DE102005043461A1 - Vehicle`s component e.g. exhaust silencer, reliability predicting method, involves using actual and reference loads, number of faults between preset and future time, and correlation between load in test and in future use for prediction - Google Patents

Abstract

The method involves subjecting a component in a test of a load, and measuring a mass for the load as an actual-load which sets the component until a preset time during the test. Another mass is preset as a reference load setting the component between preset and future time. The reliability is predicted using the actual and reference loads, a number of faults between the time, and correlation between the load in test and in future use. Independent claims are also included for the following: (A) a computer program component loaded in an internal memory of a computer to execute a method for predicting reliability of a component (B) a digital storage medium with electronically selectable control signals, which can interact with a programmable data processing equipment to execute a method for predicting reliability of a component.

Description

The The invention relates to a method and a device for prediction the reliability a technical product. Under the reliability is an average Number of errors related to a load of the product during its Use understood.

One Method with the features of the preamble of claim 1 from WO 2005/033649 A1. A component is subjected to a trial. The experiment will be carried out according to a previously determined test program. This test program sets the sequence and the respective duration of each individual Try it. Preferably, individual tests are considered to be time-consuming Experiments performed, so that the each set test duration of a longer service life of the component equivalent. It is calculated which reliability for the component due to the Predictable experiment is detectable.

• – K. Denkmayr: „Der AVL Reliability Engineering Prozeß für die Motor- und Antriebstangentwicklung", VDI-Bericht 1713, 2002, S. 27–32,
• – K. Denkmayr et al.: „Die Load-Matrix – der Schlüssel zum „intelligenten" Dauerlauf, Motortechnische Zeitschrift (MTZ), Heft 11/2003, S. 924–928,
• – T. S. Gates & M. A. Grayson: „On the Use of Accelerated Aging Methods for Screening High Temperature Polymeric Composite Materials", Americ. Inst. Aeronautics Astronautics (AIAA) 99–1296, pp. 925–935

beschreiben.Reliability prediction methods are also used in

• K. Denkmayr: "The AVL Reliability Engineering Process for Engine and Drive Train Development", VDI Report 1713, 2002, p. 27-32,
• - K. Denkmayr et al .: "The load matrix - the key to the" intelligent "endurance run, Motortechnische Zeitschrift (MTZ), Issue 11/2003, pp. 924-928,
• TS Gates & MA Grayson: "On the Use of Accelerated Aging Methods for Screening High Temperature Polymeric Composite Materials", Americ. Inst. Aeronautics Astronautics (AIAA) 99-1296, pp. 925-935

describe.

In DE 19713917 A1 A method is described for determining reliability characteristics of a technical installation. For this purpose, a generic FMEA table, ie valid for several possible configurations of the plant, is generated first. FMEA stands for "Failure Modes and Effects Analysis." Several generic tables are created from this generic table.

Out US 2004/0044499 A1 discloses a method to be continuous a technical system, eg. As a motor to monitor. With the help of stochastic Methods are derived from the results of continuous monitoring Reliability characteristics derived.

Of the Invention is based on the object, a method with the features of the preamble of claim 1 and a device with the features of the preamble of the claims 14 and 15, which provides early predictions of the reliability enable.

The The object is achieved by a method having the features of the claim 1 and a data processing system with the features of the claim 14 and a computer program product having the features of the claim 15 solved. Advantageous embodiments are specified in the subclaims.

• – der Anzahl von Fehlern bezogen auf die Belastung während der Benutzung und
• – der Anzahl von Fehlern bezogen auf die Belastung während des Versuchs.

The method predicts as the reliability an average number of errors related to a load of the product during its use. The product is subjected to stress in at least one trial. This trial extends over a given period of time. Determined is a computer-evaluable relationship between

• The number of errors related to the load during use and
• - the number of errors related to the load during the experiment.

• – der Ist-Belastung,
• – der gezählten Anzahl von Fehlern bis zum ersten Zeitpunkt,
• – der Soll-Belastung,
• – einer zu erwartenden Anzahl von Fehlern zwischen dem ersten und dem zweiten Zeitpunkt und
• – des Zusammenhangs

vorhergesagt.A first and a second time of the time period are specified, wherein the second time is after the first time. Measured as an actual load is a measure of the load that the product is exposed to during the test until the first time point. Furthermore, it is counted how many errors occurred in the experiment until the first time. As a target load, a measure is given as to which load the product will be exposed to during the test between the first time and the second time. The reliability is being used

• - the actual load,
• - the number of errors counted until the first time,
• - the target load,
• An expected number of errors between the first and second times and
• - the context

predicted.

The Procedure says the reliability using a measured actual load and a counted number from actually occurred errors before.

The Invention shows a way to the predicted reliability systematically calculate. In contrast to WO 2005/033649 A1 known method flow in the prediction of reliability actual Values that have been measured up to the first time and not only planned values determined before the start of the experiment.

Consequently allows it the method according to the invention, the continuation of the experiment after the first time when needed reschedule, z. For example, by supplementing the experiment with additional individual experiments.

in the Contrary to known methods provides the inventive method a reliability characteristic already at the first time and not after the end of the experiment. This reliability characteristic but does not refer to the first time, but to one future second time. So the procedure says at the first time using future experimental history, future reliability at the second time before.

