full processing is implemented and resulting outputs are provided to the staff managing Radio-Electronic Systems (RES). Transition from pre-emergency Technical State (TS) of REE to accident does not allowed. This requires the on-line collection of diagnostic information of control objects. At the same time, the tasks of TS estimation and identification of failure location in REE [3-5] are solved. The special properties are as follows:
Volume of processed by MI information is increased.
It is required to process several MI streams under the restricted resources of standard control elements.
Necessary control and measuring information, as well as, specialists for the analysis of REE technical condition may by unavailable sometimes in control centers.
This requires a development of new approaches for MI assessment in ACS, IMS, and telemetry systems.
For reduction of redundancy of measuring information under completeness of object control through communication channels, it is reasonable to transmit not all measuring results but only messages about an exit of object parameters outside of the limits of the set admissions. The systems
realizing such method of collecting telemetric data are called the adaptive systems of prestart control [6]. Achievement by the controlled parameter of threshold level in random time instant is an event initializing alarm signal [7]. In this case, the outlier value of parameter over threshold level is also a random value.
Conventional telemetry systems are divided in Remote Signaling Systems (RSS) and Telemetry Measurement Systems (TMS). However, we propose a complex model of alarm signal processing. It includes integration of the existing classes of systems as follows: the alarm signal is formed only in the case of exceeding by controlled parameter x the predefined threshold level (as in RSS) with the subsequent measurement (as in TMS) of the value of the emission above the threshold [6]. In the integrated system, random variables are the time instances tt of alarm signals and levels of these signals u .
For control of the technical condition of electronic equipment in the this work the approach based on integration of measuring information, arriving from group of polytypic sensors is offered. Their models promoting increase in informational content of monitoring of emergencies, reliability and the accuracy of control into account of false positive result (FPR) and false negative result (FNR).
The modelled representation of the refusal in REE consists in the following:
1) Complex use of the polytypic sensors (for example, sensors of temperature, tension of magnetic field, tension and humidity of air);
2) The processing of emergency signals in control systems integrating properties of systems of the remote signaling (SRS) and telemetry (TMS);
3) Mistakes check of FPR and FNP.
In this work, the adaptive control method of autonomous REE, which is based on estimation of their TS, is presented. It occurs by means of integration of critical results of measurements of parameters (emissions) by various sensors, ellipsoidal approximation of area of working capacity of the controlled object (CO) and formation of multidimensional refusal. At the same time mistakes of FPR and FNP are considered that allows to reduce redundancy of MI and also to increase reliability and accuracy of control.
The novelty of work consists in realization of the new approach to adaptive control of autonomous REE based on the complex estimation of their TS.
The theoretical and practical contribution consists in the following: the offered approach on control of the TS of REE allows to increase reliability and accuracy of definition of deviations of controlled parameters for prevention of emergencies.
Article has the following structure. Section 2 includes the analysis of works in the field of monitoring and estimation (recognition) of types of the TS of REE. Problem definition and a method of the decision are given in Section 3. Section 4 considers mechanisms of functioning of the offered adaptive control system and results of researches. Conclusions and the directions of the further researches are given in Section 5.
2 Related works
An analysis of the works [1, 2, 5, 8] shows that in order to ensure the effective functioning of REE while reducing the cost of their life cycle, it is necessary to introduce tools and methods for automated monitoring and diagnostics of TC. It is also necessary to use effective methods and means of ensuring the safety and reliability of the operation of REE.
At the same time, the specifics of functioning of control systems of the TS of REE with accounting of their operating modes rely on use of nondestructive ways of control and diagnostics. These methods are used in various industries of industrial electronics and electrical equipment [15]. Their disadvantage is the high probability of denial of service. This is due to the fact that the control thresholds are assigned without taking into account the general state of the communication system and the load level of the buffer devices in the switching nodes. These conditions cause blocking of knots in the loaded network, rather low productivity and high coefficient of idle time. It
is caused by the fact that for control of complex technical systems and identification of their state it is necessary to perform the measurement, transformation and processing of a large number of parameters connected with shutdown of a system and its idle standing.
