RU2579422C1 - Method of hidden evaluating and monitoring of hazardous psychophysiological states of human-machine system operator during professional activities - Google Patents

Abstract

FIELD: psychophysiology.

SUBSTANCE: perform hidden assessment and monitoring of hazardous psychophysiological conditions (PPC) of human-machine system operator during professional activities (PA) taking into account compared parameters of cardiogram (ECG) when hiring operator or re-evaluating knowledge on object-simulator when performing different types of PA and on the real object. First, parameters are evaluated in initial state - before the beginning of PA, then - during PA on object-simulator, then -on the real object. Based on comparison results the decision on operator`s PPC is taken. Upon that ECG is registered without the knowledge of a person being tested. Stationary sections of ECG are defined with subsequent formation of probability density of heart rate variability (HRV) parameters on each of them. Their are accepted as references of operator`s adequate PPC in the initial state and at the stage of performing standard PA tasks on object-simulator. Prior to PA on the real object perform authentication of operator while simultaneously recording his ECG and separating stationary sections and forming on each of them the distribution density of HRV indices detected at previous stages. Define degree of similarity of the obtained probability density of these indices at the initial stationary section with standard distribution sample obtained earlier when operator was in the initial state prior to PA on object-simulator. Tolerance to PA at real object is performed under two conditions: presence of operator`s authenticator and adequate PPC which is determined, if the degree of similarity of HRV probability density exceeds the specified level. Otherwise the operator`s permit to work at real object is locked. In process of PA at he real object upon ECG recording the probability density of HRV indices obtained on each following stationary section is compared with appropriate previously received distributions-references. Distribution which is the most close to the current distribution is specified. Based on this distribution PPC is defined in which tested person is currently residing. If the degree of distribution similarity does not exceed the preset level, the operator is in inadequate state - dangerous for PA continuation.

EFFECT: method provides high reliability of operator`s PPC evaluation during PA.

1 cl, 4 dwg

Description

The invention relates to safety systems for the functioning of human-machine systems and is intended to alert security services about dangerous psychophysiological conditions of an operator interacting with a system.

The main cause of emergency situations in any human-machine system is the human factor. The operator may be in an inadequate psychophysiological state: alcohol intoxication, excessive agitation, at the stage of falling asleep, etc. [Human factor. In 6 volumes. T. 1. Ergonomics - a comprehensive scientific and technical discipline / Ed. G. Salvendi. - M .: Mir, 1991. - 600 p.].

A known method for assessing the psychophysiological state of a working operator by the frequency of eye blinks, which changes with the accumulation of fatigue [http://news.bbs.co.uk/hi/Russian/sci/tech/newsid_2246000/2246943.stm].

The disadvantage of this method is the high probability of false alarm and the inability to organize a hidden mode of formation of standards for dangerous psychophysiological conditions of the operator.

A known method for recognizing the occurrence of fatigue of a working subject, developed by Mercedes-Benz [http://www.berlinauto.ru/index.php?p=technologies&id=1230]. According to the method, the subject’s movements are continuously monitored, and based on the received data, typical behavior typical for him is modeled. If the behavior differs from that obtained as a result of the simulation, a signal is issued recommending the subject to rest.

The main disadvantages of the method are the limited identifiable situations (fatigue), suitability for use in one type of system, high complexity, the need to introduce the subject into psychophysiological conditions that are dangerous for him (for example, drug use) at the stage of training the identification system.

A known method of non-contact detection of changes in the psychophysiological state of the subject, based on a sequential analysis of images of individuals in the far infrared optical spectrum [Pat. 2233558 RF, IPC А61В 5/04. The method of non-contact detection of changes in the psychological state of the subject [Text] / Vilenchik LS and etc.].

The disadvantages of the method is the low reliability of decisions and the inability to create standards of dangerous psychophysiological conditions of the subject.

