Road sensing torque simulation method and system for automobile steer-by-wire system

Abstract

The invention relates to the technical field of automobile steer-by-wire and road sensing torque control thereof, in particular to a road sensing torque simulation method and system for an automobile steer-by-wire system. The vehicle state parameters at least comprise vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement and torque information of a steering execution motor; calculating a basic rack force by adopting a fusion estimation strategy; according to the vehicle state parameters, multiple compensation torque components are calculated in parallel, and the compensation torque components comprise return compensation torque, friction compensation torque and inertia compensation torque; according to the steering wheel turning angle, soft dead center limiting torque is calculated; and performing weighted fusion on the basic rack force, the return compensation torque, the friction compensation torque, the inertia compensation torque and the soft dead point limiting torque to generate a final target hand wheel feedback torque. According to the invention, the delay feeling of steer-by-wire is effectively inhibited, and the control experience with rich information and natural feedback is provided for a driver.

Description

Technical Field

[0001] This invention relates to the technical field of steer-by-wire and its road feel torque reconstruction control, and particularly to a method and system for simulating road feel torque in a steer-by-wire system. Background Technology

[0002] In traditional automotive steering systems, the steering wheel and steering wheels are rigidly connected by mechanical structures such as the steering column and steering gear. When the driver turns the steering wheel, they can clearly perceive information from the road surface through this mechanical connection, such as the friction between the tires and the ground and the vehicle's self-centering force. This perception is usually called road feel, and a good road feel is a key factor in ensuring driving safety and improving the driving experience. With the development of automotive technology, steer-by-wire (SBW) systems have emerged. In SBW systems, there is no mechanical connection between the steering wheel and the steering actuator. The driver's steering intention is converted into an electrical signal by sensors, which is then processed by the controller (ECU) to drive the steering motor to complete the steering of the wheels. This design brings many advantages, such as modular design, variable steering ratio, and easy integration of advanced driver assistance systems (ADAS).

[0003] However, because the mechanical connection is eliminated, road feel information cannot be naturally transmitted to the driver. Existing steer-by-wire systems usually simulate road feel by installing a feedback motor at the steering wheel. However, most current road feel simulation methods are simple models that only consider one or a few factors, resulting in unrealistic simulated road feel and harsh feedback. Drivers find it difficult to accurately judge the vehicle's status and may even feel unsafe at high speeds or in complex road conditions. For example, the steering wheel may feel too heavy at low speeds and too light at high speeds; the return torque may be nonlinear and lack damping; and there may be a lack of clear boundary feel at the steering limit.

[0004] Therefore, how to establish an accurate and comprehensive road feel model to provide objective data support for subjective feelings, thereby simulating a delicate and realistic steering feel for drivers that is close to that of traditional mechanical steering systems, is a technical problem that urgently needs to be solved in the field of steer-by-wire technology.

[0005] CN113799872A discloses a control method and system based on steer-by-wire road feel simulation. The road feel motor controller receives steering wheel angle and torque signals, and combines the EPS motor output torque, damping / return / friction / soft stop compensation components obtained from steering wheel angle and vehicle speed, as well as thermal protection, road surface simulation and other information to form a total torque request and output it to the road feel motor. This scheme tends to rely more on the parallel superposition of multi-source compensation for the formation path of key underlying quantities such as basic load / rack force. It lacks a strong emphasis on the structured fusion between observation channels and model channels and the principle of working condition weight transfer. Therefore, when facing scenarios with low speed and strong friction, high speed and high adhesion changes or drastic changes in input steering dynamics, it may still produce problems such as unclear dominance of the feel component and strong reliance on calibration experience for feel consistency at different vehicle speeds.