The inventive method let yourself at any first time, and provides a prediction for each any second time.

The Procedure shows a systematic way to get started to decide if the further experimental program until the second Time can be continued as planned or whether to the first Time a rescheduling is required. A rescheduling can in particular then be required if the procedure at the first time a reliability which is less than a given target reliability is. Thus, can be the continuation of the measured values available up to the first time dependent do. Such a rescheduling can for example consist of that to first time the test program is modified so that the product until the second time a higher load than originally planned is subjected.

Preferably the method provides a reliability prediction that is detectable at the second time point will be. This prediction is valid at least if the test program is carried out until the second time as planned at the first time. The procedure makes it possible thus avoiding "over-testing" the product more than necessary Is tested.

Of the Relationship between the number of errors related to the load while the use and the number of errors related to the load while of the experiment is preferably in the form of a recruitment factor. The Using such a factor of retrieval - or some other form of context - makes it possible to interpret the individual experiments as time-consuming experiments. This saves time and allows it, the reliability earlier predict.

Farther allows it is the embodiment with the context, in the experiment another Measure of the burden to use as the load measurement on which the reliability is related. The Raffungsfaktor then has a dimension, z. B. 1000 driving km per oscillation process.

Preferably the relationship is determined empirically in two preliminary experiments. This way of identifying the context saves you from an analytical model for set up and validate the loads in the trial and in the use to have to.

The Procedure can be also use to reverse a value for reliability pretend and determine at first, if at all and if so at what second time this predetermined reliability value is predicted.

The Procedure can be also for that use to recalculate, at what time the attempt should have started to a desired one verifiable reliability predict.

in the The following is an embodiment the invention described in more detail with reference to the accompanying figures. Showing:

1 , an exemplary stress-time function for a component;

2 , a time course of the planned accumulated equivalent distance before the start of the experiment;

3 , the time course of the detectable error rate before the start of the experiment according to the planning of 2 ;

4 , the time course of the accumulated equivalent distance according to a modified planning, the first time is August 2005, and dashed lines the time course of the equivalent distance of 2 ;

5 , the time course of the detectable error rate according to the equivalent distance of 4 ;

6 , the timeline of the actual and planned equivalent journey in accordance with alternative modified planning, the first being in August 2005;

7 , the time course of the detectable error rate according to the time course of the equivalent distance of 6 ;

8th , the time course of the accumulated equivalent distance according to the planning of 6 as well as dashed lines the time course of the equivalent distance of 4 with the first date in November 2005;

9 , the time course of the detectable error rate according to the equivalent distance of 8th , where an error has occurred, and dashed to compare the temporal of the detectable error rate of 5 ,

in the embodiment is the term "error" a deviation understood the behavior of a component of the required behavior. According to DIN EN ISO 8402, a viewing unit has an error, if one the quality characteristics the viewing unit assumes an undesirable value.

in the Below is a "degradation" of a component a Deviation of a quality value understood by a desired value or nominal range. In particular, can the degradation is a physical and / or chemical change be the component. examples for Degradations are wear, Aging, corrosion, structural change and deposits. If the degradation of a component has progressed so far is that the Component does not fulfill its required function at all or only to a limited extent and / or Occasion for a complaint offers, so there is a mistake. For example, the insert cause a component that its surface corrodes, what a degradation is. Is the corrosion layer thicker than a predetermined threshold and / or the corroded area greater than a predetermined barrier, so the component has an error.

In the exemplary embodiment, the method is used in each case once for different components of a truck. At least one fault type can occur on each of these components. It is possible that different types of errors occur on the same component. Before applying the method, relevant component-fault type combinations are selected. Preferably, an FMEA is used to determine in advance which component-fault-type combinations are relevant and therefore to be taken into account. Such FMEA is z. B. from WO 2005/033649 A1 and DE 19713917 A1 known. The process is then executed once for each selected component-fault type combination. It is possible to generate different predictions for different component-error type combinations.

In order to apply the method to a selected component-defect-type combination, several components are tested in the actual experiment. For example, several identical components are tested. The experiment is carried out over a longer period of time. The experiment includes n individual experiments in which each one of the same components is tested. In some of these n individual tests, one copy of the component is installed in each truck. These trucks with the built-in components are driven over longer journeys in the test and / or subjected to extensive stress on test benches. In other individual tests copies of the components in test benches are subjected to the loads which are similar to the loads during the journey. This z. B. components installed in a trial.

The surface of the component "silencer housing" can corrode, which represents a degradation of the component. This may be the fault "silencer housing visible corroded "lead.

Internal parts of the muffler may wear out due to the load to which the truck is subjected. Abrasion or enlargement of a bore are examples of degradation. If the wear is so great that the muffler when driving audible noise, eg. B. rattling, caused, so there is a mistake. In particular, vehicle-side shrinkages cause material fatigue, which can lead to wear. 1 shows an example of a stress time function for a muffler. The time is plotted on the x-axis and on the y-axis a time-varying vibration acceleration in [g].