A known method for assessing the psychophysiological state of a person by heart rhythm [Pat. 2246251 RF, IPC А61В 5/00. A method for assessing the psychophysiological state of a subject by heart rate [Text] / Godunov V.A. and etc.]. The method is based on the formula for assessing the stress index by the values of spectral powers in the low-frequency and high-frequency regions of the spectrum of the dynamic range of cardio intervals.

The lack of an interface between the indicated index and the psychophysiological states of the subject, considered in [Mashin V.A. On the classification of functional states of a person // Experimental Psychology, 2011, v. 4, No. 1, P. 40-56], limit the possibilities of using the method.

и среднеквадратическое отклонение этой частоты от средней (σ_i) при выполнении i-го вида операторской деятельности в том числе и в состоянии покоя (i=0). На стадии профессиональной деятельности уровень психофизиологического напряжения (У) определяют из соотношения

, а степень напряжения субъекта оценивают по отклонению текущей частоты ЧСС_t от

в единицах σ_i.Of the known technical solutions, the closest in the totality of features to the claimed solution is “A method for assessing and monitoring the psychophysiological state of the operator by the heart rhythm in the process of professional activity” [Pat. 2358647 of the Russian Federation, IPC А61В 5/02. A method for evaluating and monitoring the psychophysiological state of an operator by heart rhythm in the process of professional activity [Text] / Mokeev V.D. and etc.]. According to this method, preliminary (at the training stage) the parameters of the sequence of cardio-intervals of a subject in a state taken as “normal” are evaluated, the average heart rate is used as parameters for assessing the psychophysiological state

and the standard deviation of this frequency from the average (σ _i ) when performing the ith type of operator activity, including at rest (i = 0). At the stage of professional activity, the level of psychophysiological stress (U) is determined from the ratio

, and the degree of tension of the subject is estimated by the deviation of the current heart rate _t from

in units of σ _i .

The disadvantage of this method is the need to fulfill the conditions: to know which of the types of activities established at the training stage is implemented by the operator on a given period of time of professional activity. If the recorded process is characterized by medium

и σ_i в зависимости от его психофизиологического состояния до начала профессиональной деятельности (прием таблеток, алкоголя, конфликт с начальником и др.) заметно отличаются от принятых за эталонные, полученные на стадии обучения. Поэтому достоверность принимаемых решений не может быть высокой. Ограниченность используемого пространства признаков

не позволяет достичь приемлемых вероятностей ошибок 1-го и 2-го рода.between the ith and (i + 1) state, the task of determining the voltage V becomes undefined. In addition, the parameters of the working operator

and σ _i , depending on his psychophysiological state prior to the start of his professional activity (taking pills, alcohol, conflict with his boss, etc.) are markedly different from those accepted as reference ones obtained at the training stage. Therefore, the reliability of decisions cannot be high. Limited use of feature space

does not allow to reach acceptable probabilities of errors of the 1st and 2nd kind.

The aim of the invention is to increase the reliability of assessing the psychophysiological state of the operator in the process of professional activity.

This goal is achieved by the fact that when hiring an operator or recertifying knowledge, a cardiogram is hiddenly recorded in the initial state - before the start of activity, then - at the simulator object, when performing various types of professional activity, stationary sections on the cardiogram are distinguished with subsequent assessment of densities at each of them probability distributions of less than 10 accepted in medical practice indicators of heart rate variability and 25 additional indicators obtained by the original described below and illustrated in FIG. In two ways, which are taken as standards describing the adequate psychophysiological state of the operator in the initial state and when performing typical tasks of the upcoming work, at the stage of professional activity at a real object, the identity of the operator who is starting work is authenticated with the simultaneous hidden registration of his cardiogram with the allocation of stationary sections on it and building on each of them the probability densities of the HRV parameters, determining the degree of similarity of the resulting distribution the probabilities of these indicators at the initial stationary site with the distribution-standard obtained earlier in the initial state of the operator before starting work on the simulating facility, while admission to work on a real facility is subject to two conditions: the authenticator of the operator and the adequate psychophysiological state of the operator, which is established if the aforementioned degree of similarity of the probability distributions of HRV indicators exceeds a predetermined level; otherwise, the operator’s access to work on a real object is blocked;

then, in the process of the operator’s professional activity at a real object, when registering a cardiogram, they compare the probability distributions of HRV indicators obtained at each subsequent stationary section with the corresponding previously obtained standard distributions, select the distribution closest to the current one, by which it is established in what psychophysiological state the current moment is the subject, and if the degree of similarity of the distributions does not exceed a predetermined level, they find denie operator in an inadequate state - dangerous to continue the occupation.

