# Method for estimating trustworhiness of tolerance parameter control on basis of measurements results

FIELD: measuring technologies.

SUBSTANCE: method includes setting tolerance for controlled parameter, measuring physical value, associated with said controlled parameter, with numeric characteristic of its value, then measured value is compared to its tolerated values (tolerances for controlled parameter), and decision concerning level of match of measurement results to tolerances for parameter is taken, when determining tolerance for controlled parameter an affiliation function is set for phrase "parameter on basis of measurements in tolerance", and during taking of decision trustworthiness of phrase is evaluated, expressed in non-precise measure, as value of affiliation function, matching value of measured parameter.

EFFECT: higher trustworthiness.

2 dwg

The known method tolerance control parameters based on the measurement results, which are as follows:

- set the tolerance on the monitored parameter;

- measure the physical quantity corresponding to this monitored parameter, with a quantitative characteristic values;

- compare the measured value with the valid values (tolerances on the controlled variable);

- make decisions about the degree of compliance measurements tolerances on parameter estimation accuracy, expressed in probabilistic measure.

The disadvantage of this method is the inability to assess the validity of the conditions when the probabilistic model of the object of control is not adequate due to lack of statistical information.

The technical challenge is to increase the reliability of the control parameters in the absence of sufficient statistics.

This object is achieved in that in the method tolerance control parameters based on the measurement results, namely, that set the tolerance on the monitored parameter; measuring a physical value corresponding to this monitored parameter, with a quantitative characteristic values; comparing the measured value with the valid values (tolerances on to the control parameter) and make decisions about the degree of conformity of the measurement results the tolerance parameter; according to the invention, when determining admission to the controlled parameter specify the membership function statements “option on the measurement results in tolerance”, and when deciding assess the credibility of statements, expressed in fuzzy as the value of the membership function corresponding to the value of the measured parameter.

The invention is illustrated by the following drawings.

Figure 1 presents a probabilistic model of the control object in the form of laws of distribution of density of probability of the measured parameter and the uncertainty of the measurement means, the upper X_{in}and lower X_{n}maximum permissible values of the parameter, which may be operable object and the top Y_{in}and lower Y_{n}the limit values of the measurements of the parameter characterizing the error measurement means and control the tolerance parameter [1].

Figure 2 shows a variant of the membership function defined on the field of tolerance controlled parameter [X_{n}, X_{in}]. While the arrows show the conversion of the measurement result in the corresponding value of membership function μ(X_{ISM 2})=μ(X_{n}, X_{in}, X_{ISM}). When approaching the values of the measured parameter to the boundary values of the field of tolerance the degree of truth should be iatse,
and on the border of the state of the object it is impossible to say anything definite.

When using the known method, to determine the probabilistic characteristics in some cases there is no possibility of a set of sufficient statistics. As noted in [3]: “the Concept of probability is statistical. The statistical meaning of the term appears multiple implementation of the conditions under which some event, and setting the frequency of occurrence of an event. Characterizing the probability of some number, it is impossible to give the number of different real values and other practical sense than the relative frequency of occurrence of a given event with a large number of experiments. Probabilistic methods are applied in order, bypassing the study of specific phenomena, to apply directly to the laws governing mass phenomenon”.

Therefore, in many practical applications is an issue of feasibility of methods for assessing the reliability as the appropriateness of the expenditure of the available computational and time resources monitoring system. Consider this thesis as applied to the system of control of technical condition of remotely managed objects. First of all, we should note the unsteadiness of such objects running on a separate modes, each of which is its turn, consists of a set of individual algorithms and operations. The total number of States of the object is significantly increased, therefore, the number of models of control. In addition, for such objects, the source of the measurement data are telemetry, characterized by a number of basic features.

First, the essential feature of telemetrically is a multistep transformation process measurement data, implemented in a spatially distributed system with non-stationary components in the conditions of constantly changing factors.

Secondly, the vector of measured parameters is characterized by high dimensionality. However due to the limited bandwidth of the radio channel number telemetering parameters sample size does not exceed 3-5 measurements.

Third, the processing of telemetry may be simultaneously conducted on a group of control objects, and in this case, the processing system is viewed as a queueing system with applications on input from units to tens.