[0006] CN110606121A discloses a steer-by-wire road feel simulation control method, which uses dynamics to construct a steering load model to calculate steering resistance torque, and constructs a load state observer based on the output torque of the steering motor to calculate steering load torque. Then, it adaptively adjusts the proportion according to the difference between the two to form a comprehensive output torque, and combines the electric power steering system model to obtain the road feel torque. This approach does not adequately emphasize the direct mapping of vehicle speed and steering wheel rotation dynamics (such as angular velocity and angular acceleration) in the fusion weight, nor the threshold-based safety adjudication mechanism for abnormal output of the observation channel. At the same time, its systematic description of the parallel construction and unified weighted output of key subjective feel components such as return to center, friction, inertia and soft stop is relatively limited, which may lead to defects in the smoothness of the extreme turning angle boundary, the fineness of return to center, and the shaping of virtual inertia feel during rapid steering.

[0007] In summary, existing steer-by-wire road feel simulation technologies generally face problems such as unclear acquisition paths and weight transfer of basic load quantities, difficulty in controlling the impact of single-channel anomalies, and the need for more refined collaborative fusion rules for multiple feel components and soft stops. To address these shortcomings, this invention proposes a road feel torque simulation method for automotive steer-by-wire systems. Using vehicle speed, steering wheel angle / angular velocity, rack displacement, and actuator motor torque information as unified inputs, a basic rack force is formed through a fusion estimation strategy. Simultaneously, return-to-center, friction, inertia compensation, and soft stop limit components are constructed. Furthermore, a weighted fusion method is used to generate the target handwheel feedback torque and control the handwheel feedback motor to apply it to the steering wheel, thereby solving the aforementioned problems. Summary of the Invention

[0008] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section, as well as in the specification abstract and the title of the invention, to avoid obscuring the purpose of this section, the specification abstract, and the title of the invention. Such simplifications or omissions shall not be used to limit the scope of the invention.

[0009] In view of the aforementioned existing problems, the present invention is proposed.

[0010] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for simulating road feel torque for a vehicle steer-by-wire system, wherein: vehicle state parameters are acquired in real time, the vehicle state parameters including at least vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and torque information of the steering actuator motor; A fusion estimation strategy is used to calculate the basic rack force; Based on the vehicle state parameters, multiple compensation torque components are calculated in parallel, wherein the compensation torque components include homing compensation torque, friction compensation torque and inertia compensation torque; Calculate the soft stop limit torque based on the steering wheel angle; The basic rack force, return torque, friction compensation torque, inertia compensation torque, and soft stop limit torque are weighted and fused to generate the final target handwheel feedback torque.

[0011] As a preferred embodiment of the road feel torque simulation method for automotive steer-by-wire systems described in this invention, the fusion estimation strategy includes: Based on the torque information of the steering actuator motor and the rack displacement, the first rack force is estimated by a strong tracking adaptive Kalman filter observer; Based on the vehicle speed and steering wheel angle, the second rack force is calculated using a vehicle dynamics model. The first rack force and the second rack force are dynamically weighted and fused to obtain the basic rack force.

[0012] In a preferred embodiment of the road feel torque simulation method for a vehicle steer-by-wire system described in this invention, the weighting coefficients of the dynamic weighted fusion are dynamically adjusted based on the vehicle speed and steering wheel angular velocity.

[0013] In a preferred embodiment of the road feel torque simulation method for a vehicle steer-by-wire system described in this invention, the dynamic adjustment of the weighting coefficients includes: Increase the weight of the first rack force when driving at low speeds or when the steering wheel is turned quickly. When traveling at high speed, the weight of the second rack force is increased.

[0014] As a preferred embodiment of the road feel torque simulation method for automotive steer-by-wire systems described in this invention, the fusion estimation strategy further includes a safety redundancy judgment, including: presetting a rack force safety threshold; before performing weighted fusion, determining whether the absolute value of the first rack force exceeds the safety threshold; if it does, directly using the second rack force as the final base rack force, and no further fusion is performed.

[0015] In a preferred embodiment of the road feel torque simulation method for a vehicle steer-by-wire system described in this invention, the return-to-center compensation torque is proportional to the return-to-center torque coefficient and the steering wheel angle, and the return-to-center torque coefficient increases with increasing vehicle speed.

