DE102007016395B3 - Vehicle-specific quantification function determining method for track, involves determining regression coefficients for vehicle reaction by satisfying preset vehicle-specific quantification equation - Google Patents

Abstract

The method involves utilizing track variations with different form, amplitude and length. Characteristic parameters associated to the variations are determined. Limit values of a temporal lapse of a vehicle reaction for the variations are calculated under variation of vehicle speed and/or a track curvature. Regression coefficients for the vehicle reaction are determined using a regression analysis by satisfying a preset vehicle-specific quantification equation which relates the speed, the curvature and the parameter.

Description

The The invention relates to a method for determining vehicle-specific Evaluation functions for the effect-related assessment of the position quality of a Track, with the deviations of the track from its nominal position measured as disturbances and the corresponding ones calculated vehicle reactions are evaluated.

At the complete system "vehicle-track" will be determined Requirements regarding Safety, stress and ride comfort. To meet these requirements guarantee to be able to are the system reactions (wheel forces and Wagenkastenbeschleunigungen) or the factors of the Systems (disorders) to limit. This is in the production by the constructive Design of vehicle and track components achieved. Because both Vehicles as well as the track are subject to wear, the warranty must the system requirements in addition be ensured by appropriate maintenance measures. Regarding the driveway here is mainly the maintenance of the Track situation to call. By regular inspections of the track have to unacceptable Track position error detected and then repaired.

Around To ensure an economical railway operation, the effort should for the Track maintenance as possible low, without jeopardizing system security and to affect the ride comfort in an unacceptable way. Therefore, for many years, an inspection process is being sought, the best possible these requirements Fulfills.

A method for condition data acquisition of driveways is known, wherein the location assignment of the condition data to the driveway in the mobile condition data acquisition system is effected by first determining the geographical position by means of a satellite-bound position determination system, the assignment between a coordinate of the position determination system and the driveway by specifying at least one fixed reference point and the direction of movement on the guideway is carried out by the inspector and that the distance determination along the guideway by re-calculating the coordinates of the positioning system on the guideway by the distance traveled to the reference point, or is converted from this to a position on the guideway ( DE 101 25 515 A1 ).

In addition, a method for simulating the state of transport routes is known in which, for a prognosis of the state development, input data are acquired in one step and these are sent for processing in a subsequent step, using a neural network to be optimized, which is generated using arbitrary input data a division of the transport path into homogeneous sections that show the same state or a similar behavior with respect to one or more parameters, and for this a classification and / or a prognosis of the state development created ( EP 1 271 364 A2 ).

• • Die Gleislagegeometrie wird lediglich anhand empirisch festgelegter Maßstäbe beurteilt, deren Zusammenhang mit den resultierenden Fahrzeugreaktionen und somit deren Auswirkung auf Sicherheit und Fahrkomfort oft unzureichend belegt ist.
• • Es wird stets nur die Amplitude der Gleislageabweichungen beurteilt. Die Form und die Länge der Gleislagefehler gehen in die Beurteilung nicht ein, obwohl sie die Auswirkung auf das Fahrzeug bei gleicher Amplitude erheblich beeinflussen.
• • Die bisherigen Beurteilungsmaßstäbe berücksichtigen nicht das evtl. gleichzeitige Auftreten verschiedener Gleislagefehler (z. B. Überlagerung von Längshöhen- und Richtungsfehler).
• • Die bisherigen Beurteilungsmaßstäbe berücksichtigen nicht den Einfluss der Gleistrassierung, obwohl sich die gleichen Gleislagefehler im Bogen anders auswirken als im geraden Gleis.

The currently used methods for inspecting the track geometry have serious weaknesses:

• • The track geometry is assessed only on the basis of empirically determined standards, the connection with which is resulting vehicle reactions and thus their impact on safety and ride comfort is often insufficiently documented.
• • Only the amplitude of track deviations is assessed. The shape and length of the track position errors are not included in the assessment, although they significantly affect the effect on the vehicle at the same amplitude.
• • The previous assessment criteria do not take into account the possible simultaneous occurrence of different track position errors (eg superposition of longitudinal height and direction errors).
• • The previous evaluation standards do not take into account the influence of the track assignment, although the same track position errors affect the bow differently than in the straight track.

