A self-adapting calibration method and system for road feedback of steer-by-wire
Patent Information
- Application Number
- CN202610507827.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-17
- Publication Date
- 2026-08-21
- Estimated Expiration
- 2046-04-17
AI Technical Summary
1、个体差异显著:不同批次电机、减速器齿侧间隙、轴承预紧、润滑与装配偏差会导致摩擦参数离散,表现为同车型不同车辆手感不一致
[0027] In this invention, the online update of the mechanical friction state variables relies solely on the steering wheel output resistance torque and the torque command of the previous cycle, decoupling it from the ideal road surface feedback resistance torque and ensuring the independence of friction estimation. Furthermore, feedforward compensation is performed using the predicted mechanical friction value of the next cycle, combined with feedback correction to generate torque commands, achieving adaptive compensation for friction in the execution link. By storing the current cycle torque command for the construction of the next cycle's observations, an execution closed loop is formed, enabling the friction parameters to be automatically updated throughout the entire lifecycle without manual calibration.
Smart Images

Figure CN122043974B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent chassis control for new energy vehicles, and in particular to a method and system for adaptive calibration of road surface feedback for steer-by-wire. Background Technology
[0002] As a trend in the development of modern intelligent vehicles, steer-by-wire technology not only saves installation space in the front of the driver's cab but also expands the design space for steering force and angular transmission characteristics. However, because the steer-by-wire system eliminates the mechanical connection between the steering wheel and the steering actuator, the driver's steering input and road surface interaction information are transmitted in a closed loop through sensors and controllers. Since the mechanical connection is eliminated, information such as changes in road surface adhesion, roughness, impact, tire return to center, and lateral deviation needs to be actively synthesized and output by the tactile feedback interaction module on the steering wheel side to form road feedback torque.
[0003] In engineering applications, steering wheel road feedback actuators typically include a road feedback motor, steering column / bearing structure, angle / torque sensor array, and controller. This actuator chain inevitably involves equivalent friction and damping, such as Coulomb friction, viscous damping, low-speed stick-slip, and nonlinear friction caused by assembly preload. Therefore, it faces the following challenges: 1. Significant individual differences: Different batches of motors, reducers, tooth backlash, bearing preload, lubrication and assembly deviations can lead to discrete friction parameters, resulting in inconsistent feel in different vehicles of the same model.
[0004] 2. Time-varying throughout the entire life cycle: Changes in temperature, wear and aging, and lubrication conditions can cause friction parameters to drift over time, leading to the gradual failure of factory calibration during use, increased road feel tracking error, and user experience issues such as zero-crossing stickiness and hysteresis.
[0005] 3. Real-time performance and resource constraints: Mass-produced ECUs have limited computing power and bandwidth. Overly complex online observation / adaptive control structures can lead to problems such as time latency sensitivity, difficulty in parameter tuning, and sensitivity to noise, and are difficult to deploy quickly across multiple vehicle platforms.
[0006] 4. Lack of an engineered online update gating mechanism: In scenarios such as driver intervention, insufficient low-speed observability, and high impact noise, parameter updates are easily contaminated, leading to compensation divergence or the introduction of new torque jitter.
[0007] Existing solutions focus on two main approaches: One type relies on manual / semi-manual calibration combined with fixed parameter compensation, but it is difficult to cover individual differences and aging drift, and consistency is difficult to guarantee. Moreover, it is time-consuming and labor-intensive to adjust the friction compensation parameters for each vehicle at the factory stage, and it is highly dependent on tooling and experience, making it difficult to standardize and scale up. Another type employs more complex online compensation and observation of friction parameters, but often requires more model parameters, state observers and robust control components. The algorithm structure is complex, the engineering implementation cost is high, and the real-time pressure is great, which is not conducive to mass production and rapid after-sales implementation.
[0008] Therefore, while ensuring the accuracy of road feel output, there is an urgent need for a road feedback friction calibration and compensation method with fewer parameters, lighter calculations, and adaptive updates throughout the vehicle's life cycle, so that the steering wheel output road feel torque can be consistent with the ideal synthesized torque, and the cost of manual calibration at the factory can be significantly reduced. Summary of the Invention
[0009] The technical problem to be solved by this invention: (1) How to achieve online identification and adaptive calibration of the equivalent friction of the steering wheel road feedback execution subsystem using only a few parameters, so as to automatically update the compensation amount under different individual differences and different usage intensity conditions, and avoid manual calibration for each vehicle.