The proposed method is illustrated by illustrations.

; d - показатель сходства текущего распределения вероятности параметра с одним из его распределений в базе эталонов, У_P - минимальный уровень сходства сравниваемых распределений, ниже которого допуск оператора к работе в системе блокируется (з); Р(С_1р) - плотность распределения вероятностей параметра С_1P, характеризующая кардиограмму до появления решения о смене ее свойств в момент времени

.In FIG. 1 shows a timing diagram of the system that implements the sequence of operations according to the proposed method: changing the cardiogram C _1E in time at the stage of checking the professional skills of the operator in the initial state C _0E and when performing various operations on the simulator C _1E , C _2E , ..., C _NE ( and ); decision making moments С _Р about changes in the cardiogram properties at points t ₀ , t ₁ , ..., t _N (b); probability density distribution of HRV indicators, i.e. With _iE , before making decisions at the appropriate time t _i (c); operations d) -i) reflect the sequence of the operator’s work on a real object, with C _A (t) - the decision-making function of the system to present the authenticator (password, smart card, etc.) (d), C _iP (t) - changing the cardiogram in the process of work (e), C _P (t) - decision signals on changing the properties of the cardiogram (e), P (C _0P ) - distribution density of the parameter C _0P cardiogram before starting professional activity at the time

; d is an indicator of the similarity of the current probability distribution of a parameter with one of its distributions in the database of standards, _P is the minimum level of similarity of the compared distributions, below which the operator’s access to the system is blocked (h); P (C _1p ) is the probability distribution density of the parameter C _1P , characterizing the cardiogram until the decision to change its properties at the time

.

In FIG. Figure 2 illustrates the idea of expanding the number of indicators to describe HRV: R _i, R _{i + 1} - i-th and (i + 1) -th heartbeat intervals, G (M) - standard deviation of the distance between points from the center of the selected square.

In FIG. Figure 3 shows the experimental one-dimensional probability distributions of HRV indicators in the initial state of the IS operator, when he consumed 30 mg of alcohol A, when he consumed a sleep pill C, VLF is the power of the very low-frequency components of the cardiogram, LF / HF is the ratio capacities in the low-frequency and high-frequency range of the cardiogram; C ₁ is the correlation coefficient of a sequence of cardio intervals shifted by one step; Region 18 - the percentage of hit points with coordinates R _i , R _{i + 1} in the area with number 18.

In FIG. 4, a , b depicts a modernized car simulator based on a personal computer, which was used to test the operability of the method in the conditions of work of subjects in situations that take place in practice: 1 - a car steering wheel simulator, 2 - a car pedal simulator block, 3 - a simulator block gear lever, 4 - monitors, 5 - speaker system, 6 - keyboard, 7 - system unit, 8 - cardiotechnics-04-8 (M) Holter monitor, 9 - Holter monitor cover mounted on the car seat, 10 - bl ok communication from Holter monitor with PC.

The above technical result is achieved as follows. When hiring or at the next recertification of knowledge, the operator is invited to demonstrate skills at a simulator or simulator. Unlike the simulators available on the market, it is additionally equipped with a contactless cardiogram recording sensor, which allows for the implementation of a hidden identification mode for the psychophysiological state of the subject both at the stage of preliminary assessment of HRV parameters and in the process of conducting professional activities.