The Safety and system stress as well as the driving comfort become not by the geometrical course of the track position error itself, but by its effect on the vehicle (forces, accelerations) affected. When assessing the error is therefore above all else to consider their impact. In the past, therefore, have been different ways for the impact assessment of the track situation, the but not yet to the desired Success led.

at the assessment of measured vehicle reactions become the tracks at regular intervals with drive a metering train. This will be the wheel forces and car body accelerations (Vehicle reactions) measured and then assessed.

Limit values are defined for the individual reaction quantities (horizontal wheel force, vertical wheel force, horizontal acceleration, vertical acceleration). In the assessment, the points in the track at which the course of a reaction variable exceeds the specified limit value progresses, determines and outputs as excess.

• • Die Beurteilungsergebnisse beziehen sich ausschließlich auf das Messfahrzeug. Man erhält keine Aussage darüber, welche Ergebnisse andere Fahrzeugtypen liefern würden.
• • Die Messergebnisse werden zudem vom Unterhaltungszustand der Messfahrzeuge, durch die Berührgeometrie zwischen Rad und Schiene, durch den Trassierungsverlauf und durch das Wetter beeinflusst. Die Beurteilung bezieht sich somit nicht eindeutig auf Gleislageabweichungen.

This method includes the following disadvantages:

• • The results of the evaluation refer exclusively to the measuring vehicle. There is no statement as to what results other types of vehicles would deliver.
• • The measurement results are also influenced by the entertainment status of the measuring vehicles, by the contact geometry between wheel and rail, by the course of the route and by the weather. The assessment thus does not clearly refer to track position deviations.

at the assessment of calculated vehicle reactions becomes the context between track deviations and vehicle reactions based on described physical models. On the basis of the model data be for the measured track position parameters via simulation calculations the vehicle reactions and evaluate these as measured reactions (see above).

The Simulation calculation is very much because of the complex vehicle models time-consuming. The method is therefore mainly used for investigations in the laboratory used. For the use in the track position assessment is the simulation calculation for time reasons not usable.

What is known is a method and a device for determining the condition of the position of a track, with a vehicle which is equipped to determine the location caused by the location of the track local accelerations with a meter and the track section, in particular passes at a certain speed, wherein the track section, the change in the acceleration occurring when driving on the track and determined based on a uniform comparison speed, and summed as measured values to a characteristic value and compared with a comparison characteristic value. As an alternative to the measurement, the accelerations can also be calculated from measured characteristic values of the track position, but only one mass point is considered ( DE 19 19 775 B2 ).

at This method makes it difficult to understand where the used Comparative values for the respective sections to be measured come. The process generates Moreover, no statements what the measured values for concrete Affect the track. There is no consideration expected extreme values, but a section by section comparison in a statistical evaluation. The procedure describes no impact assessment of the track situation on indirect Way with regard to the effects on the vehicle reaction.

in the Paperback of Mathematics, 22nd edition, published by Harri Deutsch, Thun and Frankfurt a. M., pages 673 to 694 are generally the basics theoretical Regression characteristics and the correlation between correlation and regression.

Of the Invention has for its object to develop a method which the track track indirectly related effect based (with respect. the actual effects on the vehicle reaction) and the requirements of the trackside inspection regarding safety and ride comfort enough.

• a) K Gleislageabweichungen (Teststörungen) TS_k=1..K mit unterschiedlicher Form, Amplitude und Länge und mit unterschiedlicher Überlagerung in horizontaler und vertikaler Richtung zur Abdeckung des Spektrums realer Gleislageabweichungen benutzt werden, wobei jeweils bei einer Gleislageabweichung TS = (y, z, gh) für jede Schiene getrennt die horizontale Abweichung y und die vertikale Abweichung z von ihrer Solllage sowie die Abweichung der gegenseitigen Höhenlage gh der beiden Schienen von der Sollüberhöhung betrachtet werden,
• b) für jede Teststörung die charakteristischen Parameter p_m = {y_St, y_Ew, z_St, z_Ew, gh_St, gh_Ew) durch die jeweilige Steigung der Größen y, z und gh und durch den Extremwert der Größen y, z und gh beschrieben werden, wodurch die den K Teststörungen zugeordneten charakteristischen Parameter p_{m=1..6, k=1..K} = {y_St,k, y_Ew,k, z_St,k, z_Ew,k, gh_St,k, gh_Ew,k} ermittelt werden,
• c) für alle Teststörungen unter Variation von Fahrgeschwindigkeit v und Gleiskrümmung kr die Extremwerte der zeitlichen Verläufe der J simulierten bzw. gemessenen Fahrzeugreaktionen berechnet werden,
• d) und wobei schließlich mittels Regressionsanalyse für jede Fahrzeugreaktion j = 1..J die Regressionskoeffizienten a_j bis i_j bestimmt werden durch Gleichsetzen der fahrzeugspezifischen Bewertungsfunktionen Rj = aj + bj·ySt + cj·yEw + dj·zSt + ej·zEw + fj·ghSt + gj·ghEw + hj·v + ij·krmit den Extremwerten aus Schritt c) unter Einsetzen der charakteristischen Parameter p_m,k = {y_St,k, y_Ew,k, z_St,k, z_Ew,k, gh_St,k, gh_Ew,k}, der jeweiligen Fahrgeschwindigkeit v und der jeweiligen Gleiskrümmung kr.

