Method and test bench for determining inertia

Abstract

In order to enable a more precise operation of test benches compared to the known prior art, a test-bench basic inertia is ascertained by means of a first inertia identification test, a test-bench operating inertia is ascertained by means of a second inertia identification test, and a vehicle inertia of a vehicle (1) is ascertained from the test-bench basic inertia and the test-bench operating inertia, preferably by subtracting the test-bench basic inertia from the test-bench operating inertia.

Description

[0001] Method and test bench for determining inertia

[0002] The invention relates to a method for operating a test bench, preferably a roller test bench or powertrain test bench or flat track test bench, for a vehicle to be tested, comprising at least one load machine formed from a stator and a rotor for driving or decelerating at least one wheel of the vehicle to be tested, and a test bench control system for controlling the load machines provided on the test bench, and a test bench, preferably a roller test bench or powertrain test bench or flat track test bench.

[0003] Test rigs, such as roller test rigs, powertrain test rigs, or flat track test rigs, are well known in the prior art, e.g., from EP 0 655 617 A2, EP 3 137 872 B1, or CN 212585886 II, and are used to carry out test runs with single-track or multi-track vehicles ("test specimen"). In the case of a roller test rig, one or two rotatable rollers are generally provided for each wheel, or at least for each driven wheel, of the vehicle, on which the wheel is mounted. Often, a common roller is provided for the wheels of an axle, i.e., the front or rear axle. A roller is connected to a (usually electric) load machine, which can generate torque to drive or load the vehicle's powertrain.The torque generated by a load machine is transmitted from the load machine to the roller, from the roller to the wheel, and finally from the wheel to the vehicle's drivetrain. The available load machines on the roller dynamometer can be controlled and / or regulated by a dynamometer control unit to perform desired test runs, such as legally mandated tests. The vehicle is often fixed in place relative to the roller dynamometer during this process.

[0004] When conducting test runs on a chassis dynamometer, i.e., when simulating real driving and load profiles on the dynamometer, the aim is to generate time-dependent angular velocity profiles of the vehicle's wheels in response to acceleration or braking maneuvers (e.g., by applying the brake and / or accelerator pedal). These profiles should reflect the angular velocity profiles that would occur during real-world driving on the road in response to the same acceleration or braking maneuvers. To achieve this, a variety of parameters of the dynamometer and / or the vehicle under test are measured during operation, e.g.,Rotational speeds or velocities of a load machine and / or a torque generated by a load machine and / or a tractive force generated by a load machine and / or an angular velocity of a vehicle drive shaft and / or an angular velocity of a test bench roller and / or a torque generated by a vehicle drive unit and / or a pedal position of an accelerator pedal and / or a brake pedal position of the vehicle brake pedal, etc. From measured values ​​of this multitude of state variables, preferably a tractive force and a velocity, loads are determined using suitable models that describe a real load on the vehicle on the road. Loads determined in this way are typically forces that counteract acceleration or deceleration of the vehicle.Models used to determine load forces include, in particular, models describing the (total) driving resistance experienced by a vehicle under test in real-world operation, or models describing air resistance, or models describing rolling resistance, or models describing gradient resistance, or models describing acceleration resistance.

[0005] In order to generate wheel angular velocities that are as identical as possible to those occurring in real-world driving conditions, in response to forces and / or torques generated by the vehicle and in response to load forces determined according to the above explanations, it is important to note that the mass inertias (hereinafter referred to as "inertia") of the rollers on the roller test bench are typically different from the effective rotational inertias of a vehicle under test. As is well known from the prior art, a vehicle wheel reacts to a sum of torques acting on it in the same way on the test bench as it does on the road if and only if the inertia opposite to the wheel, in the form of the roller, is equal to the total effective rotational and translational inertias of the vehicle under test.If the inertia of the roller is smaller or larger, the roller must be additionally accelerated or decelerated by the load machine, depending on the sign convention used on the test bench. This procedure is well known in the prior art as mass simulation; see, for example, DE 19836711 A1 or AT 510 041 A4.