Several technologies for remote measurement of human cardiac activity parameters have been proposed and tested. One of them is based on the use of a chair with an integrated “electromechanical film” [Annttonen J., Surakka V. Emotions and Heart Rate While Sitting on a Chair // Proceedings of the SIGCHI conference on Human factors in computing systems. New York, NY: ACM, 2005. pp. 491-493. [Electronic resource] - URL: http://portal.acm.org/citation.cfm?id=1055040]. More widely used is the technology of laser Doppler vibrometry, based on the analysis of interference patterns formed when the skin is illuminated by laser radiation [Habib Tabatabai, David E. Oliver, John W. Rohrbaugh, Christopher Popadopoulos. Novel Applications of Laser Doppler Vibration Measurements to Medical Imaging // Sensing ant Imaging: An International Journal, June 2013, V. 14, Is. 1-2, pp. 13-28; Chen M. et al. Laser Doppler vibration measurements of physiological functions: assessment of biometric capabilities // IEEE Transcations on Information Forensics and Security, 2010, 5 (3), art. No. 5483212, pp. 449-460]. The technology allows you to measure the pulse at a distance of 20 m from various sides of the subject.

An equally effective method for remotely obtaining information about the state of the cardiovascular system, based on the analysis of video and thermal images of the skin [Jain U., Tan B., Li Q. Concealed know ledge identification using facial thermal imaging // ICASSP. IEEE International Conference on Acoustics Speech and signal Processing, 2012, March 25-30. Art. No. 6288219, pp. 1677-1680; Cross C. B., Shipper J. A., Petkil D.T. Thermal imaging to defect physiological indicators of stress in humans // Conference on Thermosense - Thermal Infrared Applications XXXV, 2013, April 30-May 01, V. 8705, Art. no. 870501. DOI: 10.1117 / 12].

Given the small size of devices that implement the above technologies, the provision of a hidden mode of obtaining information about the state of the cardiovascular system of the subject of interest is not a limiting factor.

On the simulator (object-simulator) typical situations of professional activity of a re-certified (hired) operator are modeled. The recorded parameters of the cardiogram (one of which is shown in Fig. 1, a ) are analyzed for stationarity, in violation of which signals C _p (t) are generated (Fig. 1, b), which start the process of forming the empirical probability distribution density P (C _iE ) of the selected parameters C _iE (Fig. 1, c), taken as reference. Stationarity violation algorithms are known [Nikiforov I.V. Sequential detection of changes in the properties of time series. - M .: Nauka, 1983 .-- 200 p.]. The first of the distributions P (C _0E ) characterize the state of the operator before conducting professional activities (initial state). The technology for creating the initial state by removing the "stress before the exam" is known in the printing industry and comes down to the formation of so-called neutral questions [Varlamov V.A. Polygraph. - M .: PER-SE-Press, 2004. - 352 p.]. Upon completion of the examination tests, information about P (C _iE ) is entered into the computer of the control object. This completes the learning process if access to the object is carried out by the authenticator. Obviously, the standards characterize the change in the tension of the subject when performing the i-th type of activity, being in a psychophysiological state recognized by an expert psychologist as adequate.

на фиг. 1, е) завершается построение распределения Р(С_0P), (фиг. 1, ж), отражающего исходное психофизиологическое состояние оператора перед работой. Это распределение сравнивается с аналогичным распределением Р(С_0Э), полученным на стадии обучения. Определяется мера сходства этих распределений d (фиг. 1, з), при превышении ею заданного уровня У_P запускается процесс слежения за психофизиологическим состоянием оператора. При d<У_P его допуск к работе на объекте блокируется с объяснением оператору правдоподобной причины, например, несоответствие аутентификатора зарегистрированному в памяти этого объекта. Фактическая причина несоответствия сравниваемых распределений - нахождение работника в неадекватном состоянии (алкогольное или наркотическое опьянение, прединфарктное состояние, возбуждение перед реализацией намерения незаконного изъятия конфиденциальной информации из компьютера, недопустимый уровень утомления и др.). Иными словами, доступ к работе на объекте управления осуществляется при выполнении двух условий: наличие аутентификатора и нахождение работника в адекватном состоянии.Before the start of professional activity, the operator’s admission to the controlled system by the authenticator is implemented, and if the admission has taken place (Fig. 1, d), a signal C _A (t) is generated (Fig. 1, d). At the same time, the operator’s cardiogram is hidden (Fig. 1, e) and the process of its analysis for stationarity is started. On the first violation of stationarity (moment