• – Auf Fahrbahnen mit hoher Schwingungsanregung werden n1 Bauteile binnen einiger Monate jeweils mehrere tausend km erprobt.
• – Mit n2 Bauteilen werden auf Prüfständen Dauerläufe durchgeführt, bei denen binnen sechs Monaten pro Bauteil über 100.000 km Fahrtstrecke zurückgelegt werden.
• – In Lkws werden n3 Bauteile eingebaut. Die Lkws werden unter denselben Bedingungen wie in der späteren Benutzung gefahren, dabei werden über zwei Jahre hinweg jeweils mehrere 100.000 km gefahren.
• – In einer Versuchsanordnung werden Prüfstandsläufe mit n4 Bauteilen durchgeführt. Hierbei werden zwei Monaten jeweils mehrere tausend Lastwechsel durchgeführt.

For example, within a year, the following individual experiments are performed to investigate the error "muffler causes noise":

• - On lanes with high vibration excitation, n1 components will be tested several thousand kilometers within a few months.
• - With n2 components, endurance tests are carried out on test benches in which over a distance of more than 100,000 km per component is covered within six months.
• - In trucks, n3 components are installed. The trucks are driven under the same conditions as in later use, with several 100,000 km being traveled over two years.
• - Test rig runs with n4 components are carried out in an experimental setup. In this case, several thousand load changes are carried out for two months each.

Around to test how strong the degradation "corrosion of the silencer housing" from the loads depends and how many times the error "silencer housing visible corroded "occurs, be z. B. Test bench runs and Trial runs in a very humid environment, eg. B. a climatic change chamber, carried out.

In further individual tests and validation steps become burdens simulated. Therefor a simulation model is used, which is the loaded component imitating on a data processing system. In the simulation it replicates which degradations trigger these loads. error are counted, by evaluating simulation results.

The Loads that the components in the trials including the Simulation are subject to irregular overlays of different vibrations.

The when attempting measured or calculated values in the simulations are used to set values for the reliability a selected one Component-fault-combination while to predict the productive use. Every prediction relates So on a component and a type of error.

The reliability will be a measure of the burden, which is exposed to the component related. As a measure of the load on the component is used in this example both during the trials as well as during his Use travel route in [km] used a truck with the component travels. Instead of the route can also be z. As the operating hours or use the number of load changes as a measure of the load.

Of the Experiment is preferably carried out as a time trial. in the The trucks are over trying rough terrain with many uneven bumps Dangers that cause vibrations. During later use The trucks are predominantly over flat stretches, z. B. removed roads, driven.

Therefore, in at least two preliminary experiments, a gathering factor r is determined. This refraction factor r is then used in the actual experiment. The gathering factor is the ratio of the degradation of the component by the load during the test and the degradation by the load during later use. If D1 is the degradation in the experiment and D2 is the damage in use, then r = D1 / D2. This refraction factor r usually varies from component-fault type combination to Component of error combination. The rapping factor r refers to a type of test and may vary from one type of trial to another.

If as a measure of the burden the traveled Travel distance is used, then the rapping factor r following Meaning: A degradation of the component due to a load a journey of x km in the experiment corresponds to a degradation through a journey of rx km during later use. To put it simply: x km of the experiment weigh as much as rxkm while later use. The product r · x km is hereinafter referred to as the equivalent driving distance designated.

in the In the first preliminary test, several trucks with the components become the same Loads exposed as in the actual experiment. For example These trucks are driven on rough routes in the first preliminary attempt. In the second pre-trial several trucks with the components under the same conditions as in the later use. For example These trucks are in the second pre-trial mainly over developed streets hazards.

In In both preliminary tests, at least one parameter for the load, the component in the actual trial and during the Use is exposed, predetermined or determined. This load characteristic can the one introduced above Measure of the burden, that's the reliability refers to. For example, the load characteristic is accurate like the strain measure the covered Route.

Preferably is the load characteristic, however a parameter that stronger with the degradation and thus the error rate is correlated and is harder to measure than the load measurement. For example, the load characteristic is the number of operations, which trigger the degradation. Or the load characteristic characterizes the driveway, z. B. a straight ahead on a developed road, a mountain drive on developed road or drive on a dirt road. Two more examples of the load characteristic are the engine speed of the truck and the relative humidity in the truck Surrounding the muffler. The component belongs For example, to the engine or for other reasons, the speed-dependent vibrations exposed to the engine and the humidity. Is possible, specify several load characteristics, z. B. engine speed and distance.

Of the Range of values of at least one load characteristic divided into k subdomains Tb_1, ..., Tb_k. For example, the relative Humidity divided into the k = six ranges 0-20%, 20% -40%, 40% -60%, 60% -80%, 80% -90% and over 90%. Farther if a reference value is given, z. B. the residence time in the climate change chamber, the route or the journey time.

in this connection As a rule, the at least one parameter for the load varies. It is measured to what proportion n_p (1) is the load parameter in the first preliminary test assumes a value in the subregion Tb_p (p = 1,..., k). The shares are related to the reference value. A proportion n_5 (1) of 23% means that the load characteristic is in the first preliminary attempt over 23% of the total travel distance or travel time in subarea Tb_5 was. For example, the relative humidity was over 23% the distance in the range Tb_5 between 80% and 90%. Thus, will determined a load collective to which the components in the first pre-trial be subjected.

at becomes two load characteristics correspond to a proportion n_p, q (1) measured, to which the first load characteristic Kg_a a Value in subsection Tb_p (a) and the second load characteristic Kg_b a Value in subregion Tb_q (b).