[0006] To implement the aforementioned additional accelerations or decelerations of the rollers for the purpose of mass simulation, it is also necessary to determine target values ​​that can be achieved by the load machines provided on the roller test bench. These target values ​​for the load machines can include, among others, target values ​​for a load machine air gap torque, target values ​​for a tensile force measurable by a load cell under the load machine, target values ​​for a load machine angular velocity, or target values ​​for a roller angular velocity, etc. As explained above, for the operation of a roller test bench, it is essential to know the masses and consequently the mass inertias (or simply "inertias") of the rollers, load machines, and vehicles under test precisely, particularly to be able to perform an accurate mass simulation.In this context, it is important to note that the inertias of rollers and load machines are usually known with high accuracy, while the inertias of vehicles under test are either unknown a priori or only available as rough estimates, due to factors such as changing test specimens or modifications made to the test specimen during testing. According to the SAE J2264 standard, for internal combustion engine vehicles, the rotating masses of the powertrain can be assumed to be approximately 1.5% of the vehicle mass per axle. However, this rule is not applicable to electrically powered vehicles, as their rotating masses are in the range of 3-5% of the vehicle mass per axle – due to different powertrain topologies, primarily the electric motors used in the powertrain, which have rotors with significantly higher inertias than those of internal combustion engines.For example, the SAE-J2264 standard requires that for vehicles with an effective rotational inertia of more than 1.5% per axle, such as two-wheeled trucks or hybrid / battery electric vehicles, an appropriate estimate or determination of the actual effective mass of the rotating components of the powertrain is carried out.

[0007] In current technology, calculated values ​​from CAD data are typically used to estimate or determine the actual effective mass of a vehicle. However, this approach is particularly problematic for third parties, such as certification authorities, who do not have access to such CAD data. If incorrect inertia values ​​are used as the basis for tests on the test bench, this can distort the mass simulation and consequently the test results, leading to incorrect conclusions. For example, this can result in incorrect conclusions about the achievable range of electric vehicles.

[0008] The problem described above is not limited to roller test benches. It is also well known that the operation of powertrain test benches, flat track test benches, and other test bench concepts often requires precise knowledge of the inertias acting on the test bench. For example, it is common practice to reduce the effective inertia of heavy asynchronous machines used as load machines to a predetermined value using control technology when operating powertrain test benches. However, specifying meaningful values ​​for a control-engineered simulated inertia is only possible if the effective inertia of the vehicle or powertrain under test is precisely known.

[0009] Therefore, one of the objectives of the present invention is to enable more precise operation of test benches, in particular roller test benches for testing electric vehicles.

[0010] This problem is solved by the features of the independent claims. Specifically, for the aforementioned method and for an aforementioned test bench, it is provided that a test bench base inertia is determined by means of a first inertia identification test, a test bench operating inertia is determined by means of a second inertia identification test, and a vehicle inertia of the vehicle is determined from the test bench base inertia and the test bench operating inertia, preferably by subtracting the test bench base inertia from the test bench operating inertia.

[0011] The inventive method makes it possible to determine exact values ​​for the effective inertia of a vehicle under test, thus solving all of the aforementioned problems. If a vehicle inertia determined according to the invention is used for operating the test bench after identification, a more accurate and precise mass simulation can be implemented. Simulation models calculated during testing, e.g., within the framework of hardware-in-the-loop operation, can be more accurately parameterized and consequently simulated more precisely. Furthermore, controllers used on the test bench can be parameterized more accurately. Even if the sole aim is to determine a vehicle inertia, e.g., for certification purposes, this can be done with particularly high accuracy and precision according to the invention.

[0012] Preferably, no vehicle is positioned on the test bench during the first inertia identification test, so that only the inertia of the test bench itself can be identified. During the second inertia identification test, however, the vehicle is positioned on the test bench, but preferably does not generate any torques or forces, e.g., through drive components located in the vehicle's drivetrain, in order not to distort or influence the inertia identification.

[0013] Further advantageous embodiments of the invention and a method mentioned at the outset for solving the technical problem at hand are specified in claims 5-15.

[0014] The present invention is explained in more detail below with reference to Figures 1 to 2a-d, which show exemplary, schematic, and non-limiting advantageous embodiments of the invention. Figure 1 schematically shows a roller test stand for testing two-track motor vehicles.

[0015] Figs.2a-d Signal waveforms during the identification of inertias according to the invention on a roller test bench.