[0010] (2) How to ensure real-time performance under the condition of limited ECU resources in mass production: Construct an online identification and compensation architecture with low computational load, simple update rules, and insensitivity to noise and time delay.
[0011] (3) How to ensure the safety and robustness of online updates: In cases of driver intervention, low-speed unobservable conditions, and abnormal faults, the calibration updates are reliable and do not introduce new road feel instability factors through mechanisms such as gating weight restrictions, residual criteria and parameter projection constraints.
[0012] To address the technical problems to be solved, this invention proposes an adaptive calibration method and system for steer-by-wire road surface feedback. The calibration system of this invention achieves consistency in road surface feedback across different batches of actuators, different assembly states, and different lifecycle conditions through controlled calibration excitation, lightweight real-time identification, and online updating of calibration parameters. It also provides a hardware structure design to enhance the observability and safety of the calibration.
[0013] In a first aspect, the present invention proposes a steer-by-wire road surface feedback adaptive calibration method, comprising: In the current control cycle, the steering wheel output resistance torque variable and steering wheel angle variable are collected, and the ideal road feedback resistance torque variable is generated based on the steering wheel angle variable and the preset road feedback strategy. The mechanical friction observation variable is obtained based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle. Based on the observed mechanical friction variables, the mechanical friction state variables of the current control cycle are updated online, and the predicted mechanical friction values for the next control cycle are output according to the updated mechanical friction state variables. Based on the ideal road surface feedback resistance torque variable, the mechanical friction prediction value variable for the next control cycle, and the steering wheel output resistance torque variable, the torque command variable for the current control cycle is generated. The torque command variable of the current control cycle is sent to the road feedback system for execution, and the torque command variable of the current control cycle is stored for constructing mechanical friction observation variables in the next control cycle.
[0014] Furthermore, the online update of the mechanical friction state variables for the current control cycle includes: Mechanical friction observation residual variables are constructed based on the mechanical friction observation variables and prior prediction values of the current control cycle; the prior prediction values are generated based on the mechanical friction state variables and control cycle variables after correction and update of the previous control cycle. Residual gate weight variables are constructed based on the residual variables observed from mechanical friction and the preset gate scale parameters; Based on the mechanical friction observation residual variables, residual gating weight variables, and preset observer correction gain parameters, the mechanical friction state variables of the current control cycle are corrected and updated.
[0015] Furthermore, the mechanical friction state variables include mechanical friction estimate variables, mechanical friction rate of change estimate variables, and comprehensive bias estimate variables.
[0016] Furthermore, the prior prediction value includes the prior estimate of the mechanical friction correction value before the current control cycle. The prior estimate of the rate of change of mechanical friction during the current control cycle before correction. Prior estimate of the comprehensive bias correction for the current control cycle The formula for calculating the prior prediction value is as follows:
[0017] in, The variable representing the estimated mechanical friction value from the previous control cycle; The variable representing the estimated rate of change of mechanical friction in the previous control cycle; The variable representing the overall bias estimate of the previous period; This indicates a control periodic variable.
[0018] Furthermore, the formula for calculating the residual variable of the mechanical friction observation is as follows:
[0019] in, This represents the mechanical friction observation residual variable for the current control cycle; This represents the mechanical friction observation variable for the current control cycle.
[0020] Furthermore, the formula for calculating the residual gating weight variable is as follows:
[0021] in, Represents the residual gating weight variable; This represents the gating scale parameter.
[0022] Furthermore, the calculation formula for correcting and updating the mechanical friction state variables is as follows:
[0023] in, This represents the mechanical friction estimate variable updated after the current control cycle correction. This represents the estimated rate of change of mechanical friction after the current control cycle correction update. This variable represents the overall bias estimate after the current control cycle correction update. This represents the observer's master state correction gain parameter variable; This represents the observer rate of change correction gain parameter variable; This represents the observer bias correction gain parameter variable.
[0024] Furthermore, the torque command variables for generating the current control cycle include: The resistance torque error variable is calculated based on the ideal road surface feedback resistance torque variable and the steering wheel output resistance torque variable. By combining the resistance torque error variable and the error integral state variable of the previous control cycle, the error integral state variable of the current control cycle is updated and obtained. The intermediate variable for feedback correction is obtained by performing proportional-integral calculations based on the resistance torque error variable and the error integral state variable. The intermediate variable of feedforward friction compensation is obtained by the difference between the ideal road surface feedback resistance torque variable and the mechanical friction prediction variable for the next control cycle. The torque command variable for the current control cycle is obtained by superimposing the intermediate variables of the feedforward friction compensation and the feedback correction.