Among the existing ways (strategy) of monitoring by the most optimum the control of the TS focused on the reliability of the CO, providing adjusting and the anticipating, that is preventive, predicting control methods is considered [10].
So in [11] TS of the controlled objects characterize by comparison of an emergency with the reference table of conditions of emergence of malfunction.
Along with the known methods of estimation [3-5, 12, 13, 16-19], in [20] procedures are offered will predetermine the REE models. They have properties of nonlinearity and multiconnectivity and also allow to build adaptive control algorithms with identification or reference model.
Predetermination of the assessment of the TS of REE in [2] is carried out due to comparison of the measured value and values of the predetermined weight coefficients characteristic of the controlled equipment. However, in many cases of the prior information, it is not enough for the implementation (acceptance) of this or that assessment of the TS of REE, and selection of a posteriori data is small for some statistical conclusions.
In these conditions receiving enough reliable results provide methods of statistical classification [16], the theory of neural networks [4], intellectual agents [17] and others. These methods have the merits and demerits which are used at the solution of problems of assessment of the TS, forecasting of changes of the REE controllable parameters. So, in [18] for controlling a robotic arm the neural network with back-propagation is used. In [19] it is offered to use the positioning ontology which models the spatial and temporal relationship between the observations from different sensors for assessment of a condition of the Internet of Robotic Things elements. It says about the efficiency of use of the adaptive methods for technical estimation of REE.
Monitoring of autonomous objects is, as a rule, characterized by the automated wireless exchange of MI [21] that allows to reduce considerably a time resource and participation of the person. At the same time, the configuration of the system of continuous monitoring allows to carry out adjustment of the parameters, by response to sudden fluctuations in a state of the CO or sharp changes in resource requirements [22].
An alternative to the aforementioned methods is the collection and processing of MI, implemented in multi-level monitoring systems TS RES, in which the collection and processing of MI is based on its comprehensive assessment. And the MI collection stage is presented in the form of transmission and processing of data on the output of object parameters beyond the specified tolerances This stage promotes reduction of redundancy of MI, due to integration of RSS and TMS. Such integration in a complex with classification of emergencies and mistakes check FPR and FNR promote increase in productivity (efficiency) in a control system.
In general, the conducted researches in the field of control of the TS of RES, recognition of types of refusals and their forecasting are characterized by quite wide range of approaches in this subject domain.
3 Mathematical backgrounds: representation of the results of control
For implementation of assessment of the TS of REE, it is offered to use an area of working capacity which dimension is defined by the number of output parameters. The TS of REE is defined by finding of parameters, characteristic of concrete type of the equipment, within admissions. Detailed submission of information on the TS of REE requires implementation of complex accounting of characteristic signs and mistakes by transfer, receiving and processing of MI.
3.1 The complex nature of the transmitted signals
When carrying out monitoring of the TS of REE, at a stage of the collecting MI, for estimation of controlled parameters it is offered to use a way of complex statistical control of the TS of REE which has to meet the following conditions:
1) to carry out integration of indications of sensors so that the signs characterizing the TS of the CO in one parameter, «invisible» to one type of sensors, but identified by sensors of other type could be found;
2) to reveal emergency signals at multidimensional statistical control (at various levels of the system);
3) to increase scopes of technical means of control and diagnostics.
The complex nature of control consists in obtaining MI about the TS of REE. It is based on versatile signs j: temperature, electromagnetic response, humidity of air, tension, etc. Obtaining MI is carried out, respectively, from sensors of temperature, the tension of magnetic field, humidity of air, the voltmeter, etc.
Considering that the transfer of MI about the controlled equipment from sensors on blocks of processing of MI is carried out constantly, during the normal operation of the equipment, its volume will be superfluous. For elimination of redundancy, it is necessary to involve one type of the sensors removing information at the moment of time on the most critical for a certain type of the equipment and its operating mode to the parameters. In those time points when the controlled parameters go beyond allowable limits, information arriving from sensors will confirm an emergency (prefault conditions) condition of the equipment. In this case it is important that the data arriving from the each sensor supplemented each other, giving fuller picture.