[0016] Fig. 1 schematically shows a roller test stand 20 on which a vehicle 1 is arranged. For the sake of clarity, only the front axle of the vehicle 1 is shown schematically. The roller test stand 20 is regulated and / or controlled by a higher-level test stand control unit 21 and comprises at least one test stand roller 10, which is driven by a load machine 30. As explained earlier, only one test stand roller 10 can be provided for one axle of a vehicle 1, but separate test stand rollers 10 can also be provided for each wheel 9 of a vehicle 1, as indicated in Fig. 1.

[0017] For tests on a roller test bench 20, as carried out during vehicle development, a so-called road load simulation 3 is typically provided. This can be integrated into the higher-level test bench control 21 of the roller test bench 20, as shown in Fig. 1, or it can be provided separately from the test bench control 21. Using the road load simulation 3, a load is calculated, such as a load force FL or a load torque ML, as shown in Fig. 1, which is applied to the vehicle 1 by the roller test bench 20, specifically the test bench roller 10.To determine the load in the road load simulation 3, for example, a speed v and an acceleration a for a vehicle 1 to be tested can be specified by the higher-level test bench control 21 and passed to the road load simulation 3, or measured values ​​of state variables of the roller test bench 20 can also be passed directly to the road load simulation 3, such as an angular velocity of the test bench rollers 10 co. 10 , or an angular velocity of a wheel 9 co9, or a measured test bench load in the form of a measured force Fist or a transmitted torque Mj S These sizes are of course only examples and can be supplemented or replaced by other sizes.

[0018] Vehicle 1 is positioned with at least one wheel 9 on the test rig roller 10 such that forces and / or torques can be applied both from vehicle 1 to the test rig roller 10 and vice versa. A torque Mj currently transmitted from the roller test rig 20, specifically from the test rig roller 10, is... SThe force t is generated by the load machine 30 and determined via a measuring device 51. This device can be in the form of a load cell (e.g., arranged under the load machine 30) or in the form of a torque measuring flange (as shown, arranged in the shaft connection between the load machine 30 and the test rig roller 10), the use and properties of which are well known. Typically, synchronous or asynchronous machines, and sometimes also DC machines, are used as the load machine 30. In a known manner, these machine types have a stator and a rotor, with the inertia of the rotor being decisive for the rotational inertia introduced by the load machine 30. The inertia of the rotor on a roller test rig typically contributes 5–10% to the total inertia. Furthermore, a measuring device 52 can also be provided for determining the position or angular velocity of a test rig roller 10, e.g.,B. in the form of a rotary encoder, as is well known from the prior art. It is also known to detect the rotation angle of a wheel 9. If the rotation angle of the wheel 9 is also detected, any slippage between wheel 9 and test bench roller 10 can be correctly taken into account.

[0019] Wheel 9 is a conventional vehicle wheel, i.e., a combination of a rim with a certain stiffness and an elastically deformable tire mounted on the rim. The circumferential surface of the tire forms the contact surface with the road surface when the vehicle 1 is on a roller test stand 20 in a known manner relative to the test stand roller 10. However, a wheel 9 is not limited to this type of design and can also be designed in other ways, for example, without an elastic tire.

[0020] As also indicated in Fig. 1, the drive train of vehicle 1 includes a drive unit 11 that generates a drive torque Mi and drives the shafts of the drive train and ultimately the wheels 9. The drive unit 11 can be an internal combustion engine, but in particular an electric motor, such as a synchronous machine or an asynchronous machine.

[0021] The aforementioned exemplary parameters can be specified for vehicle 1 and roller test bench 20, for example, by means of a predefined driving cycle that includes corresponding phases with different speeds v and accelerations a, also in combination with a simulated test environment. For example, a negative acceleration a and / or the resistance that a test bench roller 10 opposes to a wheel 9 of vehicle 1 can be increased to simulate uphill driving. In this context, the road load simulation 3 generates, using the parameters specified by the higher-level test bench control 21 of the roller test bench 20, a road load FL or ML to be generated by the roller test bench 20 from the load machine 30 and applied by the test bench roller 10, where the current test bench load Fist or Mj applied by the roller test bench 20 St, as can be seen in Fig. 1, is taken into account in a classic control loop. In the aforementioned control loop, the roller test stand 20 and the vehicle 1 are controlled via a force and / or torque controller 2, and the load machine 30 acts as an actuator.

