Method for operating a steer-by-wire steering system and steer-by-wire steering system
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- VOLKSWAGEN AG
- Filing Date
- 2024-04-24
- Publication Date
- 2026-04-22
AI Technical Summary
Steer-by-wire (SbW) steering systems require higher reliability than conventional systems due to the absence of a mechanical connection, necessitating robust error detection and redundancy to prevent complete system failure, especially with the integration of automated driving assistance systems.
The method involves a transition from a main control path to a redundant control path in a SbW steering system, where the redundant path is activated based on driving dynamics and steering gear measurements to estimate and set a manipulated variable, such as torque, during a transition phase, thereby reducing torque gaps and ensuring a smooth switch.
This approach enhances the reliability and comfort of the transition by providing immediate manipulated variables, avoiding extreme values, and gradually aligning the redundant control path with the main path's variables, ensuring continued system functionality and driver experience.
Smart Images
Figure EP2024061303_26122024_PF_FP_ABST
Abstract
Description
[0001]Description Method for operating a steer-by-wire steering system and steer-by-wire steering system. The invention relates to a method for operating a steer-by-wire steering system and a steer-by-wire steering system. Steer-by-wire (SbW) steering systems must have significantly higher reliability than conventional electromechanical steering systems. This is due to the fact that SbW steering systems have no mechanical connection between the steering handle and the steering gear. A complete failure of the steering system would render the vehicle unsteerable. Therefore, a complete failure must be prevented and fault effects detected early. Furthermore, automated driving assistance systems also require increased safety in the vehicle components. An SbW steering system comprises a steering wheel module with a force feedback actuator (FFA).which detects a desired steering angle at the steering handle and generates a realistic feedback torque (hand torque) for the driver. Furthermore, the SbW steering system includes a steering module with a steering gear (road wheel actuator, RWA), which converts the driver's input into a wheel steering movement. To increase the fault tolerance of the overall system, the control of an actuator, which is responsible for position control in the steering gear (RWA) or torque control in the force feedback actuator (FFA), is typically implemented redundantly. In particular, the actuator has at least two independent paths for controlling or regulating the actuator. A heterogeneous implementation of these paths can increase fault tolerance against common-cause errors. If a main control path fails,A transition to a redundant control path must take place, and the main control path is deactivated. For this purpose, the redundant control path is transferred from a passive state to an active state. Due to the preceding passive operation of the redundant control path and the resulting necessary activation, a control variable difference can occur. The invention is based on the object of creating a method for operating a steer-by-wire steering system and a steer-by-wire steering system in which a transition between a main control path and a redundant control path is improved. This object is achieved according to the invention by a method having the features of patent claim 1 and a steer-by-wire steering system having the features of patent claim 9. Advantageous embodiments of the invention emerge from the subclaims. In particular, a method for operating a steer-by-wire steering system is provided,The steer-by-wire steering system has redundantly designed control paths. Upon the occurrence of a fault in a main control path, the system switches from the main control path to a redundant control path, and the redundant control path is switched from a passive state to an active state. A control variable in the redundant control path is estimated during a predetermined transition phase based on measured variables of a vehicle dynamics and a steering gear. Furthermore, a steer-by-wire steering system is provided, comprising a steering wheel module configured to detect a steering input at a steering handle and to generate a feedback torque at the steering handle, a steering module configured to set a steering angle on at least one steerable wheel based on the detected steering input and to determine the feedback torque, and a communication connection between the steering wheel module and the steering module.wherein