• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    In order to reduce the excitation of vibrations and resonances in a test stand (1) for a real component (4) and a virtual component (5), one of the method steps is a) determining a first correction value (K1) from the measured variable (M), wherein the first correction value (K1) is added to the measured variable (M) and the sum is transmitted as corrected measured variable (M *) of the virtual component (5) for calculating the control variable (S), b) determining a second correction value (K2) from the calculated control variable (S), wherein the second correction value (K2) is added to the calculated control quantity (S) and the sum is transmitted as a corrected control quantity (S *) to the actuator (3), c) determining a third correction value (K3) from the measured variable (M ), wherein a parameter (P) of the equation of motion is changed with the third correction value (K3).
    公开号:AT512483A2
    申请号:T50369/2013
    申请日:2013-06-03
    公开日:2013-08-15
    发明作者:
    申请人:Avl List Gmbh；
    IPC主号:
    专利说明:

    Printeci: 04-06-2013 E014.1
    10 2013/50369 AV-3543 AT
    Method for reducing vibrations in a test bench
    The present invention relates to a method for reducing the excitation of unwanted oscillations and resonances in a test stand for a real component 5 and a virtual component, wherein the real component provides a measured variable of the real component to the virtual component and receives a control variable for an actuator of the test bench from the virtuaikomponente wherein a simulation model with an equation of motion is implemented in the virtual component, which determines the control variable from the measured variable. In vehicle technology, the process of testing is often done so that real components, such as real internal combustion engines, real tires, real gears, real batteries, real steering systems, real powertrains, real vehicles, etc., are placed on test benches. This real component to be tested often also specifies the name of the test bench. These test benches allow, for example, the development of internal combustion engines, of vehicle components or the detection of errors in networked vehicle control units, which can affect the overall behavior of the vehicle. Testing is a process by which greater certainty is to be gained as to whether technical objects, technical systems or processes, the real component or the virtual component function within certain boundary conditions and / or whether certain properties and / or requirements be fulfilled. Performed tests thus simulate or anticipate real processes in simulated environments. In the most general case, the simulated environment exchanges material flows with the tested real component (eg a medium, such as oil, water, etc.), energy flows (eg electrical current / voltage, rotational speed / torque, etc.) and information flows (eg measured data, etc.) and thus enables the investigation of technical processes without presupposing, impairing or jeopardizing the future real environment of the real component. Therefore, a test result is never absolutely valid, but always represents an approximation. The quality of the approximation depends, among other things, on the quality of the simulated environment and on the quality with which the actual exchange of energy and information takes place - And streams can be modeled. This simulated environment is referred to below as a virtual component. The real component and the virtual component together are referred to as the test item. The test object and the test stand together are often used as a hardware-in-the-loop system (HiL system) or more specifically as an "X-in-the-loop system", where X for the respective candidate stands. -1-
    Nntei WimmiS E014.1 102013/50369
    AV-3543 AT
    A virtual component consists of simulation models that are essentially implemented as software with implemented algorithms and mathematical or physical models that are executed on a simulation unit, usually a computer. As a rule, actuators (a number of actuators) and sensors (a number of sensors) are also present on the test bench for carrying out the tests, as well as possibly a sequence control (eg a test bench control unit, an automation unit, etc.) and peripherals (such as a Data logger, etc). The sensors measure physical, chemical or information technology states or state changes ("measured variables") of the real component, and the actuators impose certain chemical, physical or information technology states or state changes ("nominal values") on the real components. Actuators are thus the signal converter's counterpart to sensors. Actuators and sensors connect the real with the virtual world of the test object, ie the real component and the virtual component. Examples of actuators are electric, pneumatic or hydraulic Belas-15 processing units for impressing speeds, torques, speeds or paths, variable electrical resistances, oil conditioning equipment, air conditioning, etc. Examples of sensors are torque sensors and encoders.
    Real component, virtual component, actuators and sensors are dynamic systems with a certain transfer behavior. Thus, a hardware-in-the-loop system 20 as interconnection of these components is also a dynamic system.