in FIG. _1e ), the construction of the distribution P (C _0P ), (Fig. _1g ), which reflects the initial psychophysiological state of the operator before work, is being completed. This distribution is compared with a similar distribution of P (C _0E ) obtained at the training stage. A measure of the similarity of these distributions d is determined (Fig. 1, h), when it exceeds a predetermined level, _P starts the process of monitoring the psychophysiological state of the operator. When d <Y _P, his access to work on the object is blocked with an explanation to the operator of a plausible reason, for example, the authenticator does not match the memory registered in this object. The actual reason for the mismatch between the compared distributions is that the employee is in an inadequate state (alcohol or drug intoxication, pre-infarction state, agitation before realizing the intention to illegally remove confidential information from the computer, an unacceptable level of fatigue, etc.). In other words, access to work at the control object is subject to two conditions: the authenticator is in place and the employee is in an adequate condition.

, формируется распределение параметров на интервале

;

, т.е. P(C_1P) (фиг. 1, и). Сравнение его с эталонными Р(С_1Э), Р(С_2Э), Р(С_3Э) и т.д. по максимуму сходства d устанавливается, в каком состоянии i в указанный период находился оператор, а по дисперсии

- уровень колебания напряжения нервной системы при выполнении i-го задания.In the process of accomplishing the tasks, the same parameters C _iP (t) are hidden, which are recorded in the learning process (Fig. 1, e), and the algorithm for finding the time instants of changing the properties of the registogram is implemented (Fig. 1, f). The appearance of the first decision about the appearance of such changes at the moment

, the distribution of parameters is formed on the interval

;

, i.e. P (C _1P ) (Fig. 1, and). Comparing it with the reference P (C _1E ), P (C _2E ), P (C _3E ), etc. by the maximum of similarity d, it is established in what state i the operator was in the indicated period, and by the variance

- the level of fluctuations in the tension of the nervous system when performing the i-th task.

To test the operability of the proposed technical solution, it is necessary first of all to select the parameters (signs in the terminology of the theory of pattern recognition) cardiograms that describe the psychophysiological state of the subject.

It is believed that a set of electrocardiographic parameters can serve as an objective indicator of a person’s emotional stress [Simonov V.P. The emotional brain. - M.: Science, 1981. - 167 p.]. In 1996, the European Society of Cardiology and the North American Electrophysical Society proposed standards for measuring and interpreting heart rate variability with recommendations for the clinical use of this source of information about human conditions [Heart Rate Variatility of measurement, physiological interpretation ant clinical use // Europlan Heart Journal, 1996, k. 17, pp. 354-381]. In domestic literature, experience in the study of HRV is systematized in the methodological recommendations set forth in [Bayevsky PM, Ivanov GG, Chireikin L.V. To help a practitioner. Analysis of heart rate variability using various electrocardiographic systems. Methodical recommendations // Bulletin of arrhythmology. - 2001. - No. 24. - S. 65-87]. To date, to describe the psychophysiological state of a person by the flow of cardio intervals, you can use the stress index of regulatory systems or the stress index SI, the standard deviation of the RR intervals SDNN, the average value of the RR intervals M, the stress index IS proposed in [Pat. 2246251 RF, IPC А61В 5/00. A method for assessing the psychophysiological state of a subject by heart rate [Text] / Godunov V.A. et al.], the correlation coefficient between the arrays RR _N and RR _{N + 1} C ₁ and others (more than 20 parameters in total).