the steering wheel module and / or the steering module have redundantly designed control paths, wherein a control device of the steering wheel module and / or a control device of the steering module is configured to switch from the main control path to a redundant control path upon the occurrence of a fault in a main control path and to switch the redundant control path from a passive state to an active state for this purpose, and to estimate a manipulated variable in the redundant control path during a predetermined transition phase based on measured variables of a driving dynamics and a steering gear. The method and the steer-by-wire steering system enable an improved transition between the main control path and a redundant control path when a fault occurs in the main control path. In particular, a torque gap that may arisebecause the redundant control path is activated from a passive state and therefore cannot initially provide a control variable corresponding to the previous control variable of the main control path (i.e., in particular, it has not yet reached steady state), must be handled and, in particular, reduced during the transition phase. For this purpose, it is provided that a control variable in the redundant control path is estimated during the specified transition phase based on measured variables of a driving dynamics and a steering gear. The estimated control variable, in particular a torque, is then set as a control variable in the redundant control path. This makes it possible to provide a control variable immediately after the activation of the redundant control path. Furthermore, the transition can be made more comfortable for a driver, since, in particular, the two extreme cases, i.e., a maximum control variable, in particular a maximum torque, and a control variable of zero,in particular, a torque of zero, can be avoided. The manipulated variable is, in particular, a torque. In principle, however, the manipulated variable can also be another variable, for example, a force or a voltage. The controlled variable is, in particular, an angle or a position. The duration of the specified transition phase depends, in particular, on a step response of the respective (redundant) controller. The step response is a measure of how long the controller needs for a transient response to decay after activation. A duration is selected accordingly (for a settling time of 1 s, for example, also 1 s). Once the duration of the specified transition phase has elapsed, the manipulated variable is no longer estimated but specified by the activated redundant controller. Parts of the steer-by-wire steering system, in particular the control device(s),can be implemented individually or collectively as a combination of hardware and software, for example as program code executed on a microcontroller or microprocessor. However, it can also be provided that parts are implemented individually or collectively as an application-specific integrated circuit (ASIC) and / or field-programmable gate array (FPGA). In one embodiment, it is provided that for the duration of the transition phase, a manipulated variable limitation in the redundant control path is set to the estimated manipulated variable. This allows the manipulated variable to be gradually adjusted to the manipulated variable present in the main control path when the error occurs. In one embodiment,that the estimation and setting of the manipulated variable is carried out repeatedly during the transition phase. This allows an updated manipulated variable to be repeatedly estimated and provided. In particular, the manipulated variable limitation can also be gradually reset and adjusted in this way. A torque gap can thereby be gradually closed and the redundant control path can thereby, in particular, gradually approach the value of the manipulated variable present at the time of the error occurrence. In one embodiment, it is provided that a tire lateral force is calculated based on measured variables of the driving dynamics and a linear single-track model for a vehicle, wherein the estimation of the manipulated variable is carried out based on the calculated tire lateral force. The manipulated variable is in particular a torque in the steering module. The idea behind this is,that two forces act on a steering gear in particular: a tire lateral force and a torque of an electric motor used to position the at least one steerable wheel. The tire lateral force is calculated; the torque is estimated and used as a control variable. In a further embodiment, it is provided that the tire lateral force is reduced by means of a lateral acceleration-dependent factor. This can prevent a value for the calculated tire