    An example of a test is a virtual test drive of a hybrid vehicle (combustion engine and electric motor) on the Großglockner High Alpine Road with realistic replica of the air humidity, the air temperature, the speed and torque behavior of the real component "internal combustion engine", which is mounted on an engine test bench is. The aim of this test drive is the assessment of the dynamic behavior of the electric motor and the temperature behavior of the traction battery, which are simulated as a virtual component, for a particular driver type, e.g. a sporty driver with aggressive shifting behavior. The test track (here the Grossglockner High Alpine Road), the handling and the driving environment are also simulated. In this test drive, the hardware-in-the-loop system is excited to vibrate over road bumps, gusts, driver braking and steering and / or combustion. However, due to the dynamic behavior of the sensors and actuators and due to the simulation accuracy of the virtual component, which is always limited by the simulation, these oscillations will not be exactly identical to the vibrations that occur during a real journey with the hybrid vehicle via the Großglockner High Alpine Road. -2-
    AV-3543 AT
    Another example is shown in EP 1 037 030 B1, which discloses a method for simulating the behavior of a vehicle on a roadway on a powertrain test bench using a vehicle model and a tire model (virtual components) for simulation.
    In practice, the virtual components are often retrofitted to existing test bed infrastructures. A classic, traditional test bench, which has so far been able to impose only simple setpoint profiles, thus becomes a powerful X-In-The-Loop test environment, which makes it possible to present new test tasks, such as the Großglockner High Alpine Ride described above with different framework conditions. The existing Prüfstandsaktuatorik and test bench sensors with their superimposed dynamic subsystems and controller structures here (for example, for cost reasons) often remain unchanged or it is unknown to the supplier of the virtual component. The same virtual component is often also different «! Test stands with different dynamic transmission characteristics or used on different test bench types. Likewise, a virtual component may be replaced by another virtual component (e.g., modified models).
    Another problem with such virtual components can arise at the test stand, if the virtual components are to represent extreme load cases that go to or beyond the limits of the implemented actuators, sensors or the real component.
    Due to the dynamic transmission behavior of the actuators and sensors installed on the test stand, but also due to the interference always present in the available measurements (eg measurement noise, limited resolution, etc.), unwanted, unexpected and unrealistic oscillation and resonance phenomena of the overall dynamic system often occur. which can adversely affect the test results and, in extreme cases, even make the use of virtual components even fail.
    Classically, this scenario could be counteracted by the use of filters (e.g., Bessel filters, Butterworth filters, etc.) for vibration damping, however, limiting the available dynamic of the test bench, which is undesirable. Test situations with high dynamics, e.g. a very rapid speed or torque change, could then no longer be carried out. Another important negative feature that occurs when using such filters is the distortion of important dynamic states during testing. As an example, in mechanical / rotary test benches (for example, powertrain), the angular momentum is called, which is exchanged between the real component and the virtual component. The use of filters here causes the actually applied angular momentum (for example from the internal combustion engine) to be incorrectly introduced into the virtual component, which in turn leads to -3-
    AV-3543 AT leads to incorrect test results (for example too high / too low fuel consumption). In addition, filters always cause a phase shift, which, among other things, negatively influences the stability reserve of the HiL system.
    It is therefore an object of the subject invention to provide a method, can be operated with the virtual components on test benches largely without limiting the dynamic behavior and largely without unwanted vibration and resonance effects.
    This object is achieved in that, at least from the measured variable, a first correction value is determined which is added to the measured variable and the sum is transmitted as corrected measured variable of the virtual component for calculating the control variable or a second correction value is determined from the calculated control variable calculated control variable is added and the sum is transmitted as a corrected control variable of the actuator or a third Korrektunwert is determined, which changes a parameter of the equation of motion. The first, second or third correction value can also be combined as desired.
    Thus, unwanted vibration and resonance effects on the test stand can be effectively suppressed, as far as possible without restricting the dynamics of the test stand and without interfering with the controller structure or the test bench sensor system underlain by the actuators. This also makes it possible to retrofit existing test environments with Virtu-alk components or to replace virtual components with other virtual components without having to make any changes to the existing test stand infrastructure (sensors, actuators). In this way, any desired system-dynamic interventions in the test bench can be carried out by interventions or additions in the virtual component, and, above all, independently of the already existing test bed infrastructure.
    The torque of a shaft between the real component and the actuator is very particularly advantageously used as the measured variable. This allows the "Shapes " the torques measured at the test bench by virtue of additional virtual torques, so that the torque impressed on the virtual world changes in a suitable manner continuously (as a function of time) so that no unwanted vibrations occur in the virtual system.
    Likewise advantageously, a speed is used as the control variable. The rotational speeds resulting in the virtual world of the simulation are suitably "reshaped" in such a way that no unwanted vibrations or resonances occur in the real system.