[0025] In a second aspect, a steer-by-wire road surface feedback adaptive calibration system is provided for performing the method described in the first aspect, comprising: The signal acquisition module is used to acquire the steering wheel output resistance torque variable and steering wheel angle variable in the current control cycle, generate the steering wheel angular velocity variable, and store the torque command variable of the previous control cycle; the steering wheel angular velocity variable is generated based on the steering wheel angle variable and the control cycle variable. The ideal road surface feedback torque generation module is used to generate the ideal road surface feedback resistance torque variable based on the steering wheel angular velocity variable and the preset road surface feedback strategy. The mechanical friction prediction observer is used to obtain the mechanical friction observation variable based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle. Based on the mechanical friction observation variable, the mechanical friction state variable of the current control cycle is updated online, and the mechanical friction prediction value variable of the next control cycle is output based on the updated mechanical friction state variable. The composite control module is used to generate the torque command variable for the current control cycle based on the ideal road surface feedback resistance torque variable, the mechanical friction prediction value variable for the next control cycle, and the steering wheel output resistance torque variable. A road surface feedback execution system is used to receive and execute the torque command variables.
[0026] Furthermore, the mechanical friction prediction observer includes: The measurement construction unit is used to construct the mechanical friction observation variable based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle. The prior prediction calculation unit is used to generate prior prediction values based on the corrected and updated mechanical friction state variables and control cycle variables from the previous control cycle. Residual construction unit, used to construct mechanical friction observation residual variables based on mechanical friction observation variables and prior prediction values; The gate weight calculation unit is used to construct residual gate weight variables based on the mechanical friction observation residual variables and preset gate scale parameters; The gain register unit is used to store the observer correction gain parameters required for correcting and updating the mechanical friction state variables; The residual correction calculation unit is used to correct and update the mechanical friction state variables of the current control cycle based on the mechanical friction observation residual variables, residual gating weight variables, and preset observer correction gain parameters. The state register unit is used to store the mechanical friction state variables before and after the correction update.
[0027] In this invention, the online update of the mechanical friction state variables relies solely on the steering wheel output resistance torque and the torque command of the previous cycle, decoupling it from the ideal road surface feedback resistance torque and ensuring the independence of friction estimation. Furthermore, feedforward compensation is performed using the predicted mechanical friction value of the next cycle, combined with feedback correction to generate torque commands, achieving adaptive compensation for friction in the execution link. By storing the current cycle torque command for the construction of the next cycle's observations, an execution closed loop is formed, enabling the friction parameters to be automatically updated throughout the entire lifecycle without manual calibration.
[0028] This invention can automatically track friction changes and update friction compensation using a small number of parameters. It is suitable for batch standardized calibration when vehicles leave the factory. The friction compensation calculation is lightweight, insensitive to noise and time delay, and suitable for real-time deployment under the condition of limited ECU resources in mass production. It can ensure the real-time performance and stability of compensation. Attached Figure Description
[0029] Figure 1 This is a flowchart of a road surface feedback adaptive calibration method for steer-by-wire proposed in this invention; Figure 2 This is a structural diagram of a road surface feedback adaptive calibration system for steer-by-wire proposed in this invention. Detailed Implementation
[0030] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments. The embodiments are implemented based on the technical solution of the present invention, providing detailed implementation methods and specific operation processes. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them, and the scope of protection of the present invention is not limited to the following embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0031] This invention proposes an adaptive calibration method for road surface feedback in steer-by-wire, which will be described in detail below with reference to specific embodiments.
[0032] Before implementing this method in the calibration system, the internal state of the mechanical friction prediction observer must be initialized: Estimated variables of mechanical friction Estimated value of the rate of change of mechanical friction and the combined bias estimate variable Initialize to zero.
[0033] Error integral state variables in the composite control module Initialize to zero.
[0034] In a specific embodiment, this method includes the following: After the vehicle is powered on, it performs an initial self-test. If the self-test fails, it will either handle the error or go into standby mode. Once the self-test passes, it will enter the following high-frequency control cycle.
[0035] S1. Collect multi-dimensional state data during the current control cycle, including the steering wheel output resistance torque variable. Steering wheel angle variable Steering wheel angle variable Combined with control period variables Generate steering wheel angular velocity variable Based on steering wheel angular velocity variable Generates ideal road feedback resistance torque variables based on preset road feedback strategies. .
