Abstract
A method and vehicle test stand comprising at least one actuator for moving the vehicle in a longitudinal direction. During the test, a rotational movement which is carried out by a wheel or a powertrain of the vehicle is measured in real time, a longitudinal acceleration corresponding to the measured rotational movement is ascertained, and the at least one actuator is actuated based on the ascertained longitudinal acceleration. In at least one example, the vehicle may be connected to an acceleration sensor, an acceleration signal of the acceleration sensor is detected during the test and a low-frequency longitudinal acceleration component is calculated based on the ascertained longitudinal acceleration and a position control loop for controlling the actuator, and the vehicle longitudinal acceleration is ascertained based on the detected acceleration signal and the calculated low-frequency longitudinal acceleration component.
Claims
1. A method for ascertaining a vehicle longitudinal acceleration during a test of a vehicle in a vehicle test stand, comprising at least one actuator for moving the vehicle in a longitudinal direction, wherein, during the test, a rotational movement which is carried out by a wheel or a powertrain of the vehicle is measured in real time, the vehicle longitudinal acceleration corresponding to the measured rotational movement is ascertained, and the at least one actuator is actuated based on the ascertained longitudinal acceleration, wherein the vehicle is connected to an acceleration sensor an acceleration signal of the acceleration sensor is detected during the test and a low-frequency longitudinal acceleration component is calculated based on the ascertained longitudinal acceleration and a position control loop for controlling the actuator, and the vehicle longitudinal acceleration is ascertained based on the detected acceleration signal and the calculated low-frequency longitudinal acceleration component.
2. The method according to claim 1, wherein the vehicle longitudinal acceleration is ascertained by superposing the detected acceleration signal and the calculated low-frequency longitudinal acceleration component.
3. The method according to claim 1, wherein the low-frequency longitudinal acceleration component is calculated as the difference between the ascertained longitudinal acceleration and a high-frequency longitudinal acceleration component ascertained from a manipulated variable of the position control loop.
4. The method according to claim 3, wherein the manipulated variable specifies the force to be applied to the vehicle by the actuator, wherein the high-frequency longitudinal acceleration component is calculated from a sum of the manipulated variables of the at least one actuator divided by the mass of the vehicle.
5. A vehicle test stand for ascertaining a vehicle longitudinal acceleration during a test of a vehicle, comprising: a measuring device for measuring a rotational movement of a wheel or of a powertrain of the vehicle, an actuator for moving the vehicle in a longitudinal direction, and a control device, which is connected to the measuring device and to the actuator and, during the test, is designed to ascertain a corresponding longitudinal acceleration in real time from a measured rotational movement obtained from the measuring device and to transmit a control signal to the actuator based on the ascertained longitudinal acceleration, wherein the control device is connected to an acceleration sensor in or on the vehicle and is designed, during the test, detect an acceleration signal of the acceleration sensor calculate a low-frequency longitudinal acceleration component based on the ascertained longitudinal acceleration and a position control loop for controlling the actuator, and ascertain the vehicle longitudinal acceleration based on the detected acceleration signal and the calculated low-frequency longitudinal acceleration component.
6. The vehicle test stand according to claim 5, wherein the control device is further configured to ascertain the vehicle longitudinal acceleration by superposing the detected acceleration signal and the calculated low-frequency longitudinal acceleration component.
7. The vehicle test stand according to claim 5, wherein the control device is further configured to calculate the low-frequency longitudinal acceleration component as the difference between the ascertained longitudinal acceleration and a high-frequency longitudinal acceleration component ascertained from a manipulated variable of the position control loop.
8. The vehicle test stand according to claim 7, wherein the manipulated variable specifies the force to be applied to the vehicle by the actuator, wherein the control device is further configured to calculate the high-frequency longitudinal acceleration component from the sum of the manipulated variables of the one or more actuators divided by a mass of the vehicle.
Description
BRIEF DESCRIPTION OF THE FIGURES
(1) The invention will be explained in further detail hereinafter on the basis of particularly preferred exemplary embodiments, to which the invention is not limited, however, and with reference to the drawings. Specifically, the drawings show schematically:
(2) FIG. 1 a total vehicle test stand with actuators for moving the vehicle in a longitudinal direction;
(3) FIGS. 2a and 2b (simulated) the temporal progression of a vehicle longitudinal acceleration in a tip-in test, on the road (FIG. 2a) and calculated from a vehicle model (FIG. 2b);
(4) FIG. 3 a heavily simplified block diagram for ascertaining a vehicle longitudinal acceleration on a test stand according to FIG. 1;
(5) FIG. 4 the temporal progression of the signals in the block diagram according to FIG. 3;
(6) FIG. 5 a block diagram for ascertaining a vehicle longitudinal acceleration on a test stand with a control device for an actuator for transmitting a longitudinal force to a vehicle; and
(7) FIG. 6 a block diagram for ascertaining a vehicle longitudinal acceleration in accordance with the invention on a test stand with a control device for an actuator for transmitting a longitudinal force to a vehicle.