    The correction values can advantageously be determined by optimizing a target function according to the respective correction value. Such objective functions can be optimized with known mathematical methods, preferably on real-time computers in real time. -4-
    Printed: 04-06-2013 £ 014,1 tlÖ201ä / S0369
    AV-3543 AT
    In order to determine the first or third correction value, a linear combination of a first and a second objective function is preferably optimized, since in this way different physical effects similar influencing variables, such as e.g. Energy or angular momentum, can be considered. For this purpose, a quadratic quality function as a function of the angular velocity or a derivative thereof is advantageously used as the first or third objective function. The second objective function advantageously evaluates the angular momentum introduced by the first or third correction value or the altered kinetic energy, thereby ensuring that the correction does not cause excessive distortions of the rotational movement or the energy balance or the momentum equations of the shaft. To determine the second correction value, a target function is preferably implemented which evaluates the deviation between the control variable calculated in the virtual component and the actual value of this control variable. For many test bed types, such as Powertrain test benches or engine dynamometers, it is advantageous to determine a correction torque as the first correction value and / or to determine a correction speed as the second correction value. Torque and speed are the usual measurement and control variables and usually available as measured values in such test benches, so that their use is advantageous.
    As the third correction value, preferably a massed parameter of the equation of motion is used, e.g. an inertial momentum or a mass, with which the virtual component can be easily influenced by the equation of motion.
    Very particularly advantageous boundary conditions for the consideration of predetermined restrictions of the virtual component or the real component or the actuator system can be taken into account in the optimization. In this way, physical limits of the test stand can be taken into account, which also effectively protects the components of the test bench from damage, e.g. due to excessive torques, accelerations, speeds, etc. represents. In addition to optimizing the objective functions, the optimization algorithm in this case will typically consider equality or inequality constraints. As a result, undesired vibrations in the HiL system can be reduced on the one hand, especially in demanding test scenarios (for example, driving over bumps or sleepers), while on the other hand, it is possible to guarantee that test bench limitations are minimized while at the same time maximizing the reality claim.
    The subject invention will be explained in more detail with reference to Figures 1 to 4, which show exemplary, schematic and non-limiting advantageous embodiments of the invention. This shows -5-
    Printed: 04-06-2013 i0l4,1 102013/50369
    AV-3543 AT
    1 and 2 examples of a configuration of a hardware-in-the-loop test environment,
    3 shows the correction of the measured variable or the control variable according to the invention, and FIG. 4 shows the determination of the first correction value using the example of a wheel simulation model. Figure 1 illustrates the basic configuration of a hardware-in-the-loop test environment. On a test bench 1, e.g. an engine test bench, is a real component 4, e.g. an internal combustion engine, which is connected via a connecting shaft 2 to an actuator 3, e.g. a loading machine in the form of an electric dynamometer, is connected.
    The virtual component 5 consists of a simulation model 21, e.g. a vehicle simulation model 6, an environmental simulation model 7, a driver simulation model 8, a road simulation model 9, a wheel simulation model 10, etc., which are provided as software in a simulation device 17, e.g. in the form of a computer with required software and implemented algorithms. Depending on the test run, different and several such component simulation models, which together form the simulation model 21, can be used. In the virtual component 5, the vehicle or a component thereof is moved through a virtual world. Real component and virtual component interact via input interfaces 11 (data from sensor 18) and output interfaces 12 (data to actuators). At the test stand 1, the respective current virtual state is controlled by the virtual component 5 at the real component 4 and set at the actuator 3 so that the real component 4 experiences the states from the virtual component 5, ie the virtual world, and via the time sequence of these states is tested.
    For this purpose, at the test bench 1, e.g. the torque T between real component 4 and actuator 3 or the speed n of the real components or the actuator 3 (e.g., in the form of an electric loading machine) are measured by suitable sensors 18, e.g. via a torque measuring device 25 on the connecting shaft 2 or a speed measuring device of the simulation device 17, and the virtual component 5 via an input interface 11 is provided. From this measured variable M (torque T or rotational speed n), the simulation model 21 calculates, in the simulation device 5, a control variable S for the actuator 3, for example after a suitable signal conditioning. a target rotational number n, a control variable for the real component 4, e.g. a throttle position a, etc.
    These control variables S are transferred via an output interface 12 of the simulation device 17 to the test bench 1 and set on the test bench 1 of the actuator 3 and possibly other suitable actuators, not shown, possibly by means of suitable control units. However, the measured variable in the sense of the present method does not have to be measured directly, but can also be derived or formed from other measured quantities.
    Printed: 04-06-2013 IE014.1 10 2013/50369
    AV-3543 AT the, e.g. in the virtual component 5. An example of this is a known torque estimator, which estimates the torque T of the connecting shaft 2 on the basis of the actual measured rotational speed n of the connecting shaft 2, or of the actuator 3 connected thereto. As a rule, the directly measured signal is also not used as the measured variable, but instead a correspondingly processed < filtered) signal.