[0036] Among them, the steering wheel angular velocity variable The calculation formula is:
[0037] in The steering wheel angle variable was collected in the previous control cycle; This indicates a control periodic variable, which is also a data acquisition periodic variable.
[0038] Among them, the ideal road surface feedback resistance torque variable This represents the target road feel torque that the system expects to present to the driver from the steering wheel side, without considering frictional losses in transmission link 2. It is obtained based on the analytical road adhesion coefficient. This ideal road feedback resistance torque variable... Used as a reference input for subsequent composite control.
[0039] Specifically, the ideal road surface feedback resistance torque variable Generated based on vehicle operating parameters, the low-speed driving condition is designed according to changes in steering wheel angle, while the high-speed driving condition is designed according to changes in lateral acceleration. A vehicle speed weighting function is used to achieve flexible switching between the two conditions.
[0040] in, The equivalent electromagnetic torque term of the steering actuator motor; For low-speed feedback functions, It is a high-speed feedback function; The vehicle speed weighting function; It is lateral acceleration; As an additional term, it is preferably composed of a commutation compensation term driven by the steering wheel angular velocity variable and a damping term that increases with vehicle speed.
[0041] S2. Steering wheel output resistance torque variable based on the current control cycle. Torque command variable from the previous control cycle The difference between the two values yields the mechanical friction observation variables for the current control cycle. , Mechanical friction observable variables It is used to reflect the difference between the steering wheel output resistance torque in the current control cycle and the command torque in the previous control cycle, which includes the influence of mechanical friction.
[0042] S3, Based on mechanical friction observation variables The system updates the mechanical friction state variables for the current control cycle online and outputs the predicted mechanical friction values for the next control cycle based on the updated mechanical friction state variables. .
[0043] Specifically, the online updates of the mechanical friction state variables for the current control cycle include: S21. Mechanical friction observation variables based on the current control cycle Constructing mechanical friction observation residual variables from prior predictions The prior predicted values are updated based on the corrected mechanical friction state variables and control period variables from the previous control cycle. generate.
[0044] Specifically, the mechanical friction state variables include the estimated mechanical friction values, the estimated rate of change of mechanical friction values, and the comprehensive bias estimated values. Prior predictions include the predicted mechanical friction values before correction for the current control cycle. The predicted value of the rate of change of mechanical friction in the current control cycle before correction. Predicted values of the current control cycle before comprehensive bias correction .
[0045] Among them, the predicted value variable before mechanical friction correction The formula for calculation is:
[0046] Among them, the predicted value variable before correction of the rate of change of mechanical friction. The formula for calculation is:
[0047] Among them, the predicted value variable before comprehensive bias correction The formula for calculation is:
[0048] in, The variable representing the estimated mechanical friction value from the previous control cycle; The variable representing the estimated rate of change of mechanical friction in the previous control cycle; The variable representing the overall bias estimate of the previous period; This indicates a control periodic variable.
[0049] Specifically, mechanical friction observation residual variables This reflects the deviation between actual observations and prior predictions, and is used to drive subsequent corrections and updates to the mechanical friction state variables. The formula for calculating the mechanical friction observation residual variables is:
[0050] in, This represents the mechanical friction observation residual variable for the current control cycle; This represents the mechanical friction observation variable for the current control cycle.
[0051] S22. To improve robustness under abnormal operating conditions, the residual variables observed under mechanical friction are... and preset gating scale parameters Constructing residual gated weight variables .
[0052] Specifically, residual gate weight variables The calculation formula is:
[0053] in, Represents the residual gating weight variable; This represents the gated scale parameter, which is a preset positive calibration parameter. Its dimensions are the same as the mechanical friction observation residual variable. The same applies, and the residual variables observed under normal steering wheel system operating conditions are used. The typical amplitude is determined. This residual gating weight variable... The value range is (0,1], when the mechanical friction observation residual variable The absolute value is much larger than the gating scale parameter. hour, Approaching 0, thereby suppressing / freezing abnormal observation residuals, so that the observer can maintain stable update logic under sudden impact, signal glitch or abnormal operating conditions, and avoid friction estimation being dragged off by outliers and causing road feel torque oscillation.