This is achieved according to the invention in that
the deviations of the track from its nominal position are measured as disturbance variables and evaluated on the basis of the associated calculated vehicle reactions. The vehicle-specific evaluation functions required for this purpose are determined by simulation calculation on the basis of a vehicle model and / or from the results of driving and / or test bench tests with a vehicle, wherein

• a) K track deviations (test disturbances) TS _{k = 1..K are used} with different shape, amplitude and length and with different superimposition in the horizontal and vertical direction to cover the spectrum of real track position deviations, each with a track deviation TS = (y, z gh) separately for each rail the horizontal deviation y and the vertical deviation z from its nominal position and the deviation of the mutual height gh of the two rails from the nominal cant,
• b) for each test disturbance, the characteristic parameters p _m = {y _St , y _Ew , z _St , z _Ew , gh _St , gh _Ew ) through the respective slope of the quantities y, z and gh and by the extreme value of the quantities y, z and gh, whereby the characteristic parameters p _{m = 1..6, k = 1..K} = {y _{St, k} , y _{Ew, k} , z _{St, k} , z _{Ew, k} , gh assigned to the K test disturbances are described _{St, k} , gh _{Ew, k} } are determined
• c) the extreme values of the time lapses of the J simulated or measured vehicle reactions are calculated for all test disturbances while varying the vehicle speed v and the track curvature kr,
• d) and finally by regression analysis for each vehicle reaction j = 1..J the regression coefficients a _j to i _{j are} determined by equating the vehicle-specific evaluation functions R j = a j + b j · y St + c j · y Ew + d j * z St + e j * z Ew + f j · gh St + g j · gh Ew + h j · V + i j · kr with the extreme values from step c) using the characteristic parameters p _{m, k} = {y _{St, k} , y _{Ew, k} , z _{St, k} , z _{Ew, k} , gh _{St, k} , gh _{Ew, k} }, respective travel speed v and the respective track curvature kr.

The method described here, in which the track position on indirect Is judged on an impact-related basis, meets the requirements of the track position inspection in terms of Safety and ride comfort. It is based on extensive research the physical connections of vehicle and driveway. The procedure allows a very detailed Evaluation of track deviations. This allows unnecessary repair measures avoid.

Surprisingly, the approach has been found that safety and system stress as well as the ride comfort does not depend on the time or path-dependent course of the vehicle _reaction R, but on their extreme value R _Extr , which _sets due to the track _deviation .

Of the Relationship between the track deviation and the corresponding Extreme value of the vehicle reaction is first based on physical Models by simulation calculation or on the basis of measurement results of driving and / or test bench tests examined with real vehicles. Based on the examination results will be afterwards mathematical functions for approximate Description of the relationship derived. With the help of the evaluation functions are the track position deviations measured during the inspection trip assessed. The relation between track position error and its effect on the vehicle is thereby produced by indirect means.