lateral force from becoming too large. It has been shown that the linear single-track model above a lateral acceleration of approximately 4.5 m / s, 2delivers values that are too large because the model assumptions no longer apply there. As a result, the manipulated variable estimated on this basis, in particular an estimated torque, would also be too large. To prevent this, the calculated tire lateral force is reduced using the lateral acceleration-dependent factor. The factor can be determined, for example, on the basis of a characteristic curve stored in the control device, for example, which includes a dependency of the factor on the lateral acceleration. The values and a course of the characteristic curve can be determined, for example, based on more complex vehicle models. In particular, the factor ensures that the manipulated variable, in particular the torque, is attenuated. In one embodiment, provision is made for the manipulated variable to be estimated using an Unknown Input Observer (UIO). This allows particularly good results to be achieved.In particular, it is assumed that an estimated state vector of the respective part of the steer-by-wire steering system under consideration (steering wheel module or steering module) used in the Unknown Input Observer (UIO) converges to a real state vector. In particular, the state vector is estimated using the UIO; based on the estimated state vector, the manipulated variable, in particular the torque, is then calculated. In a further embodiment, it is provided that a steering target input detected at a steering handle of the steer-by-wire steering system is fed as an input to the Unknown Input Observer (UIO). In one embodiment, it is provided that the method is used in a steering wheel module and in a steering module of the steer-by-wire steering system. This can prevent the occurrence of a deviation in the manipulated variables, in particular a torque gap, in both the steering wheel module and the steering module.Further features for the design of the steer-by-wire steering system will become apparent from the description of embodiments of the method. The advantages of the steer-by-wire steering system are the same in each case as in the embodiments of the method. Furthermore, in particular, a vehicle is also created, comprising a steer-by-wire steering system according to one of the described embodiments. The vehicle is in particular a motor vehicle. The invention is explained in more detail below using preferred exemplary embodiments with reference to the figures. Herein: Fig. 1 shows a schematic representation of an embodiment of the steer-by-wire steering system; Figs. 2a, 2b show schematic representations to illustrate the estimation of the manipulated variable. Fig. 1 shows a schematic representation of an embodiment of the steer-by-wire steering system 1.The steer-by-wire steering system 1 comprises a steering wheel module 2, a steering module 3, and a communication connection 4 between the steering wheel module 2 and the steering module 3. The steer-by-wire steering system 1 is arranged in particular in a vehicle 50. The steer-by-wire steering system 1 is configured to carry out the method described in this disclosure. The method is described in more detail below with reference to the steer-by-wire steering system 1. The steering wheel module 2 is configured to detect a steering input at a steering handle 51 and to generate a feedback torque at the steering handle 51. The steering input is detected by a sensor 5 of the steering wheel module 2. The feedback torque is applied to the steering handle 51 by an electric motor 6 of the steering wheel module 2. The steering module 3 is configured to set a steering angle on at least one steerable wheel 52 based on the detected steering input and to determine the feedback torque.The required rack position of a rack 53 is detected directly or indirectly by a sensor 7 of the steering module 3. The steering command is applied to the rack 53 and thereby to the wheels 52 by an electric motor 8 of the steering module 3. This is done via a steering gear 11. The steering command and the feedback torque are each transmitted via the communication link 4. The steering module 2 further comprises a control device 9, which provides two controllers 9-1, 9-2, which enable redundant control of the feedback torque. The two controllers 9-1, 9-2 are designed differently, in particular, to avoid common-cause errors (heterogeneous design). The controllers 9-1, 9-2 can both provide a manipulated variable 20, which is fed to the electric motor 6 for implementation.During normal operation, in particular, only controller 9-1 is used as the main control path, while controller 