[0016] As a preferred embodiment of the road feel torque simulation method for automotive steer-by-wire systems described in this invention, the friction compensation torque includes a Coulomb friction torque that is opposite to the direction of the steering wheel angular velocity and a viscous friction torque that is proportional to the steering wheel angular velocity.

[0017] A second aspect of the present invention provides a road feel torque simulation system for a vehicle steer-by-wire system, wherein: The parameter acquisition module is used to acquire vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and torque information of the steering actuator motor in real time. The core processing module is used to calculate and generate the target handwheel feedback torque; The torque execution module is used to drive the handwheel feedback motor to apply feedback force to the driver based on the target handwheel feedback torque.

[0018] As a preferred embodiment of the road feel torque simulation system for a vehicle steer-by-wire system according to the present invention, the system further includes one or more processors; The memory stores operable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations, including the flow of the road feel torque simulation method for a vehicle steer-by-wire system as described above.

[0019] A third aspect of the invention provides a computer-readable medium for storing software, wherein the software includes instructions executable by one or more computers, the instructions causing the one or more computers to perform operations, the operations including the flow of the aforementioned method for simulating road feel torque for a vehicle steer-by-wire system.

[0020] The beneficial effects of this invention are as follows: This invention combines the advantages of actual load observation and theoretical model calculation by fusing strong tracking adaptive Kalman filtering with a vehicle dynamics model. In particular, the introduction of the strong tracking adaptive filter enables the system to respond quickly to dynamic events such as sudden changes in road adhesion coefficient, and the tracking of rack force is more accurate and faster, significantly improving the transient realism of road feel feedback. Through dual-channel redundancy design and a safety switching mechanism, the system can seamlessly switch to a reliable backup solution in extreme conditions such as sensor malfunction or model inapplicability. The strong tracking adaptive Kalman filter itself has the ability to adapt to model errors and noise changes, which further enhances the reliability of the main estimation channel, together forming a dual safety guarantee for the system. The dynamic weight adjustment strategy enables the road feel model to intelligently adapt to different driving conditions. The application of strong tracking adaptive Kalman filtering makes the method more adaptable to different types of road surfaces (such as asphalt, cement, ice and snow, and gravel roads), without the need for frequent recalibration of model parameters. The accurate and fast-response rack force estimation, combined with delicate multi-compensation, effectively suppresses the electronic feel and lag of steer-by-wire, providing the driver with an information-rich and natural-feedback driving experience. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a schematic diagram of the road feel torque simulation method and system for automotive steer-by-wire systems as shown in the present invention. Figure 2 This is a schematic diagram of the co-simulation principle shown in this invention; Figure 3 This is a schematic diagram of the basic feedback torque shown in this invention; Figure 4 This is a schematic diagram of the return-to-center compensation torque shown in the present invention; Figure 5 This is a schematic diagram of the rack force estimation results under sinusoidal steering conditions as shown in this invention; Figure 6 This is a schematic diagram of the rack force estimation results under the step steering condition shown in this invention; Figure 7 This is a schematic diagram of the uniaxial tensile test and torque tracking effect shown in this invention; Figure 8 This is a schematic diagram of a unidirectional hysteresis loop illustrating the force-displacement characteristics of the present invention. Figure 9This is a stiffness identification diagram of rack force-path-sensing simulation based on STAKF as shown in this invention. Figure 10 This is a step response diagram of the rack force abrupt change as shown in this invention; Figure 11 This is a schematic diagram of the high-frequency band 18~20Hz for the frequency sweep time-domain test shown in this invention; Figure 12 This is a schematic diagram of the amplitude-frequency characteristics shown in this invention; Figure 13 This is a schematic diagram of the phase frequency characteristics shown in this invention; Figure 14 This is a schematic diagram illustrating the steering ease, steering stiffness, and limiting torque of the present invention. Figure 15 This is a schematic diagram of the central region maneuvering stability shown in this invention; Figure 16 This is a schematic diagram of the steering deviation road feel analysis according to the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0023] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0024] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0025] It's easy to understand that steer-by-wire (SBW) eliminates the mechanical coupling between the steering wheel and the steering wheels, converting the driver's steering intentions entirely into electrical signals. These signals are then directly driven by a motor to drive the steering rack, achieving steering. SBW offers numerous advantages in vehicle layout, steering characteristics, autonomous driving, and active safety design, making it a key focus of current steering system research. In SBW systems, rack force reflects the real-time force state of the steering wheels and is core data for road feel simulation design, directly determining the realism and accuracy of road feel feedback. However, direct measurement of rack force presents several challenges: firstly, dedicated measurement equipment such as tie rod force sensors is expensive, hindering cost control in mass-produced vehicles; secondly, physical sensors may suffer from signal delays, and additional hardware placement is limited by vehicle space and system reliability requirements. Therefore, research on rack force estimation based on existing sensor signals such as motor torque current and torque angle in SBW systems has become a research direction for achieving high-fidelity road feel simulation in steer-by-wire systems.