On the basis of the statistical analysis of the measured parameters Xi of several samples REE (for example, radio-electronic modules) establish reliability range - the area of operating states D representing an interval of dispersion of values of the parameters (proceeding from installation of allowable limits XH, XB) corresponding to the operating state of RES in general.
3.2 Creation of the MI three-dimensional model at the touch level
For combination of the indications of various types of sensors the method based on a grid of emissions and the Bayesian conclusion, modified for creation of three-dimensional model of the TS of REE on the basis of a surface of points is used. The measurements received from each of sensors are presented in the form of a surface of points in three-dimensional space (fig. 1), at the same time each point of r of a surface is presented by the following sizes:
1) mathematical expectation of the position of the point in three-dimensional space rx, ry, rz ;
2) matrix of the covariance r8 setting dispersion of three-dimensional normal distribution of the provision of a point;
3) probability of the parametrical refusal rref ;
4) probabilities of the receiving measuring signal of precritical conditions N REE elements
rd , where ii,...,N.
MI, received from the magnetic field sensor is presented in the form of several points with situation r^ = M^M;1 v , where Mr - the matrix setting position of the sensor concerning the REE
controllable element; M;1 - matrix of the TS dimension H*H, where H - length of v of vectors (with coordinate z) the digital sequence which represent h of measurements of the magnetic field sensor instant values, where h = 1,2,..., H.
The results of measurements in the form of thermograms of the REE controllable element, receive from temperature sensor. In process the analysis of control of parameters of the REE element results of measurements add to a surface of points in three-dimensional space (fig. 1). At the same time carry out identification of possible violations of stability of its TS on the basis of existence of nonrandom structures and use of borders of range of reliability [3, 4].
The results of the voltage measurements, register in the form of a separate point for which with a sufficient measure of trust the point interfaced by it is found. Formulas of calculation of a point of measurement in three-dimensional space are similar to the points for of the magnetic field sensor.
07
V y
■r
Fig. 1. Example of three-dimensional model of assessment of the TS of RES: a) hypersurface of TS of control object; b) the quality ellipse; c) the entered ellipsoid; d) the described ellipsoid
Information obtained from the sensor of humidity of air can be presented in the form of change of color range of points of a surface of the three-dimensional image (fig. 1). It occurs at corresponding change of humidity in the field of a controlled element which is characterized by an exit of the parameter (humidity) out of reliability range limits.
Coordinates of points can be expressed as r^ = M-lM-lRxRx¡í, where Rx¡¡Y - matrix of
deviations of the controlled parameter to the lower bound of admission concerning axis Y; R^Z -
matrix of deviations of controlled parameter to the upper bound of admission concerning axis Z of area of operating states D .
The covariance matrix in the global system of coordinates is set by expression r5 = M-lM-lRx rRx ZM¡. At the same time the value of diagonal of a matrix Ms corresponding to
axes X, Y in the local system of coordinates of all sensors are set proceeding from the chosen approximation step, and the value corresponding to axis Z is set proceeding from the accuracy of the concrete sensor.
This model can be expressed mathematically as follows. Let employment of a point of r is defined by a variable of a state Cell, which accepts one or the other values: Cell(r) = {it is busy, free}.
Emission - a signal of an exit of controlled parameters for reliability range borders (an accident signal) arriving from of the magnetic field sensors, tension, humidity of air and tension in some direction 9, is defined by the Cond variable: Cond(r, 9) = {accident signal, norm signal}.
All possible directions 9 (from 0 to 360°) breaks into q of the sites designated 9 . Variable conditions of Cell and Cond unite through logical operation of following. Let's consider the following assumptions:
0:Cell(r) = it is busy;
R¿:Cond(r, 9) = emission.
Then O is expressed through Ri as follows:
R vR v... vRn_! vRn ^ O . (1)
For determination of the probability that the point is occupied the Bayes method is used. For each point of r the validity of offer O is defined. As O treats R through logical following, probability can be determined as
P(O) = P(R vR v... vRn). (2)
Let's find probabilities of offers O and Ri:
P(O) = l-n(l-P(R)). (3)
y
Expression (3) can be used for calculation of the probability that the point is occupied (an exit of controlled parameter for reliability range borders) if probabilities of emissions P(R) are known. At practical application expression (3) can be written down on the basis of a formula of total probability concerning value P (r).