9-2 is passively connected as a redundant control path and is only activated in the event of a fault. The steering module 3 further comprises a control device 10, which provides two controllers 10-1, 10-2 that enable redundant control of the rack position. The two controllers 10-1, 10-2 are designed differently in particular to avoid common-cause errors. The controllers 10-1, 10-2 can both provide a manipulated variable 21, which is fed to the electric motor 8 for implementation. During normal operation, in particular, only controller 10-1 is used as the main control path, while controller 10-2 is passively connected as a redundant control path and is only activated in the event of a fault.The control devices 9, 10 are each configured to switch from the main control path to the redundant control path upon the occurrence of an error in the main control path and to switch the redundant control path from a passive state to an active state for this purpose. The control device 9 switches the controller 9-2 from the passive to the active state and deactivates the controller 9-1. The control device 10 switches the controller 10-2 from the passive to the active state and deactivates the controller 10-1. Depending on where the error occurs, this occurs only in the steering wheel module 2 or only in the steering module 3 or in both modules 2, 3. The control devices 9, 10 are further configured to estimate the respective manipulated variable 20, 21 in the redundant control path during a predetermined transition phase based on measured variables 30 of a driving dynamics and the steering gear 11.The measured variables 30 of the driving dynamics are queried and / or provided, for example, by a vehicle control system. After a predetermined duration of the transition phase has elapsed, the respective manipulated variable 20, 21 is used for control by the redundant control path, i.e., by the controllers 9-2, 10-2. It can be provided that, for the duration of the transition phase, a manipulated variable limitation (i.e., a maximum value for the manipulated variable) in the redundant control path is set to the estimated manipulated variable 20, 21. It can be provided that the estimation and setting of the manipulated variable 20, 21 is carried out repeatedly during the transition phase. In particular, it can be provided that the estimation and setting are repeated at regular time intervals. For example, this can occur 10, 100, 1000, ... times during the duration of the transition phase.It can be provided that a tire lateral force is calculated based on measured variables of the driving dynamics and a linear single-track model for a vehicle, with the control variable being estimated based on the calculated tire lateral force. In a further development, it can be provided that the tire lateral force is reduced using a lateral acceleration-dependent factor. The factor can be determined, for example, based on a characteristic curve stored in the control devices 9, 10. It can be provided that the control variable is estimated using an Unknown Input Observer (UIO). This embodiment is schematically illustrated in Figures 2a and 2b and described in more detail below. The steer-by-wire steering system 1 is modeled on the steering module 3 side using a model that is schematically shown in Figure 2a. The model includes the vehicle 50, the steering gear 11, and the electric motor 8.In the model, two quantities act on the steering gear 11: the tire lateral force ^^. ௬௩ and the torque ^^ ெ^௧ of the electric motor 8. This torque ^^ ெ^௧ is not known as a manipulated variable when activating the redundant control path and must be estimated (the estimated torque ^^ ெ^௧ is then set as a control variable in the redundant control path according to the procedure). Furthermore, the following influencing variables act on the vehicle 50: the rack position ^^ ^^௧ and the vehicle speed ^^. From the measured variables of the steering gear 11 (in the example, in particular the rack position ^^ ^^௧ ) and the driving dynamics (that is, in the example, the vehicle speed ^^) the torque required for the vehicle position ^^ ெ^௧ The non-measurable variables of the steering gear 11 are estimated based on models. The tire lateral force ^^ ௬௩is calculated from the measured variables of the driving dynamics and the linear single-track model. Depending on the lateral acceleration, the tire lateral force ^^ ௬௩in particular reduced, as already described above. The basis for the estimation is a state space representation of the controlled system: ^^^ ൌ ^^ ∙ ^^ ^ ^^ ∙ ^^ ^^ ൌ ^^ ∙ ^^ ^ ^^ ∙ ^^ where ^^ is the state vector (e.g. rotor position of the electric motor 8, rotor speed, rack position, rack speed), ^^ is the system matrix, ^^ is the input matrix and ^^ is the input variable. ^^ is the output matrix and ^^ is the throughput matrix. The state space representation represents the overall system as a dynamic model (cf. Fig. 2a). In the example, the overall system consists of the vehicle 50 and the steering gear 11. The vehicle 50 is modeled as a linear single-track model, the steering gear 11, for example, as a dual-mass oscillator. The state space representation can be used to simulate the dynamic behavior of the overall system. Fig.2b illustrates how the estimation is performed.Shown are the controlled system 40 and the unknown input observer 41. The state space is: ^^^ ൌ ^^ ∙ ^^ ^ ^^ ∙ ^^ ^ ^^ ∙ ^^ where ^^ is the non-measurable or unknown input, in this case the torque ^^. ெ^௧ , which is to be estimated. Furthermore, the state space is: ^^ ൌ ^^ ∙ ^^ ^ ^^^ ൌ ^^ ∙ ^^^ ൌ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^ For the Unknown Input Observer (UIO), the following also applies: ^ ^ ^^ ൌ ^ ^ ^ ∙ ^^̂ ^ ^ ^ ^ ∙ ^^ ^ ^^ ∙ ^^Here correspond ^ ^ ^ und ^ ^ ^ In particular, they are no longer defined with clearly interpretable physical quantities, but are specified or selected within the framework of the UIO design. With ^^̂ ൌ ^^^ ^ ^^ ∙ ^^ we get: ^ ^ ^^ ൌ ^ ^^ ∙ ^^^ ^ ^^ ∙ ^^ ^ ^^ ∙ ^^ െ ^^ ∙ ^^^^^ and ^^ can be referred to as feedback matrices (analogous to a Luenberger observer). ^^^ is the estimated state vector. With ^^^ ൌ ^^ ∙ ^^^ ൌ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ we get: ^ ^ ^^ ൌ ^ ^ ^ ∙ ^^^ ^ ^ ^ ^ ∙ ^^ ^ ^^ ∙ ^^ െ ^^ ∙ ^^ ∙ ^^ ∙ ^^ െ ^^ ∙ ^^ ∙ ^^ ∙ ^^ െ ^^ ∙ ^^ ∙ ^^ ∙ ^^ ∙ ^^The deviation (estimation error) ^^^ ൌ ^^^ െ ^ ^ ^^ corresponds to:^^^ ൌ ^^ ∙ ^^ ^ ^^ ∙ ^^ ^ ^^ ∙ ^^ െ ^ ^ ^ ∙ ^^^ െ ^ ^ ^ ∙ ^^ െ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^^^^ ൌ ^^ ∙ ^^ ^ ^^ ∙ ^^ ^ ^^ ∙ ^^ ^ ^^ ∙ ^^ െ ^ ^ ^ ∙ ^^^ െ ^ ^ ^ ∙ ^^ െ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^ ^ ^^ ∙ ^^ ∙ ^^ ∙ ^^^^^ ൌ ^^ ∙ ^ ^^ ^ ^^ ∙ ^^ ∙ ^^ െ ^^ ∙ ^^^ െ ^ ^ ^ ∙ ^^^ ^ ^^ ∙ ^ ^^ ^ ^^ ∙ ^^ ∙ ^^^ ^ ^^ ∙ ^ ^^ െ ^ ^^ ^ ^^ ∙ ^^ ∙ ^^^The estimation error should be independent of the unknown input ^^ (that is, in particular of the torque to be estimated ^^ ெ^௧ ) such that: ^^ ^ ^^ ∙ ^^ ∙ ^^ ൌ 0 and thus ^^ ൌ െ ^^ ∙ ^ ^^ ∙ ^^ ^ି^ The estimation error should continue to be independent of the measured variable or input variable ^^ (that is, in particular of the tire lateral force ^^ ௬௩ ) so that: ^ ^ െ ^ ^ ^ ^ ^^ ∙ ^^ ∙ ^^ ൌ 0 and thus ^ ^ ^ ൌ ^^ ^ ^^ ∙ ^^ ∙ ^^What remains is a term that is independent of the known quantity ^^ and the unknown quantity ^^: ^^^ ൌ ^^ ∙ ^ ^^ ^ ^^ ∙ ^^ ∙ ^^ െ ^^ ∙ ^^^ െ ^ ^ ^ ∙ ^^^ The Matrix ^ ^ ^ is chosen such that the estimation error depends only on this matrix: ^ ^ ^ ൌ ^^ ^ ^^ ∙ ^^ ∙ ^^ െ ^^ ∙ ^^The estimation error finally results in: ^ ^^ ൌ ^ ^ ^ ∙ ^^ So if the eigenvalues of ^ ^^ have negative real parts, the estimation error converges to 0. The eigenvalues can be freely chosen if the Kalman criterion for complete observability is met. The unknown input ^^ of the UIO, i.e., the motor torque ^^ ெ^௧, is calculated in such a way that the estimation error ^^^ is minimized. The estimation error here exists between the measured state variables (expressed in the state vector ^^, which includes, for example, a rack position, a rack speed, a rotor position and / or a rotor speed) and modeled state variables of the steering gear 11. In particular, the assumption is made that an actual value of the measured state variables corresponds to a steering target value (e.g., an actual position of the rack is equated with a target position, as this results from a steering target value of the driver on the steering handle), i.e., the assumption is made that the target value is converted into the actual value immediately and without delay. In other words: If the estimation error ^^ ൌ ^^ െ ^^^ disappears, the unknown input ^^ and thus the motor torque ^^ ெ^௧be calculated (estimated) using the above equations. In particular, it is provided that a steering target input detected at a steering handle 51 of the steer-by-wire steering system 1 is fed as input to the Unknown Input Observer (UIO) (here in the form of the state vector ^^). The estimated torque, i.e., the engine torque ^^ ெ^௧, thereby contains the model inaccuracies and disturbances, the torque gap can still be reduced or gradually closed. The torque that would be necessary for the model-based state variables of the steering gear 11 to correspond to the actual state variables is estimated. The result is a model-based torque with which a drop in torque after activation of the redundant control path during the transition phase can be mitigated. The example was described