    FIG. 2 shows a hardware-in-the-loop test environment for a drive train as a real component 4 as a further example. On the test bench 1 to the entire drive train is constructed. Here, it comprises an internal combustion engine 13, a clutch 14, a transmission 15 and a differential gear 16. The connecting shafts 2FL, 2fr, 2RLi 2rr are here formed by the half-shafts of the drive train and are provided with actuators 3Fl, 3Fr, 3rl, 3rr, e.g. in the form of electrical loading machines (dynamometers), connected. For the virtual component 5, the torques TFL1 TFR1 TRL, TRR of the connection waves 2fl, 2fr, 2r1, 2rr are detected here, and the virtual component 5 calculates the control variables for the real component 4 with the simulation model 21 implemented therein, in this case for the internal combustion engine 13 (eg Throttle position a), the clutch 14 (eg, a clutch signal K) and the transmission 15 (eg, a gear G), and the control variables for the actuators 3h., 3fr, 3rl, 3rR, here speeds iVl, Ofr, nRLl nRR. A test run in the hardware-in-the-loop test environment works exactly as described above with reference to FIG. Of course, other configurations for a real component 4 are conceivable, while the basic structure of the hardware-in-the-loop test environment and the performance of a test in the hardware-in-the-loop test environment remains unchanged.
    In the simulation model 21 of the virtual component 5, the measured variable M supplied by the sensor system 18 of the test bench 1, e.g. one (or more) torque T one (or 25 more) Halbwetle or a connecting shaft 2, one (or more) control variable S for the Aktuatorik 3 calculated. However, this torque T can also be a torque estimated only or calculated on the basis of the measured values of other sensors. For this purpose, in the simulation model 21 an equation of motion with at least one parameter P, e.g. in the form of a differential algebraic equation implemented several times a second, e.g. every millisecond, is solved. Of course, several equations of motion, e.g. a system of coupled equations of motion, be implemented. In general, S = f (P, M).
    In the simulation device 17, a Koirektureinheit 20 is provided (Figure 3), which consists of the corresponding processed measure M, e.g. a shaft torque Tw, preferably at each instant in which the equation of motion is solved (e.g., by numerically integrating the differential equations of motion), calculates a first correction value Ki
    AV-3543 AT net, which calms the virtual component 5 in a suitable manner and undesirable resonance phenomena, which arise due to the imperfection of the test bench 1 - in particular the actuator 3 and the sensor 18 - compensated as possible. The first correction value is therefore a continuously changing signal and is preferably added to the measured quantity M, in this case torque Tw, and the sum of the measured variable M and the correction value wird is the virtual component 5 as a corrected measured value M *, here a corrected shaft torque Tw *, for calculating the control quantity S for the actuator 3 fed.
    This "Torque Shaping" happens without interference with the overall structure of the test stand 1, 10 thus in particular without intervention in the controller structure of the test rig. 1
    The described inventive approach can be extended to any dynamic systems, wherein as measured variable M and control variable S instead of torque and speed then other physical quantities, such as. electrical voltage, mechanical force, etc., can be used. Alternatively or additionally, the correction unit 20 calculates from the calculated control quantity S, here e.g. Speed n, for the actuator 3, a second correction value K2l is added to the calculated control variable S and the sum as a corrected control variable S *, here a corrected speed n ', the actuator 3 for adjustment on the test bench 1 is provided. The second correction value "2" is preferably neither calculated at any time in which the equations of motion are solved, and again represents a signal which changes continuously. This corrected control variable S 'has the object of producing undesired vibration effects due to the imperfect transmission behavior of the signal Actuator 3 revealed to keep away from the real component 4 on the test bench 1.
    This "speed shaping" takes place again without intervention in the overall structure of the test stand 25 1, thus leaves in particular the controller structure of the test bench 1 unchanged.
    Alternatively or additionally, the correction unit 20 calculates from the measured variable M a third correction value K3 which serves to change a parameter P of the equation of motion in the virtual component 5, preferably to change a system inertia (eg the wheel inertia or vehicle inertia) or a mass (eg the vehicle mass). , This parameter P has the function of suppressing unwanted vibration effects of the virtual 5 and consequently also of the real component 4.
    This "parameter shaping" takes place again without intervention in the overall structure of the test rig 1, that is to say in particular without influencing the controller structure of the test rig 1.
    By way of example, possible methods for determining the correction values K1 (35 K2 and K3) are explained below.
