METHOD FOR PERFORMING A TEST RUN ON A TEST STAND

Abstract

Various embodiments of the present disclosure are directed to a method for carrying out a test run on a test stand. The method in some embodiments reduces a deviation between a comparison simulation value and a comparison reference value when carrying out a test run on a test stand with a test object by simulating via a simulation unit a number of simulation values using a number of specified reference values starting from a selected reference value, determining a corrected reference value which is specified to the simulation unit for simulating a corrected simulation value and determining at least one setpoint variable using the corrected simulation value.

Claims

1. Method for carrying out a test run on a test stand with a test object, the method including the steps of: wherein specifying a number of reference values to a simulation unit by a reference unit, simulating a number of simulation values by the simulation unit using the number of reference values, determining at least one setpoint variable and at least one control variable for controlling the test object from the number of simulation values, and detecting, via a detection unit, a deviation between a comparison simulation value from the number of simulation values and a comparison reference value from the number of reference values, when a deviation is detected, starting from a selected reference value (from the number of reference values, determining a corrected reference value, via a correction unit, and instead of the selected reference value, the corrected reference value is specified to the simulation unit for simulating a corrected simulation value, whereby the deviation is reduced, and in that the at least one setpoint variable is determined using the corrected simulation value.

2. The method according to claim 1, further including the step of determining at least one further setpoint variable for at least one loading machine connected to the test object using the at least one corrected simulation value.

3. The method according to claim 1, further including the step of continuously increasing or decreasing the corrected reference value.

4. The method according to claim 1, further including the step of increasing or decreasing the corrected reference value until no more deviation occurs.

5. The method according to claim 1, characterized in that the method is started at the beginning of the test run.

6. The method according to claim 5, characterized in that the method is carried out during the entire test run.

7. The method according to claim 1, characterized in that the number of reference values includes a reference position at a reference time and a reference speed at the reference time, in that the number of simulation values simulated by the simulation unit includes a simulation speed at a simulation time and a simulation position at the simulation time.

8. The method according to claim 7, characterized in that the simulation position serves as a comparison simulation value, the reference position serves as a comparison reference value, and the reference speed serves as the selected reference value.

9. The method according to claim 8, characterized in that a corrected reference speed that is lower than the reference speed is specified to the simulation unit if the simulation position is greater than the reference position and the simulation speed exceeds a speed threshold.

10. The method according to claim 8, characterized in that a corrected reference speed that is greater than the reference speed is specified to the simulation unit if the simulation position is smaller than the reference position and the simulation speed exceeds a speed threshold.

11. The method according to claim 9, further including the step of changing the simulation time to a corrected simulation time if it deviates from the reference time and the simulation speed does not exceed the speed threshold.

12. Test stand with a test object for carrying out a test run, the test stand comprising: a reference unit configured and arranged to specify a number of reference values to a simulation unit, the simulation unit configured and arranged to simulate a number of simulation values using the number of reference values, determine at least one setpoint variable from the number of simulation values and transmit the at least one setpoint variable to a control unit, the control unit configured and arranged to specify at least one control variable for controlling the test object based on the at least one setpoint variable, a detection unit configured and arranged to detect a deviation between a comparison simulation value from the number of simulation values and a comparison reference value from the number of reference values, and a correction unit configured and arranged to, in response to a deviation (x) is detected, determine a corrected reference value starting from a selected reference value from the number of reference values and to specify the corrected reference value instead of the selected reference value to the simulation unit for simulating a corrected simulation value whereby the deviation is reduced and the at least one setpoint variable is determined using the corrected simulation value.

13. The method of claim 1, wherein the step of detecting a deviation between a comparison simulation value from the number of simulation values and a comparison reference value from the number of reference values further includes utilizing a tolerance.

14. The method of claim 9, wherein the speed threshold is zero.

15. The method of claim 10, wherein the speed threshold is zero.

16. The method of claim 11, wherein the speed threshold is zero.

17. The test stand of claim 12, wherein the detection unit is further configured and arranged to utilize a tolerance.

Description

[0034] The present invention is described in greater detail in the following with reference to FIGS. 1 to 7, which show advantageous embodiments of the invention by way of example, schematically, and in a non-limiting manner. In the drawings,

[0035] FIG. 1a shows a typical test stand setup for a test object,

[0036] FIG. 1b shows a possible execution of the simulation unit,

[0037] FIG. 2 shows a reference drive by way of example,

[0038] FIG. 3 shows a time-based specification of the simulation speed,

[0039] FIG. 4 shows a location-based specification of the simulation speed,

[0040] FIG. 5 shows a test stand setup according to the invention,

[0041] FIG. 6 shows a correction according to the invention of the reference speed,

[0042] FIG. 7 shows a correction according to the invention of the simulation time in a stop phase.