Let's apply the rule of Bayes for determination of the probability P(R / r) on new measurement of controlled parameter r :
P^^MPP^ (4)
where P (R) - initial probability of the received signal of emission. In the Bayesian rule it usually is accepted eaual 1/2 as is impossible to define initially if the point is busy or free.
Complex model of the sensor. The value P(rR) is called sensor model. The model of the sensor determines the probability of receiving measurement of r if it is known that offer R - is true.
Sensor PN (rR) model for of the magnetic field sensor can be presented as follows. Let Xi -the random variable characterizing controlled parameter ustai (tension fixed by of the magnetic field sensor), and f (ustaf) - function of density of probability of stay X, in the reliability range, then
fa, if f (u_f )< 0;
P (rR) =J , f ( stmf) ; (5)
N (R) jl, if f (ustmf )> 0. ()
At the registration of the controlled parameter r in this range emission is formed. Taking into account fixing of violation of range of reliability the sensor at an exit of the controlled parameter from the range of admissions at the same time of two and more REE elements on the decisive device several signals of emissions can arrive at the same time, and several points in the three-dimensional space are respectively received. Generally, the number of such points will be in the proportion to the measured of the magnetic field sensor. Thus, the probability that the point displays a parameter exit in this direction of range of reliability is inversely proportional to of the magnetic field.
Sensor PT(r|r.) model for the sensor of temperature can be presented as follows. Let X = (X...x )T- a vector of averages in the t -th instant selections of measurements of temperature (t = l,...,m), Xj - average value in the t -th instant selection in the parameter j .
The main criterion of the violation of stability of process - an exit of controlled parameter for threshold level (for reliability range borders), then
iwlr>\ I1, f Tt < Tthreshold; /¿-\
Pt (r\R) = j_ ,,T „ (6)
j0, It It > Tthreshold.
The model of the sensor of tension Pu (rR,.) can be described by the expression similar to of the magnetic field sensor model:
R (rR) j0, if f(umeas^ed )< 0; (?)
u(R) jl, iff(u_d)>0, ()
where u - the measured value tension sensor.
Sensor Pv(rR,) model for the sensor of the humidity of air to within values of probability can be presented as follows. Let s - be the color range corresponding to the range of change of the humidity of air in the field of the controlled element (fig. 1, a) which is in various states, and f (5) - function of density of probability of a range of s, then
0,3, if f (5)< £ A 5 < r;
Pv(rR) = j 0,5, if f (5) < £ A 5 < r; (8)
0,95f (5), if f (5)> £.
Formation of a complex signal of TMI. As a rule, the TS of REE is characterized not only by one parameter, but the whole group of parameters. The method of purpose of independent intervals to each parameter existing now separately does not allow considering the correlation of the REE parameters. For accounting of correlation of parameters, and, as a result, increase in reliability of
control, the area in space of parameters in which with the set probability there are values of controlled parameters is offered to consider. Therefore at assessment of the TS of REE is offered to use an area of working Dp capacity which dimension is defined by the number of the output parameters characterizing the TS of REE in general [4, 5].
Influence of the external indignations which exact properties are unknown and cannot be described by simple ratios, leaves a mark on area of working D capacity, washing away its borders. In this regard, values of the REE controllable parameters always define with a final error.
Taking into account normal distribution of the real measurements in the presence of errors FPR and FND in which with the set probability there are values of the measured output parameters of the CO is expedient to approximate an area of working capacity an ellipsoid [14-16] (fig. 1, c, d).
3.3 Formation of an ellipsoid of working capacity taking into account mistakes of the FPR and FNR
Control of the REE separate parameters, without their interrelations, or does not provide the required size of reliability of control, or excessively overestimates operational indicators, at the same time numerous false signals of an accident are possible [6-9].