[0022] One way to implement a road load simulation 3 is, for example, by using one of the formulas corresponding model, in which Fz represents a so-called tractive force, a a speed-independent component of the tractive force, b a coefficient for a linearly speed-dependent component of the tractive force, c a coefficient for a non-linearly speed-dependent component of the tractive force, v a vehicle speed, n a variable exponent, ITIF a vehicle reference weight, Av / At an acceleration, g the acceleration due to gravity and mF*g*sina a component of the tractive force for overcoming the road gradient.

[0023] Using a roller radius rg, an equivalent tensile torque Mz = Fz * rg can be determined from such a tensile force Fz. This torque can be used, for example, in the case of torque measurement using a torque measuring flange 51, as shown in Fig. 1, to regulate a load. If, on the other hand, a force is measured directly using a load cell, the tensile force Fz can also be used directly to regulate a load. A load cell supports the reaction torque of the electric motor via a lever arm in a known manner; in addition, the roller radius provides another parameter from which the tensile force at the roller crown can be calculated.

[0024] In order to achieve the aforementioned goal on the test bench, starting from a tractive force Fz and / or an equivalent tractive torque Mz determined in this way, of generating angular velocities of the wheels 9 that correspond to real angular velocities which would also result in real operation under the influence of the same torques, these real angular velocities can first be determined. If, for this purpose, a torque Mi generated by vehicle 1 and a total rotational inertia Ji of the drivetrain of vehicle 1 are assumed, then the following results for this angular velocity:

[0025] However, if the common situation on roller test benches arises where the inertia of the test bench roller 10 and the total inertia of the drive train Ji do not match, it is necessary to apply a different torque and / or force using the load machines 30, which, despite the deviation in the aforementioned inertias, results in speeds similar to those in real-world driving on the road. A simple calculation shows that for this purpose it is sufficient to scale the torques acting on the test bench 20, i.e., the torques ultimately regulated by the load machines 20, specifically with the ratio of the aforementioned inertias:

[0026] When a load moment ML determined in this way is adjusted, the difference between the actual installed, equivalent roller mass and the desired vehicle mass, and consequently the difference between the actual installed, equivalent roller inertia and the desired vehicle inertia, is applied in the course of the aforementioned mass simulation.

[0027] In light of these considerations, it is immediately apparent that precise knowledge of the masses and inertias of the load machine 30, the test bench roller 10, and the vehicle 1 is essential for enabling accurate and meaningful test bench operation. The invention addresses this need and allows for the precise determination of the aforementioned masses and inertias entirely without the need for, for example, CAD data that would describe the mechanical and structural design of the vehicle 1 under test and enable a computational determination of the inertias.

[0028] According to the invention, a test bench base inertia Jbas is determined by means of a first inertia identification test, and a test bench operating inertia J is determined by means of a second inertia identification test. op determined, and from the test bench base inertia Jbas and the test bench operating inertia J op A vehicle inertia Ji of vehicle 1 is determined, preferably by subtracting the test bench base inertia Jbas from the test bench operating inertia J. op .

[0029] During the first inertia identification test, vehicle 1 is not positioned on the roller test stand 20, but during the second inertia identification test it is, preferably in neutral (i.e., in neutral for a vehicle 1 with automatic transmission), so that the inertias determined by the two identification tests differ only by the inertia of vehicle 1. As mentioned above, the inertia Ji of vehicle 1 can then be determined by a simple subtraction of the form Ji = Jo. P -Jbas will be determined.

[0030] Preferably, the load machines 30 used for the identification tests are warmed up before the start of an identification test, so that no cold start occurs which could lead to a change in operating behavior during operation due to heating phenomena. Furthermore, it is advantageous not to perform a road load simulation during an identification test, which could distort the identification results.

[0031] For the concrete implementation of the identification tests according to the invention, it can be particularly preferably provided that at least one test bench roller 10 is set to a predetermined first test bench roller angular velocity co 10 1 to accelerate at least one test bench roller 10 in the course of an acceleration process starting from the first test bench roller angular velocity co 10 1 with a first acceleration torque M, preferably constant a,i to a predetermined second test bench roller angular velocity co 10 2 to accelerate, from a time period At between a start of the acceleration process, at which at least one test bench roller 10 reaches the first test bench roller angular velocity co 10 1 leaves, and at the end of the acceleration process, at which at least one test bench roller 10 reaches the second test bench roller angular velocity (D 10;2 achieved, as well as from the first acceleration torque M a ,i , the first test bench roller angular velocity co 10 1 and the second test bench roller angular velocity (D 10;2 to determine the test bench basic inertia Jbas. A specific calculation method for determining the test bench basic inertia Jbas can be used here. Possible signal waveforms, such as those that can occur during such an identification attempt, are shown in Figures 2a-2d. Figures 2a-2d specifically show possible waveforms of angular velocities co (solid lines, "-") and corresponding torques M (dashed lines,