    Prihtecl: 04-0 & S013 Μ
    102013/50369 AV-3543 AT
    In general, in the case of the methods exemplified below in the correction unit 20, a target function J is implemented as a function of the first or second or third correction value Κι, K2, K3 with respect to the first, second or third correction value
    Ki, K2, K3 optimized, here minimized, becomes, so in general notation J (KX 2 3) = min £ | 2. For the determination of the first correction value Ki, at least one simulation model 21 of a part of the vehicle connected via a shaft to the real component 4 is implemented in the virtual component 5, e.g. As in Figure 4, a wheel simulation model 10 of a vehicle wheel, which is connected to the half-wave of the drive train (real component 4 in Figure 2), or a simulation model of a dual-mass flywheel or a io coupling with the crankshaft of an internal combustion engine (real component 4 in Fig.1) is connected. The measured variable M is in each case the shaft torque Tw, and optionally further measured variables, such as. the temperature used, which is either measured directly or estimated from other measures or calculated.
    In the example of a wheel simulation model 10, as shown in Figure 4, from the gemes-15 senen wave moment Tw, the control variable S for the actuator 3 on the test bench 1, z. B. as here a speed ndmd, sjm for an electrical loading machine, determined. The wheel simulation model 10 can also be used with other simulation models, such as. a tire simulation model, a road simulation model, etc., be connected and exchange with these data. The shaft torque Tw is composed of a tire torque 20 Tire between the tire and the road, a braking torque T ^ e and further optional auxiliary torques Topt, such as, for example. an electric drive torque in a wheel hub motor.
    In the correction unit 20, an objective function J is implemented as a function of the first correction value K1 (here in the form of a correction torque Τ << g.). This objective function J is minimized with respect to the first correction value K1, that is to say in general notation The correction value Ki determined in this way is added to the measured variable M from the test bench 1, in this case the shaft torque Tw, and the corrected measured variable M *, here a corrected shaft torque T *, becomes the wheel simulation model 10 of the virtual component 5 for determining the control variable Pass S for the test bench 1.
    In the correction unit 20, a first target function Jenergy could be implemented in the form of a quadratic quality function for this purpose. For this purpose, for example, an objective function that uses the i nT zJw, energy
    rsi
    Back energy, e.g. in the form of the "effect of back energy" Jmer &. or the "effect of the acceleration energy", e.g. in the form -9- 30
    Printed: (W-0 & 2Ö13 E014.1
    102013/50369 AV-3543 AT 5 10 rated. The limits of integration are used to calculate the future over a period of time T in order to counteract future, expected conditions. The correction torque Tcor adds a torque that also changes the transmitted angular momentum. In order not to falsify the simulation too much, this angular momentum, which would cause a falsification of the speed, should be as small as possible over time. Therefore, J J T ^ x ^ dTdv will use a second objective function J ^ o, e.g. in the form Jdi. - which evaluates the angular momentum introduced by the correction torque T «*. The total target function J to be minimized for determining the first correction value Kt is then written as a linear combination of the first and second target functions with the weighting factors ai, oi2, J = aiJenergy + aristo. The sought correction torque Tcor is then obtained by minimizing this target function after the correction torque Tcor , Of course, further or other objective functions may also be taken into account for the overall objective function J, e.g. Both energy-constraint-based objective functions above could be considered. For the implementation of the minimization requires the correction unit 15 20 at least the control variable for the loading machine, here the speed η ^, ^ for He mediation of the angular velocity ω. The moment of inertia of the rotating part Jw (e.g., the wheel or the clutch) may be assumed to be known. The first correction value Ki, here the correction moment ΤΜΓ, can then be stored in the virtual component 5, e.g. in the wheel simulation model 10, as described above. However, it can also be provided that a, preferably the same, wheel simulation model 15 is also implemented in the correction unit 20. The correction unit 20 can then determine a corrected Gesamtradmoment T * with the detected correction torque ΤΜΓ with knowledge of the Welienmoments Tw and the virtual component 5 passed, as shown schematically in Figure 4. For the determination of the second correction value K2, an objective function J is implemented in the correction unit 20 as a function of the second correction value K2, in this case in the form of a correction rotational speed η "». This objective function J is minimized with respect to the second correction value K2, that is, in general notation J {K2) = min. As an objective function, e.g. a function K1 may be implemented which evaluates the deviation between the control variable S calculated in the virtual component 5 and the actual value of this control variable Sact that can be measured, e.g. in the form J = J | [ί5 * (τ) - (r) 1 ^ 7. Using the example of a rotational speed n 0 as the control variable S, the target function J can be written as J =) -n << i (T) | * i / r o -10-
    Printed: 04-06-2013 E014.1 10 2019/50369
    AV-3543 AT, where nact is the actual speed of the actuator 3, e.g. a loading machine, is. The objective function J is minimized after ndmdtSim and the result of this optimization is used as the correction speed η «* as described above.