Corresponding is measured, to which proportion n_p (2) (p = 1, ..., k) the load characteristic in the second Pre-trial adopts a value in subsection Tb_p.

Farther becomes a parameter for the degradation specified or determined the component. The wear on the Component occurs due to the load, or occurring on the component Tension are examples of this characteristic.

Furthermore, an average value d_p (p = 1,..., K) for the degradation of the component is measured or calculated on the basis of a degradation model. This value d_p is the degradation that occurs on average when the load parameter has a value in the range Tb_p (p = 1,..., K). For example, d_5 is the average corrosion of a given component at a relative humidity of between 80% and 90%. These values d_p (p = 1, ..., k) are during the actual experiment and during the loading Use the same, so they apply to each individual experiment and for use. This degradation can z. B. be a measure of the relative degradation.

The corresponding measurements and / or calculations for the same degradation characteristic, the same be at least one load characteristic and the same reference carried out in a second preliminary experiment. In the second preliminary attempt the component becomes stress during exposed to use. In the actual experiment and also in the first one Pre-trial will for example, the truck over a much longer journey or driving time in high engine speeds than in use and also in the second preliminary attempt or while a considerably longer one Time exposed to high humidity.

The degradation D1 in the first preliminary test is preferably calculated according to the following calculation rule:

Accordingly, the degradation D2 in the second preliminary test is preferably calculated according to the following calculation rule:

berechnet.The gathering factor r is then calculated according to the calculation rule

calculated.

Of the Raff factor r depends from the component fault type combination so it can vary from combination to combination. He can also vary from single trial to single trial. Let n be the number of the individual tests, let m be the number of selected component-fault type combinations. Then, with r [i, j], the convolution factor for the ith individual test (i = 1, ..., n) and the jth selected Combination (j = 1, ..., m).

Preferably becomes the reliability of a component related to a fault type with the help of the error rate specified. This error rate is set to a specified period, z. As a year, and in defective components per one million Components ("parts per million, ppm) specified. An error rate of 100 ppm per year means that on average to 100 of 1 million identical components that are simultaneously are in use within one year, an error of the type of error occurs.

The error rate is related as follows to the distance traveled until an error occurs on average:

in this connection dist_a denotes the average distance traveled in [km], the a lorry covers 1 year during the period under consideration, and MTTF ("mean time to failures ") mean travel distance in [km] until an error occurs. The Characteristic MTTF depends on the component-fault type combination, hence the name MTTF [j] for used the MTTF value of the jth combination.

When Measure of reliability let yourself also directly use the middle route MTTF.

Preferably, a first quantity indicative of the load of the product during its use and a second size indicative of the load of the product during the trial are used. In this embodiment, the first size used is the distance traveled during use in [km]. This route is in the use of the component the route, which the truck covers with the component.

When second size becomes in this example, in most individual experiments, the route used that - ever after design of the individual test - the truck travels or the on trial or simulated in the simulation. In some individual experiments the number of load changes is used as second size.

• – der Belastung während des Versuchs, gekennzeichnet durch die zweite Größe, und
• – der Belastung während der Benutzung, gekennzeichnet durch die erste Größe.

The Raffungsfaktor is in this embodiment, the quotient

• The load during the test, characterized by the second size, and
• - the load during use, characterized by the first size.

• – Ein Schlechtweg-km, das ist ein km Fahrtstrecke auf einer Fahrbahn mit hoher Schwingungsanragung, entspricht 50 Äquivalent-km. Als Raffungsfaktor r wurde also r = 50 ermittelt.
• – Ein km im Dauerlauf entspricht einem km während der späteren Benutzung, also ist r = 1.
• – Für die Versuchsfahrten, die unter denselben Bedingungen wie die spätere Benutzung durchgeführt werden, gilt ebenfalls r = 1.
• – Für den Prüfstandslauf mit der Anzahl Lastwechsel als der zweiten Größe wird ein Raffungsfaktor r von 5 Äquivalent-km pro Lastwechsel ermittelt.

For example, the following factors are calculated for the component-error type combination "muffler causes noise":

• - One bad-km, which is one kilometer of driving on a high-vibration track, equals 50 equivalent km. The rapping factor r was r = 50.
• - One kilometer in continuous running corresponds to one kilometer during later use, so r = 1.
• - For the test runs carried out under the same conditions as the later use, r = 1 also applies.
• - For the test bench run with the number of load changes as the second size, a recovery factor r of 5 equivalent km per load change is determined.

berechnet.In the i-th individual test, the travel distance dist [i, j] (τ) is covered by the component of the j-th component-fault type combination up to a point in time τ (i = 1,..., N). The load in the test over the distance dist [i, j] (τ) for the jth component-error type combination corresponds to an equivalent load dist_a [i, j] (τ) = dist [i, j] (τ) r [i, j] during use. The total load in the experiment therefore corresponds to an equivalent distance dist_ges_ä [j] (τ) until the time τ during the use of the component of the jth combination. This equivalent distance is calculated according to the calculation rule

calculated.

When first size can be For example, instead of the route, the number of events use that during the use of the component have occurred, which is a degradation cause the component and therefore cause an error of the type of error can. Such events are z. B. vibration processes, inputs and Abstellvorgänge, load changes, Number of climatic changes, the occurrence of temperature peaks or the Contact of the component with corrosive, corrosive or extremely hot or cold liquids.