[0054] S23, Residual variables observed based on mechanical friction Residual gate weight variables and preset observer correction gain parameters , , The mechanical friction state variables for the current control cycle are corrected and updated using the following formula: The mechanical friction estimate variable after current control cycle correction update The updated formula is:
[0055] The estimated value of the rate of change of mechanical friction after the current control cycle correction is as follows: The updated formula is:
[0056] The updated comprehensive bias estimate variable after current control cycle correction The updated formula is:
[0057] in, This represents the observer's master state correction gain parameter variable; This represents the observer rate of change correction gain parameter variable; This represents the observer bias correction gain parameter variable.
[0058] Specifically, the predicted value of mechanical friction for the next control cycle. The calculation formula is:
[0059] S4. Based on the ideal road surface feedback resistance torque variable Mechanical friction prediction variable for the next control cycle and steering wheel output resistance torque variable Generate the torque command variable for the current control cycle. .
[0060] Specifically, the torque command variables for the current control cycle include: S41, Based on ideal road surface feedback resistance torque variable and steering wheel output resistance torque variable Calculate the resistance torque error variable The calculation formula is:
[0061] S42, Combining resistance torque error variable The state variable is the integral of the error from the previous control cycle. Update the error integral state variable for the current control cycle. The calculation formula is:
[0062] S43. Based on the resistance torque error variable Sum of error integral state variables Perform proportional-integral calculations to obtain the intermediate variable for feedback correction. The calculation formula is:
[0063] in, Indicates the proportional gain parameter variable; This represents the integral gain parameter variable. The proportional gain parameter variable... and integral gain parameter variables These are general control parameters that are uniformly determined during the design phase of the composite control module and are not manually calibrated as mechanical friction parameters for individual vehicles.
[0064] S44. Based on the ideal road surface feedback resistance torque variable The predicted value of mechanical friction in the next control cycle The difference is used to obtain the intermediate variable of feedforward friction compensation. The calculation formula is:
[0065] S45. Adjust the intermediate variable of feedforward friction compensation. and feedback correction intermediate variables The torque command variable for the current control cycle is obtained by superposition. The calculation formula is:
[0066] S5. Set the torque command variable for the current control cycle. The torque command variable for the current control cycle is sent to the road feedback system for execution and stored. This is used to construct the mechanical friction observation variables for the next control cycle. Then, it is determined whether the vehicle is powered off. If it is powered off, the loop ends; if it is still powered on, the loop restarts from S1.
[0067] Specifically, the drive command conversion unit of the road feedback motor driver is based on the torque command variable. Forming current command variables And according to the current command variable A drive control signal is generated. The road feedback motor 3 acts on the steering wheel assembly 1 via the transmission link 2, forming a new steering wheel output resistance torque variable. .
[0068] The above process is repeated in each control cycle to achieve online recursive updates of mechanical friction state variables and adaptive calibration compensation of road feedback torque. In the composite control process for generating torque commands, the ideal road feedback resistance torque serves as a reference, superimposed with feedforward compensation based on the predicted mechanical friction value of the next control cycle. Simultaneously, proportional / integral feedback correction of the resistance torque error is introduced, ensuring that the final output torque command balances response speed and steady-state accuracy. Furthermore, the torque command of the current control cycle is stored and used in the next control cycle, allowing the observer to synchronously construct observations using the steering wheel output resistance torque and the torque command of the previous control cycle within the same control cycle. This guarantees the timing consistency and feasibility of the observation-prediction-control link.
[0069] In the above method, the present invention achieves online estimation and compensation of transmission link friction through mechanical friction prediction and observation with only a few parameters. At the same time, the residual gated weight variables ensure robustness under abnormal working conditions, which can effectively improve the consistency of road feedback torque of the steer-by-wire system throughout its entire life cycle and reduce the cost of manual calibration at the factory.
[0070] In specific implementation methods, such as Figure 2 As shown, the present invention also provides a steer-by-wire road surface feedback adaptive calibration system for implementing the above method, including a signal acquisition module, an ideal road surface feedback torque generation module, a mechanical friction prediction observer, a composite control module, and a road surface feedback execution system.
[0071] The road feedback execution system includes a steering wheel assembly 1, a road feedback motor 3, and a transmission link 2. The output end of the transmission link 2 is connected to the rotation shaft of the steering wheel assembly 1, and the output end of the road feedback motor 3 is connected to the input end of the transmission link 2. The road feedback motor 3 has a drive input end, which is connected to the output of the composite control module to receive torque command variables. The corresponding drive commands. The steering wheel assembly 1 and the transmission link 2 form a mechanical link. This mechanical link has a time-varying frictional resistance torque. This time-varying frictional resistance torque, together with the resistance torque applied to the steering wheel assembly 1 by the road feedback motor 3 via the transmission link 2, causes the steering wheel to output a variable resistance torque. With torque command variable There were differences in execution.