[0043] FIG. 1 shows a typical test stand 1 for a test object 2, in this case an engine test stand for an internal combustion engine. The test object 2 is connected to a loading machine 3 in this case, for example via a connecting shaft, as indicated in FIG. 1. The test object 2 can, however, also comprise a drive train, an entire vehicle, individual or multiple components, such as gearboxes, batteries, etc. Accordingly, the test stand 1 can represent, for example, a drive train test stand or a roller test stand, it also being possible for more than one loading machine 3, for example one per driven semi-axis or also per axis, to be provided. The test object 2 is operated on the test stand 1 according to the specifications of a test run in order to obtain statements regarding specific measured variables m of the vehicle, such as the pollutant emissions, consumption, the acoustic behavior of the vehicle, the drivability of the vehicle, the durability of the vehicle, notifications on the design/optimization of individual components, etc. In this context, the measured variable is compared with a target variable. If the measured variable relates to the pollutant emissions of an internal combustion engine or the consumption of an internal combustion engine, the test object 2 naturally also comprises an internal combustion engine.

[0044] A vehicle moving on a test track is simulated on a simulation unit 4. For this purpose, a number of reference values ref is specified to the simulation unit 4 by a reference unit 5. The simulation unit 4 determines a number of simulation values sim as part of the simulation. In the simulation unit 4, again at least one setpoint variable T, for example a torque, is determined from the number of simulation values sim, the at least one setpoint variable T being able to also correspond to a simulation value sim. The at least one setpoint variable T can also represent a pedal position of an accelerator pedal or can be calculated from a pedal position. The at least one setpoint variable T is transferred to a control unit ECU; the control unit ECU also controls the test object 2 based on the at least one setpoint variable T with at least one control variable. The control unit ECU, designed in this case as an engine control unit, can specify a throttle valve position a and/or a fuel quantity k (control variable) for the test object 2 based on a torque to be generated (setpoint variable T).

[0045] The simulation unit 4 can also feed at least one further setpoint variable, for example a speed n, to a further control unit 30, the further control unit 30 being able to control a loading machine 3, as shown in FIG. 1. The actual speed of the loading machine 3 acts in this case from the loading machine 3 via the shaft on the test object 2.

[0046] Driving robots can also be provided on a roller test stand, which actuate the vehicle's operating elements, such as the accelerator pedal, brake pedal, gear shift, in accordance with the specifications of the test to be carried out.

[0047] On the test stand 1, a number of measuring sensors (not shown in detail) is usually provided, with which, for example, actual values of the torque T.sub.ist and the rotational speed n.sub.ist of the test object 2 can be detected and transmitted to the simulation unit 4.

[0048] At the test stand 1, a test run is carried out with a specific test object 2 and, for example, pollutant emissions are measured as a measured variable. Depending on the measured variable, corresponding measuring units, such as an emissions measuring unit 6, which is supplied with exhaust gas of the internal combustion engine and which measures the specific pollutant emissions such as CO.sub.2, CO, NO.sub.x, total mass of hydrocarbons (THC), and/or the number of particles (such as soot particles) and/or a consumption measuring unit 7, which measures the fuel consumption of the internal combustion engine, can be provided on the test stand 1.

[0049] The simulation unit 4 and the control unit ECU can also be designed as one unit or, as shown in FIG. 1, can be designed as independent units. The reference unit 5 is designed as an independent unit in FIG. 1, but can also be combined with the simulation unit 4. The simulation unit 4 has simulation hardware and/or simulation software with which the test drive of the vehicle is simulated. For this purpose, a simulation model, which comprises, for example, a driver model 11, a vehicle model 12, and a route model 13, is implemented in the simulation unit 4. Further models, such as a tire model, a road model, etc., can also be implemented. The simulation unit 4 thus simulates the drive of a virtual vehicle (vehicle model 12), which is controlled by a virtual driver (driver model 11), along a virtual test route (environment model 13), specific events such as traffic signs, traffic lights, foreign traffic, etc., being able to be simulated. Events are perceived by the virtual driver in the driver model 11 and implemented in the form of appropriate reactions. Part of the vehicle, such as the internal combustion engine or a drive train, is physically constructed as real hardware on the test stand 1 as the test object 2 and is operated with the specifications of the simulation according to the test on the test stand 1. This procedure for carrying out a test is well known and is often referred to as an X-In-The-Loop test, where the “X” stands for the respective test object 2 that is actually present. This type of test execution is very flexible and can come very close to the character of a real test drive with a real vehicle. Variations for worst-case estimates can also be made and evaluated. Variations can be defined very abstractly, e.g. in the form of more vehicle mass, more traffic, strong headwind, more aggressive driving behavior, etc.