When second size can be according to the number of events in the actual attempt use.

During the Trial is measured or otherwise determined, how many errors which types of errors occur on the tested components. in this connection the calculated numbers of errors are added to a type of error. With f_ges [j] (τ) becomes the total number of all errors of the error type of the jth combination denoted by the time τ in total on all copies of the component of the jth combination, that is, summed over all n individual tests.

The Method is used in the embodiment for every Component-fault-combination repeatedly applied. Each application takes place at a first time τ1. This time τ1 acts as the first date of the claims. As stated above, were the tested components up to the time τ1 total of a load exposed to an equivalent driving route of dist_ges_ä [j] (τ1). The first time τ1 for example, in the first half of the period over which the experiment extends.

With f_ges [j] (τ2; τ1) becomes the total number of all errors of the error type of the jth combination designated according to a Prediction in the period of τ1 expected to τ2 become.

In an embodiment It is assumed that between the times τ1 and τ2 no error the type of error occurs, that is f_ges [j] (τ2; τ1) = 0. This assumption is made to predict the reliability for the Case, that no Errors occur.

In another embodiment will be first dependent from the planned load the increase in the degradation of the component in the period between τ1 and τ2 predicted. This planned load is dependent on in this example the equivalent route or another load characteristic. Depending on this degradation increase is predicted how many mistakes in the period between τ1 and τ2 occur become. For both predictions are preferably models used. The first Model describes the relationship between the load and the Degradation, the second model the relationship between the occurrence of mistakes and degradation.

In each case, a prediction value for the reliability which is achieved when the test is continued up to several second times is calculated. Let τ2 be such a time. Given and used for the prediction is for each of the n individual tests, the target load that will be exposed to the respective component from the time τ1 to the time τ2. τ1 is preferably the current time, τ2 is a variable time in the future. In this embodiment, this is a planned equivalent distance of dist_plan [i, j] (τ2; τ1) for the i-th single trial (i = 1, ..., n). If, therefore, in all individual tests, the actual distance covered by the time τ2 corresponds to the planning at the time τ1, then for i = 1, ..., n and j = 1,..., M: dist_ [i, j] (τ2) = dist_plan [i, j] (τ2; τ1) + dist [i, j] (τ1)

For the equivalent distance dist_ä applies accordingly: dist_a [i, j] (τ2) = dist_plan_a [i, j] (τ2; τ1) + dist_a [i, j] (τ1)

und

sonst. Die Schranke Δ beträgt beispielsweise 0,1. Die Fallunterscheidung sichert eine stabile Prognose auch für den Fall, daß erst wenige Prozent des i-ten Einzelversuchs absolviert wurden.In one embodiment, an increase in the distance traveled is assumed proportional to the elapsed time. Thus, dist_plan_a [i, j] (τ2; τ1) becomes according to the calculation rule dist_plan_a [i, j] (τ2; τ1) = c * (τ2 -τ1) calculated. In the following, T1 [i, j] (τ1) denote the actual duration of the i-th individual test until the time τ1 and T [i, j] the planned total duration of the ith single test (i = 1, ... , n) for the jth combination (j = 1, ..., m). The equivalent distance to be traveled in total according to the planning from the beginning to the completion of the i-th individual experiment to test the j-th combination is called dist_plan_a [i, j]. The proportionality factor c is calculated according to the following calculation rule:

and

otherwise. The barrier Δ is for example 0.1. The case distinction ensures a stable prognosis even in the case that only a few percent of the i-th individual experiment was completed.

berechnet.With dist_plan [i, j] (τ2; τ1), the actual travel distance in the ith attempt is called the jth combination. By dist_plan_ges_j [j] (τ2; τ1) is meant the equivalent travel distance which, according to the planning, is covered in the time interval from τ1 to τ2 in all individual tests for the jth combination (j = 1, ..., m). This entire equivalent distance is calculated according to the calculation rule

calculated.

The total equivalent distance dist_ges_ä [j] (τ2), which has been covered and will be covered according to the schedule until time τ2, is accordingly dist_ges_a [j] (τ2) = dist_ges_a [j] (τ1) + dist_plan_ges_a [j] (τ2; τ1)

The error behavior of a component is preferably determined by a parametric error occurrence model, for example, by a Weibull distribution, described statistically. The Weibull distribution was first introduced in W. Weibull: "A Statistical Distribution of Wide Applicability," J. Appl. Mechanics, Vol. 18 (1951), pp. 293-297, which is detailed in RA Abernethy: "The New Weibull Handbook ", 4th ed., 2000.