[0072] The signal acquisition module includes a resistance torque acquisition channel, a rotation angle acquisition channel, an angular velocity generation channel, and a command recording channel.
[0073] The input of the drag torque acquisition channel is connected to steering wheel assembly 1, and the output is connected to the input of the mechanical friction prediction observer and the input of the composite control module, respectively. The drag torque acquisition channel is used to output the acquired steering wheel drag torque variable. .
[0074] The input of the angle acquisition channel is connected to steering wheel assembly 1, and the output is connected to the angular velocity channel. The angle acquisition channel is used to output the steering wheel angle variable. .
[0075] Angular velocity generation channel is used to generate data based on steering wheel angle variables. and control (sampling) periodic variables Forming steering wheel angular velocity variable Steering wheel angular velocity variable The output is sent to the ideal road surface feedback torque generation module.
[0076] The input of the command recording channel is connected to the output of the composite control module, and is used to receive and store the torque command variables of the current control cycle. .
[0077] Among them, the ideal road surface feedback torque generation module is used to generate torque based on the steering wheel angular velocity variable. The ideal road feedback resistance torque variable is generated by the preset road feedback strategy. It outputs the ideal road surface feedback resistance torque variable. The input to the composite control module. The ideal road surface feedback torque generation module is not directly connected to the mechanical friction prediction observer, and the mechanical friction prediction observer does not read the ideal road surface feedback resistance torque variable. This allows the input of the mechanical friction prediction observer to depend only on the steering wheel output drag torque variable. and the torque command variable of the previous cycle , and the ideal road surface feedback resistance torque variable Decoupling avoids the impact of changes in the ideal road surface feedback strategy on friction estimation, ensuring the independence and stability of friction estimation.
[0078] The input of the mechanical friction prediction observer is connected to the output of the drag torque acquisition channel and the output of the command recording channel, and is used to receive the drag torque variable output from the steering wheel. Torque command variable from the previous cycle The output of the mechanical friction prediction observer is connected to the input of the composite control module to output the predicted mechanical friction value for the next cycle. .
[0079] The mechanical friction prediction observer includes: The measurement unit is used to measure the steering wheel output resistance torque variable based on the current control cycle. Torque command variable from the previous control cycle The difference in mechanical friction observation variables , .
[0080] The prior prediction calculation unit is used to correct and update the mechanical friction state variables and control period variables based on the previous control cycle. Generate prior predicted values; among which, the mechanical friction state variables include mechanical friction estimate variables, mechanical friction rate of change estimate variables, and comprehensive bias estimate variables, and the prior predicted values include mechanical friction pre-correction predicted values. Predicted value of mechanical friction change rate before correction Predicted values before comprehensive bias correction .
[0081] Residual construction elements are used to construct units based on the observed variables of mechanical friction. Predicted values of variables before mechanical friction correction Predicted values before comprehensive bias correction Structural mechanical friction observation residual variables ; The gated weight calculation unit is used to calculate the residual variables observed from mechanical friction. and preset gating scale parameters Constructing residual gated weight variables Gating scale parameter variables Take mechanical friction observation residual variable The statistical mean within the normal working range.
[0082] The gain register unit stores the observer correction gain parameters required for correcting and updating the mechanical friction state variables, including the observer master state correction gain parameter variables. Observer rate of change correction gain parameter variable and observer bias correction gain parameter variable .
[0083] The residual correction calculation unit is used for residual variables based on mechanical friction observations. Residual gate weight variables and preset observer correction gain parameters The mechanical friction state variables of the current control cycle are corrected and updated.
[0084] The state register unit is used to store the mechanical friction state variables before and after the correction update.
[0085] The composite control module, officially named the feedforward friction compensation and resistance torque error feedback composite control module, can also be called a composite controller. The composite control module includes: The error calculation unit is used to calculate the feedback resistance torque variable based on the ideal road surface. and steering wheel output resistance torque variable Constructing the resistance torque error variable ; The integral state register is used to store the error integral state variable. Error integral state variable Based on the resistance torque error variable and control periodic variables Recursive update; The feedforward friction compensation synthesis unit is used to feed back the resistance torque variable based on the ideal road surface. The predicted value of mechanical friction in the next control cycle Forming intermediate variables for feedforward friction compensation ; Feedback correction synthesis unit, used to adjust the resistance torque error variable Error integral state variables Proportional gain parameter variable and integral gain parameter variables Forming feedback correction intermediate variables ; The composite generation unit, based on the intermediate variable of feedforward friction compensation. and feedback correction intermediate variables Forming torque command variables .