[0026] While existing technologies have made some progress, they still face challenges: model-based observers are sensitive to parameter perturbations and nonlinear friction, and lack robustness; although traditional Kalman filtering methods can suppress noise, their fixed noise covariance matrix makes them unable to adapt when the system model is mismatched or road conditions change abruptly, resulting in estimation lag or even divergence, meaning that the defect of slow change tracking remains significant; therefore, how to improve the accuracy, dynamic response speed and robustness of rack force estimation without adding additional sensors is the key bottleneck to achieving high-fidelity road feel feedback.

[0027] To address the aforementioned issues, this invention utilizes the strong tracking adaptive Kalman filter (ST-AKF) to track abrupt changes in battery state of origin (SOC) estimation. By applying the ST-AKF algorithm to the online estimation of rack force in SBW (Battery Swing Drive), the readily available q-axis current of the motor is introduced into state observation, and its electromagnetic torque is used as a high-dynamic force reference, significantly enhancing the observability of rack force. Furthermore, by employing strong tracking theory and adjusting the weights of historical data and current measurements in real time through a fading factor, the residual sequence is forced to remain orthogonal, thereby effectively overcoming the estimation lag problem caused by model mismatch and external abrupt changes.

[0028] According to an embodiment of the present invention, in combination Figure 1 The flowchart shown illustrates a method for simulating road feel torque in automotive steer-by-wire systems, which specifically includes the following steps: S1. Real-time acquisition of vehicle status parameters, including at least vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and steering actuator torque information. Note that the following points should be noted in this step: At the start of a control cycle, the system's controller ECU acquires a series of vehicle status parameters via the CAN bus, specifically including: Vehicle speed V: obtained from wheel speed sensors or ABS module; Steering wheel angle θ sw :Measured by an angle sensor mounted on the steering wheel assembly or steering column; Steering wheel angular velocity ω sw By adjusting the steering wheel angle θ sw The difference operation is performed to obtain the result; Steering wheel angular acceleration a sw By adjusting the angular velocity ω of the steering wheel sw The difference operation is performed to obtain the result; Driver input torque T drv :Measured by a torque sensor mounted on the steering column, reflecting the driver's steering intention and force; Torque information of the steering actuator motor (such as the motor's three-phase current Imortor): used for subsequent estimation of the motor's output torque; rack displacement x r Displacement is measured by a displacement sensor mounted on the rack or steering gear.

[0029] The above measurement methods and differential calculation methods can be carried out using existing technologies and means, and will not be elaborated further in this example.