[0032] Conversely, during the first inertial identification test, it can also be provided that at least one test bench roller 10 is set to a predetermined third test bench roller angular velocity co 10 3 to accelerate at least one test bench roller 10 during a braking process starting from the third test bench roller angular velocity co 10 3 with a, preferably constant, first braking torque Mb,i to a predetermined fourth test bench roller angular velocity co 104to brake, from a time period At between a start of the braking process, at which at least one test bench roller 10 reaches the third test bench roller angular velocity co 10 3 exits, and an end to the braking process, at which at least one test bench roller 10 reaches the fourth test bench roller angular velocity co 104 achieved, as well as the first braking torque Mb, 1, the third test bench roller angular velocity co 10 3 and the fourth test bench roller angular velocity co 104 to determine the test bench base inertia Jbas.

[0033] In this way, the test bench base inertia Jbas is determined during braking. In a particularly preferred manner, the identification methods described above (braking, acceleration) can also be combined. For example, for identification purposes, acceleration and braking can be performed, and an average value can be calculated from the two inertias thus determined. This approach has the particular advantage that frictions, which, for example, inhibit acceleration but assist braking, can be compensated for. For this purpose, it is particularly advantageous to equate the first and fourth test bench roller angular velocities, and to equate the second and third test bench roller angular velocities.

[0034] In many cases, it also proves advantageous to perform the initial inertia identification tests described above multiple times and to calculate an average value over the several inertias thus determined. Preferably, two, three, five, or ten acceleration and deceleration inertia identification tests can be carried out. To combine values ​​determined by these identifications, an arithmetic mean or a root mean square can be used, and various filtering methods can also be employed.It is also advantageous to vary the test bench roller angular velocities used in multiple initial inertia identification tests, so that, for example, a first initial inertia identification test uses a specific test bench roller angular velocity, while a subsequent inertia identification test, also based on acceleration, uses a different initial test bench roller angular velocity. This takes into account any operating-point-dependent changes in inertia that may occur.

[0035] The procedure for conducting a second inertial identification test can be analogous. The preceding explanations regarding the execution of multiple inertial identification tests, the combination of measured values, etc., apply equally to the second inertial identification test described below.

[0036] Specifically, a second inertial identification test may involve setting at least one test bench roller 10 to a predetermined fifth test bench roller angular velocity co 10 5 to accelerate at least one test bench roller 10 in the course of an acceleration process starting from the fifth test bench roller angular velocity co 10 5 with a second, preferably constant, acceleration torque M a , 2 to a predetermined sixth test bench roller angular velocity co lo ,6 to accelerate, from a time period At between a start of the acceleration process, at which at least one test bench roller 10 the fifth test bench roller angular velocity co 10 5 leaves, and an end of the acceleration process, at which at least one test bench roller 10 reaches the sixth test bench roller angular velocity co lo ,6 achieved, the second acceleration torque Ma ,2 of the fifth test bench roller angular velocity co 10 5 and the sixth test bench roller angular velocity co lo ,6 the test bench operating inertia J op to determine. For the specific calculation of the test bench operating inertia J op can advantageously refer to the formula similar to the above calculation rule. resorted to.

[0037] As explained above, the second inertial identification test can also be designed as a braking test, i.e., it can be provided that at least one test bench roller 10 is brought to a predetermined seventh test bench roller angular velocity co 10 7 to accelerate at least one test bench roller 10 during a braking process starting from the seventh test bench roller angular velocity co 10 7 with a, preferably constant, second braking torque Mb, 2 to a predetermined eighth test bench roller angular velocity colo ,8 to brake, from a time period At between a start of the braking process, at which at least one test bench roller 10 reaches the seventh test bench roller angular velocity (D 10;7 exits, and an end to the braking process, at which at least one test bench roller 10 reaches the eighth test bench roller angular velocity co lo ,8 achieved, the second braking torque Mb, 2 as well as the seventh test bench roller angular velocity (D 10;7 and the eighth test bench roller angular velocity co 10 8 the test bench operating inertia J op to determine. Here too, a combination of values ​​determined on the one hand by means of a braking test and on the other hand by means of an acceleration test is advantageous, especially to compensate for friction.