[0050] With the test run created in this way, the development of the vehicle can therefore be carried out in all development stages and the probability of compliance with specific specifications of the target variable, for example legal limit values for pollutant emissions during a check with an RDE test procedure, can be significantly increased. The same applies in an analogous manner to other measured variables, such as consumption, drivability, acoustic behavior, durability, instead of pollutant emissions.

[0051] The test run usually contains many different driving maneuvers, e.g. acceleration, deceleration, standstill, constant travel, cornering, etc. under specific boundary conditions, such as speed, torque, steering angle, road gradient, traffic, etc. understood. As driving maneuvers, starting from standstill, accelerating out of a curve, changing the vehicle speed, overtaking a slow vehicle, coasting to a red light, etc. can be implemented. Every drive of a vehicle and thus also a test run can be seen as a chronological sequence of such driving maneuvers. It is immediately evident that there can be an abundance of such driving maneuvers. The driving maneuvers are stored in the reference unit 5 and come, for example, from real, measured test drives, from simulations that have already been carried out, etc.

[0052] A test run is now created as a time sequence of such driving maneuvers. This can be done manually by a user, by randomly selecting the driving maneuvers or by a specific selection. The driving maneuvers must of course be linked to one another in such a way that there are no discontinuities, such as sudden jumps in speed, during the test run. It must also be ensured that the test object 2 can follow the desired specifications in combination with the simulation model. The test run should include many different driving maneuvers, which should preferably cover the largest possible operating range (speed, torque) of the vehicle.

[0053] Thus, there can be precise specifications as to which maneuvers must be included in which proportions.

[0054] The simulation unit 4 receives, as mentioned, a number of reference values ref from the reference unit 5 corresponding to the currently desired driving maneuver within the scope of the test run for carrying out the simulation. A reference speed v_ref and a reference position s_ref, in each case as a function of the reference time t_ref, serve as reference values ref. A reference speed v_ref corresponding to the test run is thus specified to the simulation unit 4, which reference speed is simulated by the driver model 11, for example. The driver model 11 thus follows the reference speed v_ref which is calculated using the vehicle model 12 and the route model 13.

[0055] The test run or the individual driving maneuvers are specified in the reference unit 5 as the course of the reference speed v_ref over the reference time t_ref and are transmitted to the simulation unit 4 as reference values ref. In the context of the simulation, the simulation unit 4 tries to follow the reference speed v_ref with a simulation speed v_sim. This can be location-based (i.e., the simulation speed v_sim at the simulation position s_sim always corresponds to the reference speed v_ref at the reference position s_ref) or time-based (i.e., the simulation speed v_sim at the simulation time t_sim always corresponds to the reference speed v_ref at the reference time t_ref). However, since the simulation speed v_sim can never exactly follow the reference speed v_ref, a time deviation arises in the case of a location-based approach and a position deviation in a time-based approach.

[0056] FIGS. 2, 3, 4, 6, and 7 show in each case a time-speed-path diagram, the time t being plotted on the positive abscissa, the speed v on the negative abscissa, and the position s on the positive ordinate. This results in a speed-time relationship in the left part of the diagram and a distance-time relationship in the right part of the diagram.

[0057] In FIG. 2, a reference drive results from corresponding reference values ref: A reference speed v_ref is specified for each reference time t_ref and a reference position s_ref is also specified for each reference time t_ref, with which the reference curves are formed in each case: On the left, a reference speed v_ref as a function of the reference position s_ref and, on the right, a reference position s_ref at the reference time t_ref. As a result of this relationship, a reference speed v_ref is also specified for each reference time t_ref.

[0058] In addition to the reference values ref, FIG. 3 shows the simulation values sim of a simulation, with a time-based specification of the simulation speed v_sim taking place. The simulation speed v_sim thus follows the reference speed v_ref at every point in time of the simulation time t_sim. Since the simulation speed v_sim cannot follow the reference speed v_ref exactly, there is nevertheless a speed deviation v_x between the simulation speed v_sim and the reference speed v_ref, as can be seen in the left quadrant of the graph. The simulation speed v_sim is too low in this case. As a result, a spatial deviation s_x between the simulation position s_sim and the reference position s_ref results as the deviation x, as can be seen in the right part of the graph.

[0059] FIG. 4 shows simulation values sim which are determined via a location-based specification of the simulation speed v_sim. The simulation speed v_sim thus follows the reference position s_ref at every point of the simulation position s_sim. Since the simulation speed v_sim again cannot follow the reference speed v_ref exactly, there is also a speed deviation v_x between the simulation speed v_sim and the reference speed v_ref (again shown by way of example as too low a simulation speed v_sim), which is why a time deviation t_x between the simulation time t_sim and the reference time t_ref results as the deviation x.