[0030] S2. Calculate the basic rack force using a fusion estimation strategy. Note the following in this step: This embodiment considers the limitations of traditional Kalman filtering when dealing with time-varying, nonlinear, and model uncertainties in vehicle systems. Therefore, it employs a strong tracking adaptive Kalman filter (STAKF) observer. This observer introduces a time-varying fading factor λ into the prediction and update loop of the standard Kalman filter. k By monitoring the output residuals in real time, when the residuals exceed the expected range (indicating model mismatch or abrupt state change), the fading factor λ is activated. k It will be adjusted to reduce the impact of old data on the current state estimation and improve sensitivity to new measurement data, thereby achieving strong tracking of the rack force; the observer takes the motor current Imotor as input and the measured rack displacement x as input. r The measured output is used for calibration, and the final output is the force F applied to the external rack. ext The optimal estimate is denoted as F. KF Compared to traditional filters, STAKF can more effectively suppress filter divergence and respond more quickly and accurately to situations such as sudden changes in road surface conditions.

[0031] Specifically, in order to apply the Kalman filter algorithm, a state-space model of the steering execution system containing the target estimate needs to be established. Based on the dynamic analysis of RWA, its equations of motion are:

[0032] Among them, rack mass 2.25kg, rack displacement x r Damping coefficient B is acquired in real time by the sensor. r The resistance of the rack and pinion is 651 Nm / s / rad. z The value to be determined is the motor output torque T. f For real-time calculation (based on the motor current Imotor), the reduction ratio G fm The pitch circle radius of the pinion is 16.5. p It is 0.007m; For rack displacement speed, For rack displacement The acceleration; Through RWA I q The signal and the above formula are used to calculate the equivalent driving torque T provided by the motor. f In order to transfer the rack force The estimation incorporates state variables, assuming that the rack force changes slowly within a single sampling period, i.e. Based on this assumption, an augmented state vector is constructed. The state-space equation of this physical system can be expressed as:

[0033]

[0034] in, It is the derivative of X with respect to time, and the system input U is the torque T of the motor driving the rack and pinion. f (F is force, T is torque), the system's measurable output Y is the displacement x measured by the sensor. r D is the system's direct transfer matrix (or feedforward matrix), representing the direct effect of the input on the output.

[0035] In the rack and pinion drive system described in this embodiment, since the input torque cannot directly generate instantaneous displacement, matrix D is a zero matrix. The state matrix A, input matrix B, and observation matrix C are respectively:

[0036]

[0037]

[0038] Among them, rack mass It weighs 2.25 kg and has a damping coefficient of B. r It is 651 Nm / s / rad; To facilitate iterative calculations in the controller, the above continuous-time model is discretized, resulting in:

[0039]

[0040] Among them, X k Let X be the system state vector at time k. k-1 Let U be the system state vector at time k-1. k-1 Y is the system control input vector at time k-1. k Let A be the system measurement output vector at time k. d Let B be the system's state transition matrix in discrete time (obtained by discretizing the continuous matrix A). d C is the system's control input matrix in discrete time (obtained by discretizing the continuous matrix B). d This is the system's observation matrix (or output matrix) in discrete time. For example, if the sampling time is T s The specific calculation formula is as follows:

[0041]

[0042] Among them, T s Here, e is the sampling time of the control system, and e is the base of the natural logarithm. Represents the matrix exponential function. For integration variables in integration operations, The derivative of the integral variable.

[0043] Furthermore, the standard Kalman filter is an optimal linear recursive filter that, through an iterative process of prediction and updating, provides the optimal estimate of the system state in a noisy environment. This involves predicting the state at each time step using the system model, then refining the prediction by incorporating the measured value at that time, resulting in a posterior estimate that more closely approximates the true value. Its core equation is as follows: State prediction: ; Error covariance prediction: ; Kalman gain calculation: ; Status Update: ; Error covariance update: ; in, Let A be the predicted state value at time k, that is, the current state calculated based on the state at the previous time;d B d C d The discrete state transition matrix, This is the optimal state estimate at time k-1. P is the actual control input at time k-1. k|k-1 Let P be the prediction error covariance matrix at time k, used to describe the uncertainty of the prediction; k-1|k-1 Let be the estimation error covariance matrix at time k-1. Let A be the discrete state transition matrix. d The transpose of is given, where Q is the covariance matrix of the system process noise, representing the uncertainty of the system model; K k The Kalman gain matrix at time k determines the weight of the measurement on the state correction. C is the discrete observation matrix d The transpose of the matrix is ​​given, where R is the covariance matrix of the measurement noise, representing the uncertainty of the sensor measurement; the superscript -1 indicates that the matrix within the parentheses is inverted. y represents the optimal state estimate at time k, which is the final result combining the predicted and measured values; k The actual measurement value of the sensor at time k. The residual is used to measure the difference between the actual measured value and the predicted measured value. Let I be the estimated error covariance matrix updated at time k, used for iteration at the next time step; I is the identity matrix, with its diagonal elements being 1 and the rest being 0.