[0038] As explained in detail above, the test rig basic inertia Jbas determined according to the invention describes a total rotational inertia of the test rig roller 20 provided in the roller test rig 20 and the rotors of the load machines 30 connected to the test rig rollers 10, and the test rig operating inertia J op describes a total rotational inertia of the test rig roller 20 provided in the roller test rig 20 and the rotors of the load machines 30 connected to the test rig rollers 10 as well as of the vehicle 1 to be tested, so that in particular the vehicle inertia can be deduced by means of the aforementioned difference calculation.

[0039] Within the scope of the invention, it was further recognized that while it is useful and often even necessary to ensure meaningful test bench operation, identifying the inertia of a vehicle 1 is required. Consequently, when a new and previously unknown vehicle 1 is placed on the roller test bench 20, identification according to the invention is necessary. However, it turns out that the test bench base inertias Jbas determined by means of the first inertia identification attempt often change only very slightly, for example, due to aging and wear in roller bearings or the deposition of abrasion on the roller 9. The actual influence on a test bench base inertia Jbas is, however, only minor, so that it is often sufficient to identify a test bench base inertia Jbas once, e.g., during the commissioning of a roller test bench, and to subsequently use this value as the basis for the identification according to the invention.Specifically, it may be planned to carry out the first inertial identification test during the commissioning of the roller test stand 20 or during maintenance of the roller test stand 20, and to carry out the second inertial identification test during the operation of the roller test stand 20 to test a vehicle.

[0040] The roller inertia of the test bench alone can be determined immediately before or again before determining the mass of a vehicle. This can further improve the accuracy of a vehicle's mass determination. This approach can be advantageous not only due to changes in the rotating masses of the load machine, but also due to changes in the condition of the tractive force measurement chain (e.g., zero-point offset due to load (remanence) or temperature).

Claims

Patent claims 1. Method for operating a test bench, preferably a roller test bench (20) or powertrain test bench or flat track test bench, for a vehicle (1) to be tested, comprising at least one load machine (30) formed from a stator and a rotor for driving or decelerating at least one wheel (9) of the vehicle (1) to be tested, and a test bench control unit (21) for controlling the load machines (30) provided on the test bench, characterized in that a test bench basic inertia is determined by means of a first inertia identification test, a test bench operating inertia is determined by means of a second inertia identification test, and a vehicle inertia of the vehicle (1) is determined from the test bench basic inertia and the test bench operating inertia, preferably by subtracting the test bench basic inertia from the test bench operating inertia.wherein no vehicle (1) is placed on the test bench during the first inertial identification test and a vehicle (1) to be tested is placed on the test bench during the second inertial identification test.

2. Method according to claim 1, characterized in that the vehicle (1) to be tested does not generate any torque during the second inertial identification test.

3. Method according to one of the preceding claims, characterized in that a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, wherein in the course of the first inertial identification test - the at least one test bench roller (10) on a predetermined first test bench roller angular velocity (co 10 1 ) is accelerated, - the at least one test bench roller (10) during an acceleration process starting from the first test bench roller angular velocity (co 10 1) with a first acceleration torque (M), preferably constant a ,i) to a predetermined second test bench roller angular velocity (co 10 2 ) is accelerated, - from a time period (At) between the start of the acceleration process, at which at least one test bench roller (10) reaches the first test bench roller angular velocity (co 10 1 ) leaves, and an end of the acceleration process at which the at least one test bench roller (10) reaches the second test bench roller angular velocity (“io, 2), as well as from the first acceleration torque (M a ,i) , the first Test bench roller angular velocity (co 10 1 ) and the second test bench roller angular velocity (co 10 2 ) the test bench base inertia is determined.