• • Die Gleislage wird nicht nach ihren geometrischen Abweichungen von der Solllage sondern nach ihren Auswirkungen auf das Gesamtsystem "Fahrzeug-Fahrweg", also wirkungsbezogen beurteilt.
• • Bei der Beurteilung werden die Trassierungsverhältnisse, also die Gleiskrümmung und die zulässige Fahrgeschwindigkeit des Fahrzeugs mit der örtlich tatsächlich vorhandenen Größe berücksichtigt.
• • Die Gleislagefehler werden bezüglich ihrer Wirkrichtung nicht isoliert betrachtet. Bei der Beurteilung wird die Überlagerung der horizontalen und vertikalen Gleislageabweichungen berücksichtigt.
• • Die bei einer Reaktionsmessung vorhandenen Nachteile (Einfluss der Berührgeometrie und dgl.) werden eliminiert.
• • Der Zusammenhang zwischen Gleislageabweichung und Fahrzeugreaktion wird auf der Basis von sehr genauen Fahrzeugmodellen, welche auch Nichtlinearitäten berücksichtigen, hergestellt.
• • Der Zusammenhang zwischen den charakteristischen Parametern der Gleislageabweichung und der Fahrzeugreaktion wird nur einmal untersucht. Die Fahrzeugreaktionen müssen also nicht bei jeder Inspektionsmessfahrt aufwändig per Simulation berechnet werden.
• • Der Algorithmus des Beurteilungsverfahrens ist zeitunkritisch und kann im Online-Betrieb auf den vorhandenen Geometriemessfahrzeugen angewandt werden.
• • Bei der Beurteilung kann die Auswirkung von Gleislageabweichungen für verschiedene Fahrzeugtypen berücksichtigt werden.

The solution according to the invention for indirect effect-related assessment of the track position has the following advantages over the currently used methods:

• • The track position is not assessed according to its geometric deviations from the target position but according to their effects on the overall system "vehicle driveway", that is, in terms of effect.
• • In the assessment, the track conditions, ie the track curvature and the permissible travel speed of the vehicle are taken into account with the actually existing size.
• • The track position errors are not considered in isolation with respect to their effective direction. The assessment takes into account the superposition of horizontal and vertical track deviations.
• • The disadvantages of a reaction measurement (influence of the contact geometry and the like) are eliminated.
• • The relationship between track deviation and vehicle response is established on the basis of very accurate vehicle models, which also take into account nonlinearities.
• • The relationship between the characteristic parameters of the track deviation and the vehicle reaction is examined only once. The vehicle reactions therefore do not have to be complexly calculated by simulation at every inspection test drive.
• • The algorithm of the assessment procedure is non-time critical and can be used in online mode on existing geometry measurement vehicles.
• • Assessment may take into account the effect of track deviation for different vehicle types.

Since the vehicle reaction depends on the shape and size of the track deviation (disturbance S), the driving speed of the vehicle v and the routing of the track (track curvature kr), the relationship between the input variables and the resulting extreme value of the vehicle response R can be simplified by the linear function equation R = f (S, v, kr) = a + b * S + c * v + d * kr describe. For practical application, however, this general form of the evaluation function is usually insufficient. Investigations have shown that the disturbance variable is additionally weighted with the driving speed v and that in the evaluation another component v ² · kr which is decisive for the quasi-static vehicle reaction in the curve should be taken into account: R = f (S, v, kr) = a + b * S * v + c * v + d * kr + e * v 2 · Kr.

In addition, the descriptions of the disturbance S and the vehicle response R are to be specified as follows:
The track deviation acting on the vehicle is composed of several components such as longitudinal height z and direction y of both rails and the deviation of the mutual altitude gh from the nominal cant, which are measured separately. Each measurement signal of these disturbances S _{n = 1..N} must first be broken down into individual _errors , which can then be assessed together according to their overlapping _ranges . The subdivision into individual errors can, for example, be based on the zero crossings of the mean-value-free measuring signals or also by means of the intersections of the measuring signals with one to the path axis (abscissa) parallel threshold line, cf. 1a and 1b , But there are also other subdivision methods conceivable, see, for. B. 1c ,

For each of the separated individual _errors , characteristic geometric parameters p _{m = 1..M} can be determined. As a characteristic parameter, for example, the maximum or average slope St _max or St, the extreme value EW, the height h, the area integral A u. Ä., see 2 , The appraisal function thus takes the following form: R = f (p m = 1..M , v, kr) = a + (b 1 · p 1 + ... + b M · p M ) · V + c · v + d · kr + e · v 2 · Kr.