For control of the CO parameters on several correlated indicators in works [3-5] multidimensional methods of the statistical analysis are used. It assumes correcting of values of the CO parameters during its operation by results of selective control. This procedure is necessary for maintenance of statistically operated and stable process of work CO, however at the same time there is no accounting of mistakes of errors FPR and FNR.
In the modelled system of monitoring, at a deviation of the controlled parameters of the REE elements, the comparison of parameters with threshold values within Dp is made. By the results of comparison, the normal state of REE decides on probability px or its abnormal state with probability p2. And TS recognition (critical condition) of REE is carried out taking into account mistakes of the FPR (a) and the FNR (p).
Minimization of the probability a comes down to creation of the entered rectangle BB of the maximum area (fig. 1, b) also decides by means of a method of diagonals. According to this method of top of the entered parallelepiped (fig. 1, d) are in a point of intersection of diagonals of the described parallelepiped with an ellipsoid (fig. 1, c and fig. 2). For this purpose the values of the parties of the entered parallelepiped (fig. 1, d), corresponding to the admissions on parameters of the RES elements when ensuring zero probability of a mistake p and minimum possible mistake a .
Fig. 2. Ellipsoid of working capacity (informational content)
Ensuring the required correlation a/p between the FPR and FNR errors when approximating the region D of admissible parameter values (the hatched described ellipsoid in Fig. 1, c, d), determines the tolerances for the parameters. With these tolerances, a certain probability of an
undetected failure of the REE elements is preliminarily set, or the cost of the control and monitoring system is minimized when implementing the established requirements for the TS indicator of the controlled object (quality of operation).
Thus, set of the controlled parameters which are in limits of admissions at ellipsoidal approximation in a complex look will represent a characteristic form of an ellipsoid (fig. 2). Its dimension (parameters) will also be written down in memory elements. At a deviation of parameters from norm, the proportion of a figure will change, and the dynamics of the changing parameter will be characterized by a color range from violet (normal state) to red (critical condition).
Complex use of MI, received from the diverse sources, is presented in three-dimensional space, at ellipsoidal approximation. It is applicable for the solution of problems of control of the TS of objects in the conditions of uncertainty. Such approach promotes increase in reliability of MI about a condition of the observed objects in the systems of monitoring, to expansion of a scope of technical means of control and diagnostics and also decrease in redundancy of MI at stages of its transfer. In general, it promotes increase in efficiency of process of control.
4 Realization of a method of adaptive control of the TS of RES
For definition of the stages of control of the TS of RES we will define the following basic data:
list of the controlled parameters (temperature, tension, of the electromagnetic field strength, humidity of air, etc.);
frequency of the poll of RES sensors;
greatest possible dynamics (frequency, speed) of the change of the controlled parameter.
MI RES received when functioning, is necessary to transfer for processing to remote dispatch center management (DCM) for the analysis and final definition of the TS of RES.
The solution of a problem of control of the TS of RES is presented in the form of the sequence of the following stages (procedures).
Main stages of control of the TS of RES:
I. Preprocessing of MI on the remote terminal:
stage 1 - scan poll controllers (servers) of sensors on the RES elements;
stage 2 - drawing up the knowledge base (statistical data) in controlled parameters;
stage 3 - primary estimation of the values of the received group of signals (remote signaling) from the RES controllable element: appraisal evaluating degradations of the controlled parameter (application);
II. Processing of the signals of a TMI on DCM:
stage 4 - secondary complex estimation of the received group of signals of RSS-TMS:
а) formation of a multidimensional complex image of the TS of RES (in the most critical parameters);
б) comparison of the received image with reference values of images of signals from the knowledge base;
stage 5 - identification of the TS of RES taking into the account mistakes of the FPR and FNR.
In case of an exit of different controlled parameters of objects out of the allowable limits, in the sensor located directly on elements of objects the signal of critical condition of the such objects is formed. In the existing systems of the telemetry each parameter of an object is controlled with the period To, irrespective of its speed of change. However at the increase of the speed of change of the separate parameters they can reach permissible values in a time smaller the fixed period T0. In this case the control system will not be able to react in due time to inadmissible changes of parameter that will lead to refusal of a controlled object. For efficiency of control of a condition of an object measurement and the subsequent assessment of parameter is carried out with a frequency of proportional speed of change of the parameter. Depending on exit speed (time A^, At2 achievement of permissible value) controlled parameter U out of the allowable limits the priority of a signal is defined.