[0086] The input of the composite control module is connected to the output of the ideal road surface feedback torque generation module, the output of the mechanical friction prediction observer, and the output of the resistance torque acquisition channel to receive the ideal road surface feedback resistance torque variable. Next cycle mechanical friction prediction variable Steering wheel output resistance torque variable The output of the composite control module is connected to the drive input of the road feedback motor 3, which transmits the torque command variable. The corresponding drive control quantity is sent to the road feedback motor driver for execution; the drive command conversion unit of the road feedback motor driver generates a current command variable based on the torque command variable, and generates a drive control signal accordingly. The drive control signal acts on the power switching device inside the road feedback motor driver to control the road feedback motor 3 to output the target resistance torque corresponding to the torque command variable; under the action of the driver, the road feedback motor 3 applies the target resistance torque to the steering wheel assembly 1 through the transmission link 2, thereby forming a closed loop with the signal acquisition module.
[0087] The present invention has the following technical effects: 1. By collecting only a small number of parameters, the mechanical friction state variables are updated online recursively through a mechanical friction prediction observer. Only three core mechanical friction state variables—the estimated mechanical friction value, the estimated mechanical friction rate of change, and the comprehensive bias estimate—and three preset gain parameters—the observer's main state correction gain, the observer's rate of change correction gain, and the observer's bias correction gain—are required to achieve online identification and adaptive calibration of the equivalent friction of the steering wheel road feedback execution system. Under different individual differences and lifespan drift conditions, the observer can automatically track friction changes and update the friction compensation amount, eliminating the need for manual calibration for each vehicle, significantly reducing factory calibration costs and after-sales maintenance workload.
[0088] 2. The mechanical friction prediction variable calculation process of this invention adopts a low-order linear parameterized model and a discrete recursive update architecture. Each control cycle only requires a small number of addition, subtraction, multiplication, and division operations, without complex matrix operations or iterative optimization processes. The computational load is small and the update law is simple. The construction of the mechanical friction observation variables only requires the steering wheel output resistance torque and the torque command of the previous cycle. There is no need for indirect calculation channels that are susceptible to noise interference, such as motor current conversion. It is not sensitive to sensor noise and control delay, and is suitable for real-time deployment under the limited resources of mass-produced ECUs, ensuring the real-time performance and stability of compensation.
[0089] 3. This invention introduces residual gating weight calculation in the updating of mechanical friction state variables. The residual gating weight variable is dynamically generated based on the ratio formula between the mechanical friction observation residual variable and the preset gating scale parameter. When the driver actively intervenes, the vehicle is in a low-speed unobservable zone, or the sensor experiences impact interference resulting in abnormal observation residuals, the residual gating weight variable approaches zero, effectively suppressing or freezing the impact of outliers on the observer's state update, preventing the friction estimate from being skewed and causing road feel torque oscillations. Simultaneously, the correction and updating of the mechanical friction state variables are always performed within the preset gain parameter framework. Reasonable tuning of the gain parameter can further constrain the state update amplitude, ensuring the safety and robustness of online updates and preventing the introduction of new road feel instability factors.
[0090] This invention focuses on a systematic calibration system and lightweight algorithms. By using controlled excitation and discrete recursive identification, it can quickly update key road feel parameters, transforming the consistency of road feedback from relying on manual tuning to automatic calibration and updating. This makes it more suitable for mass production and after-sales closed-loop implementation.
[0091] The term "an embodiment" or "embodiment" as used in this invention refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the invention. In the description of this invention, it should be understood that the terms "first," "second," and "third," etc., in the specification, claims, and accompanying drawings are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.
[0092] This specification provides method operation steps as shown in the embodiments or flowcharts, but based on conventional or non-inventive labor, more or fewer operation steps may be included. The order of steps listed in the embodiments is merely one of many possible execution orders and does not represent the only possible execution order. In actual system or server product execution, the method can be executed in the order shown in the embodiments or drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment), or the execution order of steps without timing constraints can be adjusted.
[0093] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.