4. Method according to one of the preceding claims, characterized in that a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, wherein in the course of the first inertial identification test - the at least one test bench roller (10) on a predetermined third test bench roller angular velocity (co 10 3 ) is accelerated, - at least one test bench roller (10) during a braking process starting from the third test bench roller angular velocity (co 10 3 ) with a, preferably constant, first braking torque (Mb,i) to a predetermined fourth test bench roller angular velocity (co 104 ) is slowed down, - from a time period (At) between the start of the braking process, at which the at least one test bench roller (10) leaves the third test bench roller angular velocity (“io, 3”), and the end of the braking process, at which the at least one test bench roller (10) leaves the fourth test bench roller angular velocity (co 104 ) achieved, as well as the first braking torque (Mb,i), the third test bench roller angular velocity (co 10 3 ) and the fourth test bench roller angular velocity (co 104 ) the test bench base inertia is determined.

5. Method according to one of the preceding claims, characterized in that a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, wherein in the course of the second inertial identification test - the at least one test bench roller (10) on a predetermined fifth test bench roller angular velocity (co 10 5 ) is accelerated, - at least one test bench roller (10) during an acceleration process starting from the fifth test bench roller angular velocity (co 10 5 ) with a second, preferably constant, acceleration torque (M a 2) to a predetermined sixth test bench roller angular velocity (co lo ,6) is accelerated, - from a time period (At) between a start of the acceleration process, at which at least one test bench roller (10) reaches the fifth test bench roller angular velocity (co 10 5 ) leaves, and an end of the acceleration process, at which at least one test bench roller (10) the sixth test bench roller- Angular velocity (co lo ,6) is reached, the second acceleration torque (M a , 2) the fifth test bench roller angular velocity (co 10 5 ) and the sixth test bench roller angular velocity (co lo,6) the test bench operating inertia is determined.

6. Method according to one of the preceding claims, characterized in that a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, wherein in the course of the second inertial identification test - the at least one test bench roller (10) on a predetermined seventh test bench roller angular velocity (co 10 7 ) is accelerated, - at least one test bench roller (10) during a braking process starting from the seventh test bench roller angular velocity (co 10 7 ) with a, preferably constant, second braking torque (Mb, 2) to a predetermined eighth test bench roller angular velocity (co lo ,8) is braked, - from a time period (At) between the start of the braking process, at which the at least one test bench roller (10) leaves the seventh test bench roller angular velocity (“io, 7”), and the end of the braking process, at which the at least one test bench roller (10) leaves the eighth test bench roller angular velocity (co lo ,8) is achieved, the second braking torque (Mb, 2) and the seventh test bench roller angular velocity (co 10 7 ) and the eighth test bench roller angular velocity (co lo ,8) the test bench operating inertia is determined.

7. Method according to claims 3 to 6, characterized in that the test rig base inertia describes a total rotational inertia of the test rig rollers (10) provided in the roller test rig (20) and of the rotors of the load machines (30) connected to the test rig rollers (10).

8. Method according to claims 3 to 7, characterized in that the test bench operating inertia describes a total rotational inertia of the test bench rollers (10) provided in the roller test bench (20) and the rotors of the load machines (30) connected to the test bench rollers (10) as well as additionally of the vehicle (1) to be tested.

9. Method according to one of the preceding claims, characterized in that the determined vehicle inertia of the vehicle (1) to be tested is used for a mass simulation.

10. Method according to claim 9, wherein a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, characterized in that, in the course of a mass simulation, a load moment (ML) output from a road load model (3) is scaled with a relation of the determined vehicle inertia and at least one roller inertia of the at least one test stand roller (10).

11. Method according to one of the preceding claims, wherein a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, characterized in that the first inertial identification test is carried out during commissioning of the roller test stand (20) or during maintenance of the roller test stand (20).

12. Method according to one of the preceding claims, wherein a roller test stand (20) with at least one test stand roller (10) is provided as a test stand, characterized in that the second inertial identification test is carried out during the operation of the roller test stand (20) for testing a vehicle.

13. Test stand, preferably roller test stand (20) or powertrain test stand or flat track test stand, for testing a vehicle (1) with - at least one load machine (30) formed from a stator and a rotor for driving or decelerating at least one wheel (9) of the vehicle (1) to be tested, - a test bench control (21) for controlling the load machines (30) provided on the test bench, characterized in that the test bench control (21) is designed, - to determine a test bench basic inertia by means of an initial inertia identification attempt; - to determine a test bench operating inertia by means of a second inertia identification attempt, - to determine a vehicle inertia of the vehicle (1) from the test bench basic inertia and the test bench operating inertia, preferably by subtracting the test bench basic inertia from the test bench operating inertia.

Images