This is set thanks to a multi-level tolerance system (the higher the tolerance level, the higher the priority of the service request).
An autonomous electrical installation (power plant) was used as the control object, and the proposed approach was used to monitor TS. As a subject to control autonomous electro installation
(power plant) for which monitoring of the TS the offered approach was used. When carrying out researches accounting of internal parameters of electro installation (output voltage, temperature of heating of the generator (anchor), humidity of air and tension of the electromagnetic field) was made.
Proceeding from the offered models of sensors, the fixed deviations of controlled parameters, are displayed in the form of an ellipsoid of various color scale. It allows, to comparison with a reference ellipsoid to display flowing the TS of a controlled object. Results of control of humidity of air on the RES controllable element are displayed in fig. 3. Given the availability of any emergency option in the knowledge base, based on the results of such a comparison, the current TS of the RES will be identified.
In general, multidimensional representation of refusal CO with use of the offered approach shows a considerable prize in reliability of MI and reduction of its redundancy.
Thus, the results of the conducted researches showed that complex idea of MI of the TS of RES is increased by reliability and informational content of the obtained resultant information.
Further development of researches of complex estimation of the TS of autonomous RES, at implementation of their control is possible by carrying out a computing experiment on the basis of methods of imitating modeling with use of software, for example, in the Anylogic programming environment. Use of software will allow to consider any aspect of the modelled system with any level of specification, and the graphic AnyLogic interface, tools and libraries will allow to create quickly models for a wide range of tasks from modeling of refusals, poll of sensors before development of all system of monitoring of the TS of RES.
5 Conclusion
The results of a research show that for implementation of monitoring of the RES parameters of, especially, uninhabited autonomous objects (for example, artificial Earth satellites) various instruments of control can be used. They provide increase in reliability of results of identification and sensitivity to detection emergency situations.
a) b)
Fig. 3. Informational content ellipsoid: a) display of the TS of the measured sample; b) «reference ellipsoid»
The approach of complex control of the TS of RES presented in article on the basis of integration of indications of several types of sensors can be used for creation of the universal automated complex of control of uninhabited autonomous objects of technological systems. Such complex of control includes the systems of technical sight and allows to estimate operability of the wide nomenclature of the REE with high reliability. Complex representation of MI, taking into account its transfer at integration of SRS and TMS, promotes decrease in redundancy of MI in the system of monitoring, to increase in reliability and accuracy of estimation of the TS of RES.
In general, the offered approach will allow carrying out support of decision-making in the control systems of the TS of RES on-line on elimination of critical conditions.
Acknowledgment
The research is executed with financial support of the Russian Foundation for Basic Research within the scientific project №. 16-29-04326 ofi_m.
References
1. Sokolov, B., Yusupov, R.: Conceptual foundations of quality estimation and analysis for models and multiple-model systems, Journal of computer and systems sciences international 46(6), 831-842 (2004).
2. Takayama, K., Kariya, S.: Autonomous Measuring by Sensing Node in Telemetry System, Measurement Science Review 3(3), 29-32 (2003).
3. Abramov, O.: Parallel algorithms for computing and optimizing reliability with respect to gradual failures, Automation and Remote Control 71(7), 1394-1402 (2010).
4. Abramov, O.: Choosing Optimal Values of Tuning Parameters for Technical Devises and Systems, Automation and Remote Control 77(4), 594-603 (2016).
5. Abramov, O., Dimitrov, B.: Reliability design in gradual failures: a functional-parametric approach, Reliability: Theory&Application 12(4), 39-48 (2017).
6. Fedorenko, V., Vinogradenko, A., Kononov, Y., Samoylenko, V., Samoylenko, I.: The Time-Probability Characteristics of a Telemetric Signal with the Variable Number of Bits, In: Proceedings of the 2017 IEEE II International Conference on Control in Technical Systems (CTS), pp. 146-149 (2017).