The selection of the vehicle reactions to be evaluated depends on the purpose of the track position assessment. In principle, all available evaluation variables can be used if they are sufficiently sensitive to track position faults. However, since the trackside assessment is mostly focused on driving safety, driving load and driving comfort, the appropriate reaction variables such as wheel-to-rail forces and vehicle accelerations should usually be used for the assessment. This also has the advantage that limits have already been set for these sizes, z. B. in the European standard EN 14363 for vehicle approval, which are used for the track position assessment directly or in a modified form as basic sizes R _base . Thus, the determined extreme values of the vehicle reactions R can also be stated as a percentage utilization of the wear stock:

Corresponding to the number of selected reaction _variables , therefore, there are several evaluation functions per vehicle R _{j = 1..J} , = f (p ₁ , p ₂ ,..., P _M , v, kr). Since for the comprehensive assessment of the track situation usually several types of vehicle (eg different passenger cars, locomotives and freight cars) must be taken into account, a relatively large number of assessment functions is necessary overall. Due to their simple structure, however, they can be evaluated very quickly so that an online assessment of the measurement results is possible even during the test drive.

Before the assessment procedure can be used, the coefficients of the assessment functions must be determined. This is preferably done using known simulation methods. For this purpose, first fictitious track position faults he testifies, which are composed of different combined with each other components with different characteristic parameters p _{m = 1..M} . These disturbances are to be chosen so that they reflect as well as possible in their shape and size the track position errors occurring in reality. Furthermore, the track curvature kr and the vehicle speed v are given in several stages, which correspond to the real operating conditions of the selected vehicles. With these input variables, the vehicle reactions to be assessed are then determined for each vehicle model to be considered by means of simulation calculation. From the calculated time _{courses of the vehicle reactions,} their extreme values R _{j = 1..J} are then determined for each calculation case.

In this way one obtains for the desired coefficients a _j , b _1j ,..., E _{j of} the assessment functions of each vehicle an overdetermined system of equations which can be solved by known methods (regression analysis).

embodiment

following the invention will be explained with reference to an embodiment.

there demonstrate:

1a to 1c - Different variants of the subdivision of the measured Störgrößenverläufe in single error

2 - various characteristic parameters of the disturbance variables (extreme value EW, average gradient St, maximum gradient St _max , error surface A, peak-to-peak value h)

The disturbance variables S take into account the measurement signals of the longitudinal height z, the direction y and the mutual altitude gh. These signals are decomposed into individual errors y _i , z _i and gh _i , which are delimited from one another by their zero crossings ( 1a ). Then the associated extreme value EW and the average slope St are determined as characteristic parameters for each individual error ( 2 ). The following vehicle reactions R _j are used to assess the track position:
R ₁ = ΣY = sum of horizontal wheel forces (system load),
R ₂ = Y / Q = ratio of horizontal to vertical wheel force (safety),
R ₃ = Qmin = minimum wheel contact force (safety),
R ₄ = Qmax = maximum wheel contact force (system load),
R ₅ = yb = horizontal car body acceleration (ride comfort),
R ₆ = zb = vertical car body acceleration (ride comfort).

With the driving speed v and the track curvature kr, the following evaluation functions result: R j = a j + (b j · y St + c j · y Ew + d j * z St + e j * z Ew + f j · gh St + g j · gh Ew ) · V + h j · v 2 · Kr + i j · V + j j · kr

The vehicle-specific regression coefficients a _j to j _j are now as described above z. B. determined on the basis of simulation calculations, which are based on the considered track (-teil-) network typical track position error and the main vehicles used there. The result is a set of evaluation functions, which for each reference vehicle, depending on its driving speed, provide the extreme values of the vehicle reactions to be expected at each point of the track. By referring the determined vehicle reactions to their intervention threshold values, the utilization of the wear stock in percent can now be specified for each point. This in turn is the basis for the planning of the maintenance measures on the examined route.