[0044] The optimality of a standard Kalman filter is highly dependent on an accurate system model and accurate process and measurement noise statistics. However, in actual steer-by-wire applications, the vehicle's speed, load, and other driving conditions, as well as road conditions (such as dry, wet, and bumpy roads), are constantly changing. This causes changes in the system model parameters, resulting in model mismatch. If a fixed model and noise parameters are used, the filter may over-rely on inaccurate predictions, leading to decreased estimation accuracy or even divergence, and failing to effectively track the real changes in rack force.

[0045] To address the performance degradation of standard Kalman filters under model mismatch, this embodiment introduces a strong tracking filter theory. This includes: when the system model is inaccurate, forcing the filter to forget some outdated historical data and instead rely more heavily on the latest measurement information, thereby maintaining a strong tracking capability of the true system state. This objective is achieved by introducing a time-varying fading factor into the error covariance prediction equation. To achieve this, the corrected error covariance prediction equation is:

[0046] Among them, Pk|k-1 Let P be the prediction error covariance matrix at time k, used to describe the uncertainty of the prediction; k-1|k-1 Let A be the estimation error covariance matrix at time k-1. d The discrete state transition matrix, Let A be the discrete state transition matrix. d The transpose of is given, and Q is the covariance matrix of the system process noise, representing the uncertainty of the system model. It is a gradually diminishing factor.

[0047] When model mismatch is detected, the system will increase... This value will increase the prediction error covariance. This reduces the filter's confidence in the current model's predictions, according to the Kalman gain K. k The calculation formula, a larger one This results in a larger Kalman gain; ultimately, in the state update step, a larger gain will be given to the residual term. The higher weighting of the deviation between measured and predicted information forces the state estimate to quickly converge to the latest measurement, thus achieving strong tracking.

[0048] It should be noted that the fading factor is introduced. The key lies in how to make it adaptively adjust. Strong tracking theory utilizes the residual orthogonality principle of optimal filters, that is, for an optimal Kalman filter, its output residual sequence It should be a zero-mean white noise sequence, meaning the residuals at different times are uncorrelated. When the system model mismatches, the residual sequence will lose its orthogonality. Therefore, by monitoring the characteristics of the residual sequence in real time, the fading factor can be adjusted in reverse to force the residual sequence to remain orthogonal in a statistical sense, thus keeping the filter in a near-optimal state. The calculation method for the fading factor is as follows:

[0049]

[0050] in, The fading factor at time k is used to adjust the prediction error covariance matrix online and force the residual sequences to be orthogonal. The candidate value for the fading factor at time k is calculated from the ratio of the residual covariance; if is the logical judgment condition, tr[ ] is the trace operation of the matrix, representing the sum of the elements on the main diagonal of the matrix; M k The covariance matrix of the theoretical output residual at time k; N k The covariance matrix of the actual output residual at time k can be approximated as:

[0051] in, Let K be the residual vector at time k, and the calculation formula is: That is, the difference between the actual measured value and the predicted measured value. For residual vectors The transpose of ; M k The calculation formula is:

[0052] in, For the observation matrix of the discrete system, Let be the transpose of the observation matrix of the discrete system. Let be the state prediction error covariance matrix (prior covariance) at time k. for The state estimation error covariance matrix at time t (posterior covariance); R is the measurement noise covariance matrix. Let be the state transition matrix of the discrete system. Let Q be the transpose of the state transition matrix of the discrete system, and let Q be the process noise covariance matrix. To represent the calculation of theoretical covariance M k At that time, the fading factor is 1.