7. Katzel, J.: Managing Alarms, Control Engineering 54(2), 50-54 (2007).
8. Fantana, N., Pettersson, L., Perkins, M., Girgis, R., Fazlagic, A.: Method and system for systematic evaluation of evaluation parameters of technical operational equipment, Patent No. US 7 239 977 B2 (2007).
9. Fedorenko, V., Budko, P., Vershkov, N.: Mathematical model of discrete communication channel under the influence of destabilizing factors, Engineering Simulation 15(1), 77-83 (1998).
10.Yang, W., Tavner, P., Crabtree, C., Feng, Y., Qiu, Y.: Wind turbine condition monitoring: technical and commercial challenges, Special Issue: Condition Monitoring 17, 673-693 (2014).
11.Yoshida, A.: Centralized network monitoring device for monitoring devices via intermediate monitoring devices by means of polling and including display means displaying screens corresponding to hierarchic levels of the monitored devices in a network, Patent No. US 5,504,863 (1996).
12. Buryi, S., Loban, A., Lovtsov, D.: Compression Models for Arrays of Measurement Data in an Automatic Control System, Automation and Remote Control 59(5), 613-631 (1998).
13. Lovtsov, D., Karpov, D.: Dynamic Planning of Navigation Determinations of Airspace and Missile Objects in an Automated Flight Test Control System, Solar System Research. PAS 45(7), 645-651 (2011).
14. Chernousko, F.: State estimation for dynamic systems. Boca Raton, Florida: CRC Press (1994).
15. Lluis, R., Assumpta, S., Federico, T.: An ellipsoidal calculus based on propagation and fusion, IEEE transactions on systems, man, and cybernetics - part b: cybernetics 32(4), 430-442 (2002).
16. Singha, J., Das, K.: Hand Gesture Recognition Based on Karhunen-Loeve Transform, In: Proceedings of Mobile & Embedded Technology International Conference, pp. 365-371 (2013).
17. Kotenko, I., Ageev, S., Saenko, I.: Implementation of Intelligent Agents for Network Traffic and Security Risk Analysis in Cyber-Physical Systems, In: Proceedings of the 11th International Conference on Security of Information and Networks (SIN'18), Article 22, 4 pages (2018).
18. Kumar, R., Kumar, S., Lal, S., Chand, P.: Object Detection and Recognition for a Pick and Place Robot, In: Proceedings of the IEEE Asia-Pacific World Congress on Computer Science and Engineering (2014).
19. Simoens, P., Dragone, M., Saffiotti, A.: The Internet of Robotic Things: A review of concept, added value and applications, International Journal of Advanced Robotic Systems I-II, 1-9 (2018).
20. Pshikhopov, V., Medvedev, M., Gaiduk, A., Neydorf, R., Belyaev, V., Fedorenko, R., Krukhmalev, V.: Mathematical model of robot on base of airship, In: Proceedings of the IEEE Conference on Decision and Control 2013 IEEE 52nd Annual Conference on Decision and Control (CDC'2013), 959-964 (2013).
21. Joseph A.: Automated wireless preventive maintenance monitoring system for magnetic levitation (MAGLEV) trains and other vehicles, Patent US 5,445,347 (1995).
22. Lawson, B., Smirni, E., Puiu D.: Self-adapting backfilling scheduling for parallel systems, In: Proceedings International Conference on Parallel Processing. Conference Location: Vancouver, BC, Canada, Canada 21 Aug. 2002. IEEE Xplore (2002).
Метод адаптивного интеллектуального контроля технического состояния
радиоэлектронных систем
Будко П.А., Винограденко А.М., Меженов А.В., Заремба В.Е.