Claims (8)

A method for determining vehicle-specific evaluation functions for the effect-related assessment of the positional quality of a track, wherein the deviations of the track from its desired position are measured as disturbances and evaluated on the basis of the associated calculated vehicle responses, characterized in that the vehicle-specific evaluation functions by means of simulation calculation based on a vehicle model and / or e) K track deviations (test disturbances) TS _{k = 1..K} with different shape, amplitude and length and with different superimposition in the horizontal and vertical direction to cover the spectrum real track position deviations are used, each with a track deviation TS = (y, z, gh) separated for each rail, the horizontal deviation y and the vertical deviation z of their desired position and the deviation of the mutual H f) for each test disturbance, the characteristic parameters p _m = {y _St , y _Ew , z _St , z _Ew , gh _St , gh _Ew } by the respective slope of the quantities y, z and gh and by the extreme value of the quantities y, z and gh, whereby the characteristic parameters p _{m = 1..6, k = 1..K} = {y _{St, k} , y _{Ew, k} , z _{St, k} , z _{Ew, k} , gh _{St, k} , gh _{Ew, k} } are determined, g) the extreme values of the time lapses of the J simulated or measured vehicle reactions are calculated for all test disturbances with variation of vehicle speed v and track curvature kr , h) and finally using regression analysis for each vehicle response to be assessed j = 1..J the regression coefficients a _j to i _{j are} determined by equating the vehicle-specific evaluation functions R j = a j + b j · y St + c j · y Ew + d j * z St + e j * z Ew + f j · gh St + g j · gh Ew + h j · V + i j · kr with the extreme values from step c) using the characteristic parameters p _{m, k} = {y _{St, k} , y _{Ew, k} , z _{St, k} , z _{Ew, k} , gh _{St, k} , gh _{Ew, k} }, respective travel speed v and the respective track curvature kr. Method for determining vehicle-specific evaluation functions for effect-related assessment of the positional quality of a track, which - apart from the definition of the evaluation functions - the same features as the patent claim 1, wherein in the vehicle-specific evaluation functions of claim 1, the quantities y _St , y _Ew , z _St , z _Ew , gh _St , gh _{Ew are} additionally weighted with the driving speed v and in the evaluation another, for the quasi-static vehicle reaction decisive component v ² · kr is taken into account, whereby the evaluation functions take the following form: R j = a j + (b j · y St + c j · y Ew + d j * z St + e j * z Ew + f j · gh St + g j · gh Ew ) · V + h j · v 2 · Kr + i j · V + j j · kr Method according to claim 1 or 2, characterized that the location quality track based on evaluation functions for different vehicles is evaluated, with the regression coefficients of the evaluation functions for the different vehicles from each other. Use of the method according to one of the preceding claims for the effect-related assessment of the positional quality of a track using the vehicle-specific evaluation functions, the deviations of the track from its desired position being measured as a disturbance variable course S = (y, z, gh) and evaluated on the basis of the associated calculated vehicle reactions, characterized in that a) the characteristic parameters p _{m = 1..6} = {y _St , y _Ew , z _St , z _Ew , gh _St , gh _Ew } from the respective slope of the quantities y, z, gh and the respective B) subsequently these characteristic parameters p _{m = 1..6} = {y _St , y _Ew , z _St , z _Ew , gh _St , gh _Ew }, the locally permissible driving speed v and the localized track curvature kr are used in the vehicle-specific evaluation functions R _j , whereby the expected extreme values of the driving c) and that these are compared with permissible values of the vehicle reactions, the result of this comparison representing the performance-related assessment of the positional quality of a track. Use of the method according to claim 4, characterized in that the disturbance variable course S = (y, z, gh) divided into sections S _n = (y _n , z _n , gh _n ) and for each section S _n a set of characteristic parameters p _{m , n is} determined. Use according to claim 5, characterized in that the subdivision of the Störgrößenverlaufs in sections S _n = (y _n , z _n , gh _n ) is based on the zero crossings of a modified Störgrößenverlaufs, wherein the modified Störgrößenverlauf results from the Störgrößenverlauf minus its average. Use according to claim 5, characterized in that the subdivision of the Störgrößenverlaufs into sections S _n = (y _n , z _n , gh _n ) based on the intersections of the Störgrößenverlaufs with a to the path axis (abscissa) parallel threshold line. Method according to one of claims 1 to 3 or use according to one of claims 4 to 7, characterized in that the evaluation functions R _j are taken into account for the following vehicle reactions: j = 1: ΣY (sum of the horizontal wheel forces), j = 2: Y / Q (ratio of horizontal to vertical wheel force), j = 3: Q _min (minimum wheel tread force), j = 4: Q _max (maximum wheel tread force), j = 5: yb (horizontal car body acceleration), j = 6: zb (vertical car body acceleration ).