The results of studies on the creation of systems for identifying the psychophysiological state of a subject using each of the parameters listed separately allowed us to conclude that they all belong to the class of uninformative ones, i.e. not providing an acceptable level of reliability (probability of errors of the 1st and 2nd kind) of identification. A similar situation is typical for the tasks of biometric identification of subjects according to the dynamics of subconscious movements [Epifantsev B.N. Hidden identification of the psychophysiological state of the human operator in the process of professional activity. - Omsk, SibADI, 2013, 198 p.]. The existing view of the problem is the decision to take based on the results of the analysis of the totality of the available signs [Ushmaev OS Problems of the use of biometric identification in integrated information protection // Information Technologies, 2009, No. 12, p. 34-38]. Ideological basis: each sign carries insignificant information about the psychophysiological state of the subject, but the use of the totality of signs in principle allows us to obtain acceptable probabilities of errors of the 1st and 2nd kind.

Various researchers use from 3 to 20 indicators of heart rate variability in their work. In solving the problem of identifying the psychophysiological state, it was necessary to increase this number. Turning to FIG. 2, and on which the pairs of coordinates in successive RR-intervals RR _i and RR _{i + 1} reflected the real picture of the test operation of the heart in different psychophysiological states. For a particular subject, the image of the cardiogram in the indicated coordinates changes shape under various psychophysiological conditions of the subject.

To describe an image of a similar type, we find its center M _RRi , M _{RRi + 1} and divide it into 25 regions (Fig. 2b): we deviate from the center in each direction by the standard deviation G (M). We will number the areas from 1 to 25, we will take the percentage of points in each of them for additional signs.

During the study, all signs were recorded, according to the experiment, histograms of the joint distribution were constructed, according to the χ ² - Pearson criterion, the type of the standard distribution was determined, the parameters of which were taken as the reference ones.

It is known that the number of images presented at the training stage should significantly exceed 20n, where n is the number of signs [Erosh I. L., Sergeev M. B., Soloviev N. V. Image processing and recognition in preventive security systems. - SPb .: SPbGUAP, 2005. - 154 p.]. This means that in order to build a standard of psychophysiological state of “alcohol intoxication”, the subject must take alcohol about 1000 times.

To demonstrate the operability of the proposed technical solution, let’s step back from the “optimal signal reception” scheme in terminology [V. Tikhonov Optimum signal reception. - M .: Radio and communications, 1983. - 320 p.], Simplifying the following operations.

At the “training” stage, instead of a joint probability distribution of the values of the attributes, we will construct one-dimensional probability distributions for each of them. In FIG. 3 and -d are examples of a few number of parameters of such distributions in three psychophysiological states: original state received instructor for the normal state, after taking 30 mg sorokogradusnuyu alcohol state after taking sleeping pills: IP, A, C.

Comparison of the current (their) distribution (s) of the probabilities of features with the reference can be implemented in several ways. The classic of them is based on obtaining an estimate of the proximity of the compared vectors by the distance between them. The easiest is to estimate the Euclidean distance

where a _i , b _i are the values of the ith feature of each of the vectors, each of which in the problem under consideration is the moment of the compared probability distributions. The distances of Mahalanobas, Tanimoto, Fisher, etc. are more complicated in design.

In another embodiment of the method, it is possible to exclude the stage of constructing probability distributions of features at the stage of professional activity, and replace the operation of allocating the reference distribution closest to the current one by replacing the operation of extracting the reference distribution that is most adequately described by the current set of features. This operation requires an algorithm.

Let A _j be an identifiable attribute, H _i be a hypothesis associated with an identifiable image, i be a hypothesis number, N be a number of hypotheses, k be a realization number of identification data generated by an identifiable image (and describe it), j be an identifier attribute number, n is the total number of features, P (A _j / H _i ) is the posterior density of the probability distribution of hypotheses.

Upon receipt of the k-implementation, a set of probabilities P _k (A _j / H _i ) is formed. The first hypothesis corresponds to the probabilities P _k (A ₁ / H ₁ ), ..., P _k (A _j / H ₁ ), ..., P _k (A _n / H ₁ ), i-th: P _k (A ₁ / H _i ), ..., P _k (A _j / H _i ), ..., P _k (A _n / H _i ), etc. Then the probability of the hypothesis H _i in all respects can be estimated by the formula

In the simplest case, the decision on the number of the correct hypothesis i is made by the maximum value q _k (H _i ). More advanced options are based on statistical decision theory.