DETAILED DESCRIPTION
(8) The technology of powertrain test stands has been further developed in recent years such that the device under test, i.e. the vehicle 2 (also “DUT” for short), experiences exactly the same loads on what is referred to as a vehicle test stand 1 or total vehicle test stand (see FIG. 1) as on the road. A “vehicle application on the test stand” in which, instead of time-consuming and costly driving tests on a test track, the same tests can be performed on the test stand is thus made possible (see Pillas J., Kirschbaum F., Jakobi R., Gebhardt A., Uphaus F.: Model-based load change reaction optimization using vehicle drivetrain test beds. Conference proceedings of the 14th International Stuttgart Symposium, pages 857-867, 2014).
(9) Here, mathematical models of the vehicle 2, of the wheels and of the tyres are also calculated online on the test stand, and these models are applied to the wheeled machines 3 on the basis of the calculated values. If a simple one-mass model is used as model 2′ for the vehicle 2, however, it has been found that the longitudinal acceleration a calculated by the model 2′ is too “smooth” on the test stand: In the case of an exemplary tip-in test (sudden increase in the acceleration pedal value), a fluctuating signal profile 4 (see FIG. 2a) is achieved on the road by the vehicle's own longitudinal acceleration sensor, whereas the longitudinal acceleration signal 5 calculated by the model 2′ does not show this fluctuation at the test stand (see FIG. 2b). This difference between road measurement and test-stand measurement is problematic if an objective assessment of the driveability of the current vehicle application is to be performed on the basis of this signal profile.
(10) The reason for the deviation of the longitudinal acceleration a calculated by the model 2′ lies in the modelling of the vehicle 2 as a one-mass model. In fact, the vehicle 2 consists of a number of masses, which are connected to one another via various spring-damper elements. For example, the axle tree bolster may vibrate relative to the rest of the body and may contribute to the signal profile 4 shown in FIG. 2a. The present invention is based on the finding that, on the vehicle test stand 1, the corresponding parts of the real vehicle 2 provided on the total vehicle test stand can be used for the ascertainment of a realistic vehicle longitudinal acceleration a.sub.S instead of the simulation or in addition to the simulation. This presupposes that the vehicle test stand 1 comprises at least one actuator 6 for moving the vehicle 2 in a longitudinal direction, and, during the test, a rotational movement performed by a wheel or a powertrain of the vehicle 2 is measured in real time, a longitudinal acceleration a corresponding to the measured rotational movement is ascertained, and the at least one actuator 6 is actuated based on the ascertained longitudinal acceleration a, in particular on the basis of a high-frequency acceleration component a.sub.HF (also “filtered” or “subjective” acceleration). A corresponding vehicle test stand 1 is described for example in WO 2015/157788 A1 and can be realised in accordance with the exemplary embodiments disclosed there.
(11) Proceeding from a vehicle test stand 1 of this kind, the present method provides that the vehicle 2 is connected to an acceleration sensor 7 (see FIG. 5). During the test the longitudinal acceleration a calculated by the model 2′ is split into a low-frequency longitudinal acceleration component a.sub.NF and a high-frequency longitudinal acceleration component a.sub.HF, for example with the aid of filters 8, 9 (see FIG. 3). The high-frequency longitudinal acceleration component a.sub.HF is applied at the test stand to the real vehicle 2 provided, via the actuators 6 (for example linear motors). During the test an acceleration signal a.sub.S,HF at or in the vehicle 2 is detected by means of the acceleration sensor 7. The acceleration signal a.sub.S,HF thus detected is then suitably superposed with the calculated low-frequency longitudinal acceleration component a.sub.NF so as to obtain the sought vehicle longitudinal acceleration a.sub.S.
(12) If only the low-frequency longitudinal acceleration component a.sub.NF were applied to the vehicle test stand 1 (which doesn't work due to the necessary distance), exactly the same profile a.sub.S,NF would be measured by the acceleration sensor 7, since the vehicle's internal vibration frequencies are higher. Thus, there is no need to apply the low-frequency longitudinal acceleration component a.sub.NF at the vehicle test stand 1, and the sought longitudinal acceleration signal a.sub.S can be obtained by superposition by means of addition 10:
a.sub.S=a.sub.S,NF+a.sub.S,HF=a.sub.NF+a.sub.S,HF
(13) FIG. 4 by way of example shows a signal profile 11 of the low-frequency longitudinal acceleration component a.sub.NF, a signal profile 12 of the high-frequency longitudinal acceleration component a.sub.HF calculated by the model 2′, and a signal profile 13 of the acceleration signal a.sub.S,HF measured on the test stand.