[0053] Using the above formula, the filter can dynamically calculate the most suitable fading factor based on the prediction and measurement results at each moment, thus achieving adaptive adjustment of the filter gain.

[0054] It should be further noted that the vehicle dynamics model mainly calculates the self-aligning torque generated by the tire sideslip characteristics and equates it to the force acting on the rack, which can be simplified as follows: F model =f(V, θ) wheel C a ) Where, θ wheel By steering wheel angle θ wheel The wheel angle C is calculated from the steering gear ratio. a It is the tire lateral stiffness, which is affected by tire pressure, load, etc. This functional relationship can be obtained through a simplified linear model and pre-stored in the controller in the form of a function.

[0055] For example, the HWA (steering wheel actuator) mainly consists of a steering wheel, a reduction mechanism, and a road feel motor. It provides the driver's desired steering wheel angle signal to the RWA (wheel actuator), and simultaneously simulates the road force feedback of the vehicle based on the rack force fed back by the RWA, providing the driver with road feel feedback information. The equations of the steering wheel actuator dynamic model are as follows:

[0056]

[0057]

[0058] in, For torque, For rotational inertia, The damping coefficient is... For frictional torque, For the corner, For output torque, To output the moment of inertia, This is the output damping coefficient. To output frictional torque, For measuring torque with TAS, To increase torsional stiffness, This is the transmission ratio.

[0059] The RWA (Rotational Weaving Controller) mainly consists of a steering actuator motor, a reducer, a steering rack, and a steering angle sensor. It receives the desired steering angle command from the HWA (Hydraulic Weaving Controller) and controls the motor to move the rack laterally, thus achieving the steering function. Its dynamic model equations are as follows:

[0060]

[0061]

[0062]

[0063] Among them, T fm For electromagnetic torque, T f For output torque, θ fm For corner, I q Torque current, J fm For rotational inertia, B fm For damping coefficient, For torsional stiffness, K ft For electromagnetic torque coefficient, M r For quality, B r For damping coefficient, x r For rack displacement, F z For rack resistance, G fm For reduction ratio, r p The radius of the pinion is given.

[0064] To ensure the functional safety of the system, a maximum safe threshold for the rack force, F_max_threshold, is set. This value is calculated theoretically or experimentally calibrated based on the structural strength of the steering system and the mechanical limits of the tires. In each calculation cycle, the system makes a judgment using the following example code: IF(F_KF>F_max_threshold)THEN F_final_rack=F_model / / Determine if channel one is faulty, directly use the result from channel two. GOTO Step 3 / / Skip the fusion step and proceed to the next step ENDIF This step ensures that when sensor signal anomalies (such as jumps or drifts) cause unreasonable maxima, the system will not output incorrect forces, but will seamlessly switch to a safe theoretical model.

[0065] If F KF Within a safe range, the results from the two channels are dynamically weighted and fused to leverage their respective advantages under different operating conditions. F final_rack =ωF KF +(1-ω) F model Wherein, the weighting coefficient ω is the vehicle speed V and the steering wheel angular velocity ω sw The function, ω=f (V, ω sw The design principle of the weighting function is as follows: at low speeds (such as parking) and when the steering wheel is turned quickly (such as emergency obstacle avoidance), the value of ω is larger (e.g., close to 1), and it relies more on the STAKF observation value that reflects the real load; at high speeds and stable driving, the value of ω is smaller (e.g., close to 0), and it relies more on the calculated value of the stable vehicle dynamics model.