Fourthly, the requirement of immediacy of control leads to the necessity of its automation, which in turn affects the problem of interpretation of testing results, translation quantitative assessments of quality, expressed in natural language. Eventually there Probl the mA conservation of semantic validity.

This poses the problem of feasibility of probabilistic indices of reliability for the decision of tasks of operational control, as the aim of the statistic and its subsequent processing can go to the category transmusicales. An alternative is the use of fuzzy measures of reliability. The degree of the truthfulness of the information will be given by the corresponding membership function.

The described method is implemented as follows. Type of membership function is selected from a known set (see, e.g., [2]) on the processing results of the opinions of experts in this field. The increase in amount of information about the problem at hand allows to reduce the degree of subjectivity and to offer some recommendations on the choice of membership functions.

Let the measurements at the disposal of the researcher has information on control box access and area health controlled parameter. In this case, the membership function can be constructed based on the known number by specifying the values of μ(X_{n}), μ(X_{in}), as well as μ(Y_{n}), μ(Y_{in}). For example, for a symmetric membership function μ(X_{n})=μ(X_{in}), μ(Y_{n})=μ(Y_{in})=A, where A∈[0.5, 1], the level of confidence in the results is the ATA measurements on the boundary of the control field tolerance.
Consider the factors that affect the value of the parameter A. the Analysis shows that when solving problems of control, the most significant of them are the following: confidence probability P_{d}which is defined average (confidence) interval [Y_{n}, Y_{in}] and the measurement error γ [4].

The condence probability is a measure of methodological validity, and to assess the instrumental component you can use the probability of a false rejection rate of P_{lo}. In this case, the truth of the statement “option on the measurement results in tolerance will decrease with the increase of the error of the first kind, which is a function of R_{lo}{f(x), γ, X_{n}, X_{in}, X_{ISM}}. Other factors considered by the influence coefficient of the k≤1. Then based on these a priori data parameter And can be set in the following form: A=k•P_{d}(1-P_{lo}).

Literature

1. Savin S. Kaliev, Nikitin A.A., Kravchenko V.I. Reliability of complex electronic control systems of aircraft. M: mechanical engineering, 1984.

2. Reliability and efficiency in the technique: Ref. Volume 3 / edited by V.F. Utkin, Kryuchkov Y. C. - M.: Mashinostroenie, 1990.

3. Levin BYR Theoretical foundations of statistical radio engineering. - M.: Owls. Radio, 1969.

4. Potapkin A. Application neced the first steps in monitoring the technical condition of aircraft. // Measurement techniques, No. 7, 2002.

A method of evaluating the reliability of tolerance control parameters based on the measurement results, namely, that set the tolerance on the controlled variable, is measured physical quantity corresponding to this monitored parameter, with a quantitative characterization of its value, compare the measured value with the valid values (tolerances on the controlled variable) and make decisions about the degree of conformity of the measurement results the tolerance parameter, wherein when determining admission to the controlled parameter specify the membership function statements “option on the measurement results in tolerance”, and when deciding assess the credibility of statements, expressed in fuzzy as the value of the membership function corresponding to the value of the measured parameter.

**Same patents:**

FIELD: computers.

SUBSTANCE: device has control trigger, random pulse generators, block for forming program of functioning of modeled multimode system, working modes and technological mode blocks, operation time counters, random pulses generators, OR block, orders counters.

EFFECT: broader functional capabilities.

3 dwg

FIELD: computers.

SUBSTANCE: system has nine registers, four address selectors, triggers, AND elements, OR elements and delay elements.

EFFECT: higher speed.

8 dwg

FIELD: automatic control, applicable in systems with excessive quality of transducers, for example, accelerometers, a failure of one of which should not result in a failure of the control system.

SUBSTANCE: the method is based on a periodic check-up of relation between the measured parameters of motion characterizing the correct operation of the transducers, fixation of the moment of failure of the relation, comparison of the readings of the transducers at this moment and at the moment preceding the moment of disturbance of the relation, and determination of the failed transducer by the results of the comparison.

EFFECT: expanded functional potentialities due to possibility of determination of the failed transducer in any excess system.

1 dwg