Аннотация. Повышение требований к качеству функционирования сложных технических объектов (радиоэлектронные системы, робототехнические комплексы, летательные аппараты и др.), а также к их безопасности и надежности сделали особенно актуальной проблему мониторинга их технического состояния, с учетом воздействия различного рода атак и дестабилизирующих факторов, старения и технологического разброса параметров. В статье предлагается новый подход к совершенствованию метода адаптивного интеллектуального контроля технического состояния радиоэлектронных систем. Данный подход основан на интервальном комплексном оценивании параметров, использовании базы знаний об аварийных и штатных состояниях, а также взаимосвязанности контролируемых параметров с учетом погрешностей и ошибок первого и второго рода. Рассматривается реализация многомерного представления технического состояния радиоэлектронных систем, что позволяет производить четкую фиксацию их технических состояний, повышает точность и достоверность результатов идентификации состояний, а также расширяет возможности применения технических средств контроля и диагностики в системах контроля.
Ключевые слова: интеллектуальное оценивание, контроль, радиоэлектронная система, техническое состояние, база знаний.
Статья поступила 05 августа 2019 г.
Information about Authors
Budko P.A. - Scientific Secretary of PJSC «Information Telecommunication Technologies». Doctor of Technical Sciences, professor. Phone: +7(812)448-95-97; +7911010-92-64. E-mail: intelteh@inteltech.ru; budko62@mail.ru. Address: 197342, Russia, Saint-Petersburg, Kantemirovskaya, 8.
Vinogradenko A.M. - Military Academy of Communications named after Marshal of the Soviet Union S.M. Budenniy. Doctorate of Technical Sciences, Associate Professor. Doctoral candidate. Phone: +79214439022. E-mail: vinogradenko.a@inbox.ru. Address: 194064, Russia, Saint-Petersburg, Tikchoretskiy, 3.
Mezhenov A.V. - Military Academy of Communications named after Marshal of the Soviet Union S.M. Budenniy (St. Petersburg). Graduated in a military academy. Phone: +79697203007. E-mail: a.mezhenov@yandex.ru. Address: 194064, Russia, Saint-Petersburg, Tikchoretskiy, 3.
Zaremba Vladyslaw Evgenevich - St. Petersburg State University of Telecommunications of M.A. Bonch-Bruevich. Master's student. Phone: +79061531748. Address: 193232, St. Petersburg, ave. Bolshevikov, 22, k.1.
Информация об авторах
Будко Павел Александрович - Ученый секретарь ПАО «Интелтех». Доктор технических наук, профессор. Область научных интересов: управление ресурсами сетей связи; морская робототехника. Тлф.: +7(812)448-95-97; +7(911)010-92-64. E-mail: intelteh@inteltech.ru; budko62@mail.ru. Адрес: 197342, Россия, г. Санкт-Петербург, ул. Кантемировская, дом 8.
Винограденко Алексей Михайлович - Военная академия связи имени Маршала Советского Союза С.М. Буденного. Кандидат технических наук, доцент. Докторант. Область научных интересов: модели и методы формирования и обмена телеметрической информацией в едином информационном пространстве РФ. Тлф.: +7(921)443-90-22. Адрес: 194064, Россия, г. Санкт-Петербург, Тихорецкий пр., д. 3.
Меженов Алексей Викторович - Военная академия связи имени Маршала Советского Союза С.М. Буденного. Адъюнкт. Область научных интересов: обеспечение контролепригодности технического состояния средств связи и РТО в едином информационном пространстве. Тлф.: 8(969)720-30-07. Адрес: 194064, Россия, г. Санкт-Петербург, Тихорецкий пр., д. 3.
Заремба Владислав Евгеньевич - Санкт-Петербургский государственный университет телекоммуникаций им. М.А. Бонч-Бруевича. Магистрант. Тлф.: +7(906)153-17-48. Адрес: 193232, Санкт-Петербург, пр. Большевиков, д.22, к.1.
Для цитирования: Будко П.А., Винограденко А.М., Меженов А.В., Заремба В.Е. Метод адаптивного интеллектуального контроля технического состояния радиоэлектронных систем // Техника средств связи. 2019. № 4 (148). С. 59-69. (In English).
For citation: Budko P.A., Vinogradenko A.M., Mezhenov A.V., Zaremba V.E. Method of adaptive intellectual control of technical condition of radio-electronic systems // Means of communication equipment. 2019. No 4 (148). P. 59-69.