The second algorithm for solving the problem is based on a modified Bayesian hypothesis formula

where Р (Г _ij ) are the a priori probabilities of the i-th hypotheses upon receipt of the j-th attribute. If they are unknown, we will consider

P (T _ij ) = N ^-1 for j = 1;

P (G _ij ) = 0.5 (P (G _ij ) + P (H _i / A _j ), with j = 2;

, при j>m.

, for j> m.

For m = 1, this formula transforms into the classical Bayes formula.

The third algorithm for solving the problem can be built on the general theorem on the repetition of experiments. According to this theorem, the probability that the event H _i in n dependent experiments appears d times is equal to the coefficient at z ^d in the expression of the generating function:

where q _ij = 1-p _ij , p _ij is the probability of the i-th hypothesis upon receipt of the j-th attribute. The coefficient at z ° is the probability P _0n that the hypothesis under consideration is not true. Then, with the probability Q _0n = 1-P _0n, it is recognized as valid.

To compare the above algorithms by the probability of identification errors provided by them, a computational experiment was carried out under the following conditions: number of signs 7, maximum number of hypotheses (psychophysiological states - 50, k = 100).

The parameters of the posterior probability generators for each feature were determined on the basis of the condition: the probability distribution densities P (A _j / H _i ) obey the normal law. To generate feature values in accordance with their distributions, the Monte Carlo random value method was used.

Based on the results of the study, it was concluded that the considered algorithms are approximately the same in terms of the probabilities of errors of the 1st and 2nd kind, and in subsequent experiments, an algorithm based on a modified Bayes formula was used.

For each hypothesis, 30,000 realizations were generated. With the significance parameter α = 0.01, the reliability of the conclusions made about the probabilities of identifying the psychophysiological conditions of the operator (IS, A, C) is 0.974. The obtained average value of the probability of the correct identification of dangerous conditions for a group of subjects from 20 people turned out to be 93.7%.

The experiments were carried out on a car simulator (Fig. 4) and included driving a vehicle in various psychophysiological conditions.

Claims (1)

A method for covert assessment and monitoring of dangerous psychophysiological conditions of a human-machine systems operator in the course of professional activity, based on comparing the cardiogram parameters when hiring an operator or recertifying knowledge on a simulator object when performing various types of activities and on a real object, for which the parameters are first evaluated in the initial state - before the start of activity, then - during professional activity at the simulator object, then - at the real object, according to the result there, comparisons of parameters make a decision on the psychophysiological state of the operator, characterized in that the cardiogram is recorded secretly from the subject, while stationary sections are distinguished on the cardiogram, followed by the construction of probability densities of heart rate variability (HRV) indicators on each of them, which are taken as standards an adequate psychophysiological state of the operator in the initial state and at the stage of performing typical tasks of professional activity on an object an e-simulator, in this case, before starting a professional activity on a real object, the operator who starts work is first authenticated with the simultaneous registration of his cardiogram with the allocation of stationary sections on it and the construction of the distribution densities of the HRV indicators recorded at the previous stages on each of them, determine the degree of similarity of the obtained probability distributions of these indicators in the initial stationary section with the distribution-standard obtained earlier in the initial state op an herator before starting work on a simulator object, while admission to work on a real object is subject to two conditions: the presence of an operator authenticator and an adequate psychophysiological state of the operator, which is established if the mentioned degree of similarity of the probability distributions of HRV indicators exceeds a predetermined level; otherwise, the operator’s access to work on a real object is blocked;
Further, in the process of the operator’s professional activity at a real object, when registering a cardiogram, they compare the probability distributions of HRV indicators obtained at each subsequent stationary section with the corresponding previously obtained standard distributions, select the distribution closest to the current one, by which it is established in which psychophysiological state at the moment the subject is located, and if the degree of similarity of the distributions does not exceed a predetermined level, ascertain the operator’s driving in an inadequate condition - dangerous for continuing professional activities.

Images