It has a distribution function F (τ) = F [α, β] (τ) = 1 - exp [(- λ · τ) ^β ] which depends on the two parameters λ and β. Here, the parameter λ defines the error rate and the parameter β ("slope") the form of the distribution function The probability that an error of a specific type of error has occurred at a component up to the time τ is just F (τ) Thus, for a year and one type of error, F (1 year) is F. The MTTF of a Weibull distribution is MTTF = 1 / λ * Γ (1 + 1 / β), where Γ (x) indicates the value of the gamma function the position x, and it holds for every natural number n:

If β = 1, the Weibull distribution goes into the exponential distribution. These has the distribution function F (τ) = F [λ] (τ) = 1 - exp [- (λ · τ)] with temporal constant error rate λ> 0.

berechnet. Wie oben erwähnt, ist f_ges[j](τ1) die gesamte Anzahl aller Fehler der j-ten Bauteil-Fehlerart-Kombination, die bis zum Zeitpunkt τ1 in allen Versuchen aufgetreten sind. Mit f_ges[j](τ2;τ1) wird die gesamte Anzahl aller Fehler der j-ten Kombination bezeichnet, die gemäß der Vorhersage im Zeitraum von τ1 bis τ2 auftreten werden. Der Nenner kann nie gleich Null werden.In one embodiment, a time constant error rate λ [j] is assumed for the jth combination and the exponential distribution is used as the error occurrence model. Preferably, then a time-dependent estimated value MTTF '[j] (τ2, τ1) for the parameter MTTF [j] according to the calculation rule

calculated. As mentioned above, f_ges [j] (τ1) is the total number of all errors of the j-th component-error-type combination that have occurred in all trials up to the time τ1. By f_ges [j] (τ2; τ1), the total number of all errors of the jth combination that will occur according to the prediction in the period from τ1 to τ2 will be designated. The denominator can never be zero.

For the temporally constant error rate λ [j], an estimate λ '[j] (τ2; τ1) dependent on the two times τ1 and τ2 is calculated, in accordance with the calculation rule

As stated above, the reliability is predicted using the distribution function F (t). As a measure of the reliability with respect to the jth component-error type combination, a failure probability U [j] = U [j] (τ2; dist_a, τ1) is predicted according to the calculation rule U [j] (τ2; dist_a, τ1) = F [λ '[j] (τ2; τ1)] (dist_a) = 1-exp [-λ' [j] (τ2; τ1) * dist_a].

To the Time τ1 is therefore predicted that a Incident probability with respect to an annual Distance from dist_a at time τ2 will be achievable, the U [j] (τ2; dist_a, τ1) is.

Preferably For example, this procedure is performed once for each of the m device-fault type combinations. To the Time τ1 are thereby generated in the forecasts U [j] (τ2; dist_a, τ1) (j = 1, ..., m). These m predictions can be compared. Thereby can at time τ1 be found out which component-fault type combination for Time τ2 will have low predicted reliability. Already at the time τ1 Is it possible, the test plan to change and to test this combination in more detail. It is also possible, contrary to original Planning not to perform individual tests if the required reliability even without these individual experiments will be demonstrated. Thereby saves the procedure attempts and thus time.

In a first alternative embodiment is for an estimate a maximum error probability α between 0 and 1 and therefore between 0% and 100%. Common values for α are 0.1 or 0.05 or 0.01, ie 10%, 5% or 1%. The confidence level 1 - α denotes the statistical security with which reliability is predicted.

berechnet.If no errors of the type of error occur up to the time τ1 and it is predicted that no errors will also occur by the time τ2, an exponential distribution is used as the error occurrence model in the first alternative embodiment. Preferably, furthermore, β = 1 is assumed men. A lower bound MTTF '[j] (τ2; α, τ1) for the MTTF value MTTF [j] which depends on the confidence level 1-α is calculated according to the calculation rule

calculated.

berechnet. Genau wie in der obigen Ausführungsform wird hieraus ein Schätzwert λ'[j](τ2;τ1) für die zeitlich konstante Fehlerrate λ[j] berechnet. Mit Hilfe des Schätzwerts für die Fehlerrate wird ein Schätzwert für die erreichbare Zuverlässigkeit bestimmt. In einer zweiten alternativen Ausführungsform wird anstelle der Exponentialverteilung eine Weibull-Verteilung verwendet. Der Formparameter β wird als bekannt vorausgesetzt. In den folgenden Rechenvorschriften werden die einzelnen Äquivalent-Fahrstrecken der n Einzelversuche verwendet. Mit dist_ä[i,j](τ) wird die Äquivalent-Fahrstrecke für die j-te Kombination bezeichnet, die im i-ten Einzelversuch bis zum Zeitpunkt τ zurückgelegt wurde (i = 1, ..., n; j = 1, ..., m).If errors have occurred up to the time τ1 f_ges [j] (τ1) or if it is predicted that errors will occur in the period from τ1 to τ2 f_ges [j] (τ2; τ1), then a non-α dependent estimate MTTF ' [j] (τ2; τ1) according to the calculation rule

calculated. As in the above embodiment, an estimated value λ '[j] (τ2; τ1) for the time constant error rate λ [j] is calculated therefrom. Using the estimate of the error rate, an estimate of the achievable reliability is determined. In a second alternative embodiment, a Weibull distribution is used instead of the exponential distribution. The shape parameter β is assumed to be known. The following calculation rules use the individual equivalent routes of the n individual tests. With dist_ä [i, j] (τ) is the equivalent distance for the j-th combination, which was covered in the i-th individual experiment until the time τ (i = 1, ..., n; j = 1, ..., m).

berechnet Ein einseitiges Konfidenzintervall für eine Weibull-Verteilung wird in R. A. Abernethy, a.a.O., Kapitel E-2, beschrieben.If no errors occur by the time τ1, in the second alternative embodiment, a lower bound MTTF '[j] (α, τ2; τ1) for the MTTF value MTTF [j] is calculated according to the calculation rule

A one-tailed confidence interval for a Weibull distribution is described in RA Abernethy, loc. cit., chapter E-2.

berechnet. Ein Schätzwert für die Weibull-Verteilung wird ebenfalls in R. A. Abernethy, a.a.O., Kapitel E-2, beschrieben.If f_ges [j] (τ1) errors have occurred by the time τ1 and f_ges [j] (τ2; τ1) errors are predicted in the period from τ1 to τ2, an estimate α MTTF '[j] (τ2 ; τ1) for the MTTF value MTTF [j] according to the calculation rule

calculated. An estimate of the Weibull distribution is also described in RA Abernethy, loc. Cit., Chapter E-2.