    The determination of the third correction value K3 can be carried out analogously to the determination of the first correction value Ki. Again, an objective function J could be used as a linear function of two objective functions. With a first objective function, the effect of the back energy or the acceleration energy (Je ^ y) could again be evaluated as described above. The second objective function could e.g. the rotation energy (Je *. *,) changed by the changed parameter P, here the moment of inertia Jw, whereby the rotational energy changed by the changed parameter P should again be as small as possible over time, in order to distort the speed, the pulse or to minimize the kinetic energy of the system. Thus, J &lt; * s »o e.g. in the form t V
    Jdlslo = j or Jdis [0 - J | ./^ (χ, ν) ω (τ, ν) ί / τί / ν written v = 0r = (l i> = Or = 0. The objective function J can then be the third correction value K3, here the correction moment of inertia Joor, are optimized, with which the moment of inertia Jw in the virtual component 5, or in the equation of motion in the simulation model of the virtual components 5, is corrected, for example, added to Jw with the correct sign becomes.
    A particular advantage in the optimization of objective functions for determining the correction values Ki, K2, K3 can be seen in the fact that optimization can very easily take into account boundary conditions, which makes it possible to comply with predetermined limitations of the virtual component 5, e.g. a maximum wheel speed, or real component 4, e.g. a maximum torque of an internal combustion engine, or the actuator 3, e.g. a maximum spin of an electric loading machine, can be considered. 25 e.g. For determining the first and third correction values Ki, K3, the following boundary conditions could be taken into account: (f <x <f + T) I® (τ) ΗώΗ ^ (^ x <f + 7y with limit values for the rotational speed and the rotational acceleration, For the determination of the second correction value K2, similar boundary conditions could be taken into account, eg -11-
    Printed: 04-06-2013 teQ141 10 2013/50369
    AV-3543 AT Μτ) -η ™ (t ^ <t + T) ΜΦή ™ (t <x <t + T) 'whereby limit values for the rotational speed and the rotational acceleration can be specified. For the optimization of the above objective functions J, there are well-known solution methods, e.g. dynamic programming, receeding horizon optimization, etc., which are not explained here. The objective function J is preferably optimized in real time, preferably on a real-time computer.
    12-
    权利要求:
    Claims (14)
    [1]
    Printed: 04-06-2013 10 2013/50369 AV-3543 AT Claims 1. Method for reducing the excitation of oscillations and resonances in a test stand (1) for a real component (4) and a virtual component (5), the real component ( 4) supplies a measured variable (M) of the real component (4) to the virtual component (5) and receives from the virtual component (5) a control variable <S) for an actuator (3) of the test bench (1), wherein in the virtual component ( 5) a simulation model (21) is implemented with an equation of motion that calculates the control variable (S) from the measured variable (M), characterized in that at least one of the following method steps a), b) or c) is set, a ) Determining a first correction value (Ki) from the measured variable (M), wherein the first correction value (KO to the measured variable (M) is added and the sum as a corrected measured variable (M *) of the virtual component (5) for calculating the control variable (S) is transmitted b) Determining a second correction value (K2) from the calculated control variable (S), wherein the second correction value (K2) is added to the calculated control variable (S) and the sum is transmitted as a corrected control variable (S *) to the actuator system (3), c) Determining a third correction value (K3) from the measured variable (M), wherein a parameter (P) of the equation of motion is changed with the third correction value (K3).
    [2]
    2. The method according to claim 1, characterized in that as a measured variable (M), the torque (Tw) of a connecting shaft (2) between real component (4) and actuator (3) is used.
    [3]
    3. The method according to claim 1 or 2, characterized in that as a control variable (S) a speed (n) is used.
    [4]
    4. The method according to any one of claims 1 to 3, characterized in that for determining the first or second or third correction value (Κι, K2l K3) a target function (J) as a function of the first or second or third correction value (Κι, K2, K3 ) is implemented, which is optimized with respect to the first or second or third correction value (Κι, K2, K3).
    [5]
    5. The method according to claim 4, characterized in that for determining the first or third correction value (K1, K3) as a target function (J) a linear combination of a first and second objective function (J «n« rgy, Jdi * u>) is used.