[0060] As can be seen in FIG. 5, according to the invention, a detection unit 7 and a correction unit 8 are provided on the test stand 1, which are connected in this case as a unit between the simulation unit 4 and the reference unit 5. This is particularly advantageous because it does not intervene in the simulation unit 4 itself; only the corresponding simulation values sim need to be available in order to compare them with the equivalent reference values ref. In FIG. 5, the simulation position s_sim is fed to the detection unit 7 from the simulation unit 4.

[0061] The deviation x between a comparison simulation value and a comparison reference value is now determined by means of a detection unit 7. In this case, the spatial deviation s_x between the simulation position s_sim as a comparison simulation value and the reference position s_ref as a comparison reference value is determined as the deviation x. Of course, a tolerance (band) can be provided. By means of the correction unit 8 advantageously integrated in this case into the detection unit 7, the reference speed v_ref is changed as selected reference value ref to the corrected reference speed v_ref′, which is then made available instead of the reference speed v_ref to the simulation unit 4. A corrected simulation position s_sim′ is thus simulated in the simulation unit 4 in the further course, which results in a smaller spatial deviation s_x.

[0062] In FIG. 6, a spatial deviation s_x, in this case a simulation position s_sim that is too small compared to the reference position s_ref, is recognized at the point in time t1. This spatial deviation s_x can arise because the simulation speed v_sim is too low, as is also indicated in the present drawings. Compared to FIG. 4, it can be seen that after the point in time t1 of the simulation unit 4, instead of the reference speed v_ref (as selected reference value ref) a corrected reference speed v_ref′, which is higher in this case, is specified, which is shown by the dashed portion. Based on this, a higher simulation speed v_sim is simulated in the simulation unit 4 since this follows the now corrected reference speed v_ref′. The corrected reference value ref (in this case the corrected reference speed v_ref) can be continuously increased or decreased.

[0063] As a result, starting from the time t1, instead of the simulation speed v_sim, the corrected simulation speed v_sim′ is specified to the simulation unit 4 and the spatial deviation s_x is thus reduced; in the illustrated case no more spatial deviation s_x occurs until time t2. The corrected reference speed v_ref is then retained in order to keep the spatial deviation s_x at zero. The simulation speed v_sim thus roughly follows the reference speed v_ref and the simulation position s_sim follows the reference position s_ref.

[0064] This has the consequence that the simulation unit 4, using the at least one corrected simulation value sim′, forwards at least one setpoint variable T, which is now corrected, to the control unit ECU. Using this at least one setpoint variable T, the control unit ECU controls the test object 2 with at least one control variable that is also corrected. Using the corrected simulation value sim′, at least one further setpoint variable n can also be determined for the further control unit 30 of the loading machine 3.

[0065] The method could of course also be started at the beginning of the test run and preferably be carried out during the entire test run. A slight deviation x thus occurs during the simulation since this is preferably corrected continuously and as best as possible.

[0066] However, it has to be always ensured in the present embodiment that the simulation speed v_sim exceeds a first speed threshold for a reduction of the reference speed v_ref (as selected reference value ref) to a corrected reference speed v_ref. For an increase in the reference speed v_ref, it must also be ensured that the simulation speed v_sim exceeds the speed threshold. In particular, the simulation speed v_sim must not be zero in this case.

[0067] At low reference speeds v_ref and thus also low simulation speeds v_sim, the possibility of a correction by adapting the reference speed v_ref as selected reference value ref is low, in particular if the simulation position s_sim is greater than the reference position s_ref. Since the reference speed v_ref is low, it can of course no longer be reduced much before it reaches zero.

[0068] In FIG. 7, the reference speed v_ref becomes zero at the time t4. Since a time deviation t_x occurs, the simulation speed v_sim does not become zero until the time t3. Thus, the speed threshold (of zero) is reached at the time t3 by the simulation speed v_sim, with which the speed threshold is no longer exceeded, which is preferably recognized by the detection unit 7. The simulation time t_sim which deviates from the reference time t_ref is thus changed, preferably by the correction unit 8. At the time t5, the reference speed v_ref and thus the simulation speed v_sim are increased again. So that these times are synchronized again, the duration in which the simulation speed v_sim is zero is reduced. In accordance with the reference, the stop duration would have to last from t4 to t5; in the simulation the stop duration was reduced to t3 to t5. The change in the simulation time t_sim can also be viewed as a faster or slower expiry of the simulation time t_sim or the reference time t_ref in the changed range of the simulation time t_sim.