In the following calculations is for each of the m component fault type combinations a required reliability demonstrated. This reliability is to be proven in the form of a required maximum error rate. The required maximum error rate in this example is visible for the combination "silencer housing corroded "150,000 ppm per year. Of 1 million muffler housings allowed at an average Travel distance dist_a per year so a maximum of 150,000 cases per Be visibly corroded during the year.

The required as well as the in 2 to 9 Predicted reliability values are purely hypothetical values that serve only to illustrate the embodiment. Actual ppm setpoints and proven ppm actual values are orders of magnitude below these hypothetical values.

On the x-axis of the diagrams of 2 to 9 in each case the second times τ2 are entered. The respective first time τ1 is indicated by an arrow. The trial will begin in February 2005. The plan is to complete the trial in December 2006. On the y-axis of the diagrams of 2 . 4 . 6 and 8th is the total accumulated equivalent distance dist_ges_ä [j] (τ2) in [km] entered in the trial, the trucks in the experiment total until the respective second time τ2 have completed or will cover according to the planning. As stated above, this entire equivalent distance is calculated according to the calculation rule dist_ges_a [j] (τ2) = dist_ges_a [j] (τ1) + dist_plan_ges_a [j] (τ2; τ1) calculated.

On the y-axis of the diagrams of 3 . 5 . 7 and 9 the respective predicted error rate U = U [j] (τ2; dist_a, τ1) is entered.

In 2 and 3 a plan is illustrated before the start of the experiment. The first time τ1 is thus z. In February 2005. The diagram of 2 shows the respective planned equivalent distance dist_ges_ä [j] (τ2) until the respective time τ2. The diagram of 3 shows the time course of the predicted error rate U = U [j] (τ2; dist_a, τ1) for the jth combination at the respective time. As can be seen from these hypothetical values, at the end of the experiment only an error rate of 330,000 ppm per year is predicted instead of the required error rate of 150,000 ppm per year.

Therefore the experimental program will be modified so that the equivalent distance will increase. This is achieved, for example, by planning the actual one Driving distance during the Attempt to increase.

In 4 and 5 the changed planning and a result of this changed planning are illustrated. The first time τ1 is now after the beginning of the experiments, z. For example, in August 2005. Up to this time, no errors were measured in the trial. 4 shows the time course of the accumulated equivalent route. Until the first time in August 2005, this is the actual equivalent distance (product of the actual distance to the first time and the factor of refraction), then the planned equivalent distance. For comparison, in 4 Dashed lines of the originally planned time course of the accumulated equivalent distance shown that already in 2 was shown. 5 shows the time course of the predicted error rate. For this purpose, the measurement result is used that no error of the type of fault has occurred up to the first time. For comparison, in 5 Dashed line of the predicted according to the original planning time course of the error rate, which is already in 3 was shown.

6 and 7 show an alternative continuation of the experiment. Also in this example, the first time after the beginning of the experiments, z. For example, in August 2005. Up to this time, no error type errors were measured in the trial. However, a smaller overall journey was actually covered by the first time than originally planned. This also reduces the equivalent distance. 6 shows the time course of the equivalent route according to the alternative course. Until the first time in August 2005, this is the actual equivalent distance (product of the actual distance to the first time and the factor of refraction), then the planned equivalent distance. For comparison, in 4 Dashed lines of the originally planned temporal course of the equivalent route shown already in 2 was shown. 7 shows the time course of the predicted error rate. For this purpose, the measurement result is used that no error of the type of fault has occurred up to the first time. For comparison, in 7 Dashed line of the predicted according to the original planning time course of the error rate, which is already in 3 was shown.

8th the time course of the equivalent route according to the alternative planning of 6 , In the example of 8th The first time is at the end of November 2005. Until the first time in November 2005 is the equivalent distance of 8th the actual equivalent distance (product of the actual distance up to the first time and the factor of refraction), then the planned equivalent distance. For comparison, was in 8th the equivalent distance of 4 entered dashed.

9 shows the detectable error rate, which according to the experiment 8th was predicted. In the example of 9 two errors were counted before the first time in November 2005. As a result, the predicted error rate deteriorates significantly. For comparison, in 9 the course of the detectable error rate of 5 entered dashed.