Claims
1. A method for adaptive calibration of road surface feedback in steer-by-wire, characterized in that, include: In the current control cycle, the steering wheel output resistance torque variable and steering wheel angle variable are collected, and the ideal road feedback resistance torque variable is generated based on the steering wheel angle variable and the preset road feedback strategy. Based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle, the mechanical friction observation variable is obtained. ; Update the mechanical friction state variables for the current control cycle, and output the predicted mechanical friction values for the next control cycle based on the updated mechanical friction state variables. The update formula for the mechanical friction state variable includes: in, These represent the estimated values of mechanical friction, mechanical friction rate of change, and comprehensive bias after the current control cycle correction and update, respectively. This represents the predicted value variable before mechanical friction correction. ; This represents the predicted value of the rate of change of mechanical friction before correction. ; This represents the predicted value variable before comprehensive bias correction. ; Indicates the control periodic variable; Indicates the sequence number of the control cycle; This represents the observed residual variable of mechanical friction during the current control period. ; This indicates the residual variable observed based on mechanical friction. and preset gating scale parameters Constructed residual gated weight variables, These represent the observer master state correction gain parameter variable, the observer rate of change correction gain parameter variable, and the observer bias correction gain parameter variable, respectively. Based on the ideal road surface feedback resistance torque variable, the mechanical friction prediction value variable for the next control cycle, and the steering wheel output resistance torque variable, the torque command variable for the current control cycle is generated. The torque command variable of the current control cycle is sent to the road feedback system for execution, and the torque command variable of the current control cycle is stored for constructing mechanical friction observation variables in the next control cycle.
2. The method according to claim 1, characterized in that, The torque command variables used to generate the current control cycle include: The resistance torque error variable is calculated based on the ideal road surface feedback resistance torque variable and the steering wheel output resistance torque variable. By combining the resistance torque error variable and the error integral state variable of the previous control cycle, the error integral state variable of the current control cycle is updated and obtained. The intermediate variable for feedback correction is obtained by performing proportional-integral calculations based on the resistance torque error variable and the error integral state variable. The intermediate variable of feedforward friction compensation is obtained by the difference between the ideal road surface feedback resistance torque variable and the mechanical friction prediction variable for the next control cycle. The torque command variable for the current control cycle is obtained by superimposing the intermediate variables of the feedforward friction compensation and the feedback correction.
3. A steer-by-wire road surface feedback adaptive calibration system, used to execute the method according to any one of claims 1-2, characterized in that, include: The signal acquisition module is used to acquire the steering wheel output resistance torque variable and steering wheel angle variable in the current control cycle, generate the steering wheel angular velocity variable, and store the torque command variable of the previous control cycle. The steering wheel angular velocity variable is generated based on the steering wheel angle variable and the control cycle variable; The ideal road surface feedback torque generation module is used to generate the ideal road surface feedback resistance torque variable based on the steering wheel angular velocity variable and the preset road surface feedback strategy. The mechanical friction prediction observer is used to obtain the mechanical friction observation variable based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle. Based on the mechanical friction observation variable, the mechanical friction state variable of the current control cycle is updated online, and the mechanical friction prediction value variable of the next control cycle is output based on the updated mechanical friction state variable. The composite control module is used to generate the torque command variable for the current control cycle based on the ideal road surface feedback resistance torque variable, the mechanical friction prediction value variable for the next control cycle, and the steering wheel output resistance torque variable. A road surface feedback execution system is used to receive and execute the torque command variables.
4. The calibration system according to claim 3, characterized in that, The mechanical friction prediction observer includes: The measurement construction unit is used to construct the mechanical friction observation variable based on the difference between the steering wheel output resistance torque variable of the current control cycle and the torque command variable of the previous control cycle. The prior prediction calculation unit is used to generate prior prediction values based on the corrected and updated mechanical friction state variables and control cycle variables from the previous control cycle. Residual construction unit, used to construct mechanical friction observation residual variables based on mechanical friction observation variables and prior prediction values; The gate weight calculation unit is used to construct residual gate weight variables based on the mechanical friction observation residual variables and preset gate scale parameters; The gain register unit is used to store the observer correction gain parameters required for correcting and updating the mechanical friction state variables; The residual correction calculation unit is used to correct and update the mechanical friction state variables of the current control cycle based on the mechanical friction observation residual variables, residual gating weight variables, and preset observer correction gain parameters. The state register unit is used to store the mechanical friction state variables before and after the correction update.
Citation Information
Patent Citations
Friction compensation in a vehicle steering system
US20040138797A1