(14) The low-frequency longitudinal acceleration component a.sub.NF can be assumed in the simplest case to be the complement of the high-frequency longitudinal acceleration component a.sub.HF to the ascertained longitudinal acceleration a. This case is shown in FIG. 5: this shows schematically a vehicle test stand 1 with a control device 14 for an actuator 6 of the vehicle test stand 1. The actuator 6 is formed by a linear motor (transmission function P(s)) for transmitting a longitudinal force to a vehicle 2 (DUT, or “device under test”) connected to the vehicle test stand 1. In order to measure a rotational movement of a powertrain or a wheel of the vehicle 2, the control device 14 is connected to a measuring device 15, for example in the form of a torque sensor. The control device 14 is designed to firstly ascertain a longitudinal acceleration a of the vehicle 2 corresponding to the measured torque value M.sub.ist according to a model 2′ of the vehicle 2. The longitudinal acceleration a thus ascertained is then modified in a high-pass filter 16, wherein low-frequency components of the acceleration a are eliminated, such that the high-frequency longitudinal acceleration component a.sub.HF corresponds to a subjective acceleration. The high-frequency longitudinal acceleration component a.sub.HF is weighted accordingly, i.e. multiplied by a vehicle mass m and divided by a number N of linear motors (oriented substantially in parallel) on the test stand. The resultant high-frequency acceleration force is added at the output of a position controller 17 as disturbance variable so to speak to the current manipulated variable value of the controller to form the desired air-gap force F.sub.LS of the actuator 6. The position controller 17 is designed for practically unnoticeable resetting of the vehicle 2, and therefore works only with slow or low-frequency accelerations. The low-frequency longitudinal acceleration component a.sub.NF of the ascertained longitudinal acceleration a damped or removed by the high-pass filter 16 is thus effectively replaced at the vehicle test stand 1 by the current manipulated variable value of the position controller 17. The entire processing starting with the measurement of the torque is performed in real time, i.e. without noticeable delays. The position controller 17 is part of a position control loop 18 with a position measurement of the actuator 6, which ascertains the current position x.sub.ist of the rotor of the actuator 6, and with a difference member 19, which compares the current position x.sub.ist with a constant, predefined setpoint position x.sub.soll and transmits the difference, which corresponds to a deflection of the rotor from the setpoint position x.sub.soll, to the position controller 17. The position controller 17 ascertains a current manipulated variable value for the air-gap force F.sub.LS of the actuator 6 from the obtained deflection. The position controller 17, to this end, can comprise a deflection controller (for example as P controller) for ascertaining a setpoint speed and, connected thereto, a speed controller (for example as PI controller) for ascertaining the current manipulated variable value for the air-gap force FLS of the actuator. The low-frequency longitudinal acceleration component a.sub.NF corresponds to the difference 20 between the longitudinal acceleration a ascertained by the model 2′ and the high-frequency longitudinal acceleration component a.sub.HF.
(15) FIG. 6 shows the vehicle test stand 1 according to the invention with a control device 14 for calculating an alternative low-frequency longitudinal acceleration component a′NF. The test method utilises the fact that the actuators 6, as shown in FIG. 1, act on the wheels, and therefore the vehicle's internal dynamics from the wheel to the acceleration sensor 7 is excited. In order to take these dynamics into consideration, the control device 14 according to FIG. 6 is designed to calculate the low-frequency longitudinal acceleration component a′.sub.NF based on the ascertained longitudinal acceleration a and additionally based on the position control loop 18 of the actuator 6, specifically as the difference 21 between the ascertained longitudinal acceleration a and a modified high-frequency longitudinal acceleration component a′.sub.HF ascertained from the manipulated variable of the position control loop 18 (i.e. the air-gap force F.sub.LS). This is calculated from the sum 22 (shown by way of simplification as factor N) of the manipulated variables of the one or more actuators 6 divided by the mass m of the vehicle 2. In other words, instead of the longitudinal acceleration components calculated purely from the model 2′, the longitudinal forces calculated from the model 2′ and modified by the position controller 17 are used on the tyres, with the advantages described at the outset. The vehicle longitudinal acceleration as is then ascertained as the sum 23 (i.e. by addition) of the previously detected acceleration signal a.sub.S,HF and modified low-frequency longitudinal acceleration component a′.sub.NF.