    [6]
    A method according to claim 5, characterized in that as the first objective function (Jenergy) a quadratic quality functional as a function of the angular velocity (ω) or -13- pmmMä & mtöou.i 102013/50369 AV-3543 AT 1 i + i of a derivative thereof, preferably in the form = -Jw J | tS (r) g </ T 2 t = t or J "* gr = jw τ = /
    [7]
    7. The method as claimed in claim 5, characterized in that the second target function (Jdisto) evaluates the angular momentum T r (rv) additionally introduced by the first correction value KO, preferably in the form Jdlsll, = II - · - dxdv v = 0 τ = 0 * Λί
    [8]
    8. Method according to claim 5, characterized in that the second target function (Jdisto) evaluates the kinetic energy changed by the third correction value (KO) f V, preferably in the form = j jj ^ (τ, ν)! { »(Τ, ν ^ ο / τάν or i = |> T = 0 / v ddlslu = J | J cor (τ> v) ω (τ, v) didv. Ν = Οτ = 0
    [9]
    9. The method according to claim 4, characterized in that as a first correction value (KO a Korrektunmoment (Τ «*) is determined.
    [10]
    10. The method according to claim 4, characterized in that as the third correction value (K3) a correction moment of validity (Jo *) is determined.
    [11]
    11. The method according to any one of claims 1 to 3, characterized in that for determining the second correction value (K2), an objective function (J) is implemented, the deviation between the calculated in the virtual component control variable (S) and the actual value of this control variable (Sact), preferably in the form.; = J | i '(r) -s "(T) |' dT. 0
    [12]
    12. The method according to claim 11, characterized in that as second correction value (K2) a correction speed (η «*) is calculated, preferably from J = j (T) _ now (T) | 2 ^ T. by optimizing for the speed (ndmd.sim) and using the detected speed (ndmd, sim) as the correction speed (n »r).
    [13]
    13. The method according to claim 2 to 12, characterized in that boundary conditions for the consideration of predetermined limitations of Virtu-alkomponent (5) or the real component (4) or the actuator (3) are used in the optimization.
    [14]
    -14-
    类似技术:
    公开号 | 公开日 | 专利标题
    DE112014002661B4|2021-12-30|Method for reducing vibrations in a test bench
    DE10200695B4|2004-09-09|Test system and method for a motor vehicle component using a dynamometer
    DE19910967C1|2000-09-21|Method for simulating the behavior of a vehicle on a road
    AT520179B1|2019-02-15|Test bench and method for carrying out a test
    DE102007016420B4|2011-04-21|Test bench and procedure for testing a drive train
    EP3224589B1|2020-04-22|Method and a device for determining the propulsion torque
    WO2008040282A1|2008-04-10|Drive train
    EP3067681A1|2016-09-14|Method for operating an engine or power train test stand
    EP1382956B1|2006-01-04|Procedure for simulating the driving performance of vehicles
    DE102017105469A1|2017-10-05|Hardware-in-the-loop system and method for the control of an internal combustion engine
    AT510041B1|2012-01-15|METHOD AND DEVICE FOR SIMULATING A TRANSLATORALLY OR ROTATIVELY MOVING BODY
    DE102016125538A1|2018-03-01|Method for verifying actuator control data
    AT507938B1|2011-04-15|METHOD FOR VERIFYING DRIVE TRAY SYSTEMS
    DE102016103643B4|2018-05-03|Method and device for checking a software of a control unit of a vehicle
    DE102015111409B3|2016-11-03|Method for determining forces and / or torques
    AT522635B1|2022-02-15|Method and control device for operating a test stand
    AT520537A4|2019-05-15|Method for operating a test bench
    DE102010047413A1|2011-06-30|Method for direct determination of power loss of e.g. vehicle transmission, involves determining power delivered by electric machines, and determining power loss of transmission from difference between input and output power
    AT519553A4|2018-08-15|Method for controlling, in particular regulating, a powertrain test bench with a real gearbox
    DE102004039738A1|2006-02-23|Dynamic gear state change detecting method for use in motor vehicle, involves calculating state variables of gear from speed signals for diagnostics purpose, and evaluating measurement signals to simulate physical system performance of gear
    DE102012202295A1|2013-08-22|Method for adjustment of application data of control device e.g. combustion engine control unit installed in e.g. motor car, involves simulating running conditions of motor car by influencing powertrain
    DE102015007632A1|2016-12-15|Method and system for inspecting a motor vehicle subsystem