For October 2005, before the two errors, an error rate of 500,000 ppm / year was predicted. For the November 2005, thus after occurrence of the two errors, became an increased error rate predicted by about 660,000 ppm / year. The predicted error rate in the following months is well above that of 5 ,

List of used reference signs and symbols

Claims (15)

A method for predicting the reliability of a technical product, wherein as the reliability is predicted an average number of errors related to a load of the product during its use, the product is subjected to stress in at least one trial and the trial extends over a predetermined period of time and a computer evaluable relationship between - the number of errors related to the load during use and - the number of errors related to the load during the test is determined, characterized in that a first and a second time of the period are given, the second Time counts after the first time point is counted as many errors occurred in the trial until the first time when an actual load is a measure of the load that the product is exposed to during the experiment until the first time point is measured as a target Bel A measure is given of the load to which the product will be exposed during the test between the first time and the second time and the reliability is automatically calculated using - the actual load, - the number of errors counted up to the first time, - the Target load, - an expected number of errors between the first and the second time point and - the relationship between the error numbers is predicted. Method according to claim 1, characterized, that the Number of errors due to a degradation due to the product the load during of the experiment occurs, in dependence from the target load an expected degradation of the product between the first and the second time is predicted and the expected number of errors between the first and the second time depending is predicted by the expected degradation. Method according to claim 1, characterized, that when expected number of errors is given, that between the first and the second time no errors occur. Method according to one of claims 1 to 3, characterized in that the reliability is calculated using a parametric error-occurrence model, the model determining the number of errors as a function of a measure of the load during use and at least one parameter of the model automatically using: - the actual load, - the counted number of errors up to the first time, - the target load, - an expected number of errors between the first and second times, and - the relationship between the error numbers is estimated. Method according to one of claims 1 to 4, characterized, that when reliability the number of errors related to a time-dependent first size, which the load on the product during of its use is calculated, as an actual load of the product during of the experiment a second size, which the load on the product during of the experiment is measured and as a target load a value for given the second size becomes. Method according to claim 5, characterized, that - the first Size the duration the load on the product during its use and - the second size the duration the load on the product during of the experiment. Method according to claim 5, characterized, that - either the load on the product during the use as well as the load of the product during the Experiment involves the occurrence of stressful events, - the first Size the number the stressful events during the use is and - the second size the number the stressful events during of the experiment. Method according to one of claims 1 to 7, characterized, that the Product is a vehicle or a component of a vehicle, the first size at covered by the use of the vehicle Track is and the second size of the vehicle during the Trial Track is. Method according to one of claims 1 to 8, characterized, that in In a first preliminary experiment, the product of the stress during the Is exposed to experiment, in the first pre-trial a first Characteristic that the load on the product during of the experiment is measured, in a second Preliminary trial exposed the product to stress during use becomes, in the second pre-trial a second parameter, the the load on the product during of use, is measured, both in the first as well as in the second pre-trial a characteristic, which by the respective Indicates degradation of the component caused degradation, is measured, and the connection between the error numbers under use - of the measured values of the first and the second parameter, - of the measured Value of the characteristic characteristic of the degradation is determined. Method according to one of Claims 1 to 9, characterized in that a setpoint value for the reliability is predefined, a number of errors to be expected after the first time point is predicted and using - the actual load, - the counted number of errors up to first time, - the target load, - the predicted number of errors after the first time and - the context it is predicted at which second time point a value is predicted for the reliability which is greater than or equal to the predetermined setpoint value. Computer program product included in the internal memory a computer can be loaded and includes software sections, with in which a method according to any one of claims 1 to 10 can be carried out, if the product is running on a computer. Computer program product on one of one Computer readable medium is stored and stored by a computer has readable program means that cause the computer to to carry out a method according to one of claims 1 to 10. Digital storage medium with electronically readable Control signals, so with a programmable data processing system can work together the existence Method according to one of the claims 1 to 10 executable is. Data processing system for the prediction of reliability of a technical product, where the data processing system related to the prediction of an average number of errors on a load of the product during its use as the reliability is designed the data processing system read access to a data store, in which a computer evaluable Log of at least one experiment is stored, in which the product has been subjected to a load and spread over a predefined time span extends, and the data processing system for the automatic evaluation of a computer-available relationship between - the number of errors related to the load during use and - the number of errors related to the load during the experiment designed is characterized in that the protocol - the number of errors that occurred in the trial until the first time are and - when an actual load is a measure of the load, the product during the experiment is suspended until the first time, comprises in the data store a computer-accessible description of a target load as a measure of what strain the product during of the experiment between the first time and a subsequent one second time of the period will be suspended is and the data processing system for the prediction of reliability under use - of the Actual load, - of the counted Number of errors until the first time, - the target load, - one expected number of errors between the first and the second Time and - of Relationship between the error numbers is designed. A computer program product for predicting the reliability of a technical product, wherein the computer program product is configured to predict an average number of errors related to a load of the product during its use as the reliability, the computer program product having read access to a data store in which a computer-analyzable log of at least one test is stored in which the product has been subjected to a load and which extends over a predetermined period of time, and the computer program product for automatically evaluating a computer-available relationship between - the number of errors related to the load during use and - the number of errors related to the load is designed during the test, characterized in that the protocol - the number of errors that have occurred in the experiment until the first time - as an actual Bel includes a measure of the load to which the product is subjected during the test until the first in the data memory, a computer-accessible description of a target load as a measure of which load the product will be exposed during the test between the first time and a subsequent second time of the period, and the computer program product for predicting the reliability Use - the actual load, - the counted number of errors up to the first time point, - the target load, - an expected number of errors between the first and the second time point and - the relationship between the error numbers.