using the steering module 3. In the same way, the estimation can also be carried out for a torque control for the feedback torque (manual torque) in the steering wheel module 2. If a different variable is to be set as the manipulated variable, the estimation is carried out in a fundamentally analogous manner. The embodiment shown in Fig. 1 provides for the method to be used in a steering wheel module 2 and in a steering module 3 of the steer-by-wire steering system 1.In principle, however, it can also be provided that the method is only used in the steering wheel module 2 or only in the steering module 3. List of reference symbols 1 Steer-by-wire steering system 2 Steering wheel module 3 Steering module 4 Communication connection 5 Sensor 6 Electric motor 7 Sensor 8 Electric motor 9 Control device 9-1 Controller (main path) 9-2 Controller (redundant path) 10 Control device 10-1 Controller (main path) 10-2 Controller (redundant path) 11 Steering gear 20 Manipulated variable (steering wheel module) 21 Manipulated variable (steering module) 40 Controlled system 41 Unknown Input Observer (UIO) 50 Vehicle 51 Steering handle 52 Steerable wheel 53 Rack ^^ System matrix ^^ Input matrix ^^ Output matrix ^^ Through matrix ^^. ௬௩ Tire lateral force ^^ ெ^௧ Torque ^^ Input variable ^^ Vehicle speed ^^ Non-measurable / unknown input (estimated) ^^ State vector (target) ^^ ^^௧ Rack position ^^ State vector (actual)
Claims
Claims 1. Method for operating a steer-by-wire steering system (1), wherein the steer-by-wire steering system (1) has redundantly designed control paths, wherein upon the occurrence of an error in a main control path, switching is carried out from the main control path to a redundant control path, and the redundant control path is switched from a passive state to an active state for this purpose, wherein a manipulated variable (20, 21) in the redundant control path is estimated during a predetermined transition phase based on measured variables (30) of a driving dynamics and a steering gear (11).
2. Method according to claim 1, characterized in that for the duration of the transition phase, a manipulated variable restriction in the redundant control path is set to the estimated manipulated variable (20, 21).
3. Method according to claim 1 or 2, characterized in that the estimation and setting of the manipulated variable (20, 21) is carried out repeatedly during the transition phase.Method according to one of the preceding claims, characterized in that a tire lateral force ( ^^. ௬௩ ) is calculated based on measured variables (30) of the driving dynamics and a linear single-track model for a vehicle (50), wherein the estimation of the manipulated variable (20,21) is based on the calculated tire lateral force ( ^^ ௬௩ ) is carried out.
5. Method according to claim 4, characterized in that the tire lateral force ( ^^ ௬௩ ) is reduced by means of a lateral acceleration-dependent factor.
6. Method according to one of the preceding claims, characterized in that the manipulated variable (20, 21) is estimated by means of an unknown input observer (41).
7. Method according to claim 6, characterized in that a steering target input detected at a steering handle (51) of the steer-by-wire steering system (1) is fed as an input to the unknown input observer (41).
8. Method according to one of the preceding claims, characterized in that the method is used in a steering wheel module (2) and in a steering module (3) of the steer-by-wire steering system (1).
9. Steer-by-wire steering system (1), comprising: a steering wheel module (2) configured to detect a steering input at a steering handle (51) and to generate a feedback torque at the steering handle (51), a steering module (3) configured to set a steering angle on at least one steerable wheel (63) based on the detected steering input and to determine the feedback torque, and a communication connection (4) between the steering wheel module (2) and the steering module (3), wherein the steering wheel module (2) and / or the steering module (3) have redundantly designed control paths, wherein a control device (9) of the steering wheel module (2) and / or a control device (10) of the steering module (3) is configured toUpon the occurrence of a fault in a main control path, switching from the main control path to a redundant control path and, for this purpose, switching the redundant control path from a passive state to an active state, and estimating a manipulated variable (20, 21) in the redundant control path during a predetermined transition phase based on measured variables (30) of a driving dynamics and a steering gear (11).
10. Vehicle (50) comprising a steer-by-wire steering system (1) according to claim 9.