    DE112007003032B4|2019-12-12|powertrain
    DE102010019301A1|2010-11-11|Torque determining method for drive train of e.g. lorry in mechanic workshop, involves calculating torque for operating state and normal operation with equation of torque equal to addition of product of gain and parameter with offset value
    AT521277B1|2019-12-15|Determination of the height of a vehicle and a road gradient from the air pressure
    同族专利:
    公开号 | 公开日
    JP6442490B2|2018-12-19|
    AT512483B1|2015-02-15|
    JP2016520841A|2016-07-14|
    DE112014002661A5|2016-03-24|
    CN105452833A|2016-03-30|
    US10317312B2|2019-06-11|
    WO2014195238A1|2014-12-11|
    US20160116367A1|2016-04-28|
    DE112014002661B4|2021-12-30|
    AT512483A3|2014-10-15|
    CN105452833B|2019-05-07|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    WO2015103658A1|2014-01-09|2015-07-16|Kristl, Seibt & Co. Gesellschaft M.B.H.|Method and device for controlling a powertrain test stand|
    EP3067681B1|2015-03-10|2018-02-14|IPG Automotive GmbH|Method for operating an engine or power train test stand|
    AT519553A4|2017-04-07|2018-08-15|Avl List Gmbh|Method for controlling, in particular regulating, a powertrain test bench with a real gearbox|US280948A|1883-07-10|Pbecy c |
    BE786393A|1971-07-19|1973-01-18|Caterpillar Tractor Co|MACHINE FOR TESTING SHOCK ABSORBERS|
    DE2217702C2|1972-04-13|1974-04-18|Carl Schenck Maschinenfabrik Gmbh, 6100 Darmstadt|Control arrangement for a test system for testing a test item with movable elements, in particular for testing motor vehicle axles|
    US4211105A|1976-04-27|1980-07-08|Magyar Vagon-Es Gepgyar|Method for measuring the damping coefficient of torsional-vibration dampers|
    US4395904A|1979-05-18|1983-08-02|Ivanov Gely M|Device for damping oscillations|
    DE3861392D1|1987-02-25|1991-02-07|Siemens Ag|ELECTRONIC OBSERVER FOR A TORQUE GENERATOR COUPLED TO A LOADING MACHINE, AND METHOD FOR DETERMINING THE TORQUE AND FOR TESTING THE TORQUE GENERATOR.|
    DE19540753C1|1995-11-02|1997-01-23|Ebern Fahrzeugtech Gmbh|Method of suppressing vibrations, especially on hydraulic clutch pedal of motor vehicle|
    DE19910967C1|1999-03-12|2000-09-21|Avl Deutschland Gmbh|Method for simulating the behavior of a vehicle on a road|
    US6637271B1|2001-12-04|2003-10-28|Automotive Products Inc.|Method and apparatus for identifying and eliminating vibration symptoms in hydraulic clutch actuators|
    AT500978B8|2003-05-13|2007-02-15|Avl List Gmbh|METHOD FOR OPTIMIZING VEHICLES|
    JP2005212706A|2004-02-02|2005-08-11|Nsk Ltd|Electric power steering evaluation system, method, and program|
    DE102005022096A1|2005-05-12|2006-11-16|Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.|Device and method for vibroacoustic examination of a motor vehicle|
    US20070260373A1|2006-05-08|2007-11-08|Langer William J|Dynamic vehicle durability testing and simulation|
    JP2008094197A|2006-10-10|2008-04-24|Nsk Ltd|Development assist system and control program of electric power steering device, and control program loading method|
    US8135556B2|2008-10-02|2012-03-13|Mts Systems Corporation|Methods and systems for off-line control for simulation of coupled hybrid dynamic systems|
    JP5727315B2|2011-07-05|2015-06-03|株式会社小野測器|Transient running simulation apparatus and method|
    CN202648904U|2012-05-25|2013-01-02|上海第二工业大学|Vehicle vibration test analysis system based on virtual instrument|JP6369596B1|2017-05-09|2018-08-08|株式会社明電舎|Control device for dynamometer system|
    AT522353B1|2019-08-05|2020-10-15|Avl List Gmbh|Test bench and method for performing a test run on a test bench|
    AT522354B1|2019-08-12|2020-10-15|Avl List Gmbh|Procedure for operating a test bench|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    ATA50369/2013A|AT512483B1|2013-06-03|2013-06-03|Method for reducing vibrations in a test bench|ATA50369/2013A| AT512483B1|2013-06-03|2013-06-03|Method for reducing vibrations in a test bench|
    CN201480039078.2A| CN105452833B|2013-06-03|2014-05-30|Method for reducing vibration in testing stand|
    US14/895,380| US10317312B2|2013-06-03|2014-05-30|Method for reducing vibrations in a test bed|
    PCT/EP2014/061262| WO2014195238A1|2013-06-03|2014-05-30|Method for reducing vibrations in a test bench|
    DE112014002661.1T| DE112014002661B4|2013-06-03|2014-05-30|Method for reducing vibrations in a test bench|
    JP2016517251A| JP6442490B2|2013-06-03|2014-05-30|Method for reducing vibration in a test bench|
    [返回顶部]