[0066] S3. Based on the vehicle state parameters, calculate multiple compensation torque components in parallel, including return-to-center compensation torque, friction compensation torque, and inertia compensation torque. It should be noted that in this step: To obtain high precision, the basic rack force F final_rack Then, in order to simulate a complete road feel, the following compensation components need to be calculated in parallel: Furthermore, the return-to-normal compensation torque T return The self-alignment effect of a vehicle is simulated by the following formula; T return =-K return (v)θ sw Among them, K return (v) is the gain coefficient that varies with vehicle speed v. At low speeds, K return(v) is relatively small, the self-centering force is weak, and the steering is easy; at high speeds, K return (v) is relatively large, the self-correcting force is strong, and the steering is stable; For example, friction compensation torque T friction The following formula simulates the damping feel of a traditional steering system; T friction =T coulomb +T viscous =-Csgn(ω sw )-B(v) ω sw Among them, T friction Friction-compensating torque is used to simulate the damping feel of a traditional steering system; T coulomb T represents the Coulomb friction torque components (static friction and sliding friction); viscous ω represents the viscous friction torque component (damping proportional to velocity); C is the Coulomb friction coefficient, representing the inherent frictional force amplitude of the system; sgn() is a standard mathematical symbol function, taking a value of 1 when the variable in parentheses is positive, -1 when it is negative, and 0 when it is 0; sw B(v) is the angular velocity of the steering wheel rotation; B(v) is the viscous damping coefficient that varies with the vehicle speed v, where v is the current vehicle speed.

[0067] Inertia compensation torque T intertia The formula used to simulate the inertia of traditional mechanical components such as steering columns is as follows: T intertia =-J virtual· a sw Among them, T inertia Inertial compensation torque is used to simulate the physical inertia of mechanical components such as the steering column and gears, preventing the steering wheel from turning too quickly or too lightly; J virtual For the set virtual moment of inertia, a sw The angular acceleration of the steering wheel, i.e., the angular velocity ω sw The derivative with respect to time.

[0068] To make the steering wheel return process more natural and guided, the system needs to determine whether the driver intends for the steering wheel to return to center automatically. The determination logic is as follows: IF (|T_drv|<T_threshold_1) AND (|\theta_sw|> \theta_threshold_2) THEN is_returning = TRUE ELSE is_returning = FALSE END IF Among them, Tthreshold1 This is the threshold value of the driver's input torque (e.g., 0.5 Nm), T threshold2 It is a small angle threshold (e.g., 5 degrees). When the driver applies almost no force and the steering wheel is not in the center position, the system determines that the vehicle has entered the straightening state.

[0069] S4. Calculate the soft stop limit torque T based on the steering wheel angle. stop In this step, it is important to note that: To simulate mechanical limit switches, a virtual soft stop is set, and the maximum steering wheel angle is set to θ. max The buffer zone inlet is θ buffer (e.g. θ) buffer =0.9·θ max The example execution logic is as follows: IF (|\theta_sw|>\theta_{buffer}) THEN T stop =-K stop (θ sw -sgn(θ sw ) θ buffer )- D stop ω sw ELSE T stop =0 END IF Among them, K stop It is a very large virtual spring stiffness coefficient, D stop It is a virtual damping coefficient. When the corner enters the buffer zone, the limiting torque will increase sharply, providing the driver with a clear and gentle sense of boundary.

[0070] S5. The basic rack force, return-to-center compensation torque, friction compensation torque, inertia compensation torque, and soft stop limit torque are weighted and fused to generate the final target handwheel feedback torque. Note that the following points should be noted in this step: The final basic rack force T final_rack Converted to basic feedback torque T base : T base =T final_rack· r pinion Where, r pinion It is the equivalent gear radius; Then, perform the overall fusion: T target =ω1T base +ω2T return +ω3T fiction+ω4T inertia +ω5T stop Among them, ω1 to ω5 are the weight coefficients of each item. These weights can be adjusted according to the driving mode (comfort, sport). Specifically, when is_returning is judged to be TRUE in step S4, the weight ω2 of the return torque will be temporarily increased to provide stronger return guidance.

[0071] The controller will ultimately calculate the target handwheel feedback torque T target As a command, the current of the handwheel feedback motor is controlled by the drive circuit, so that it can accurately output the target torque on the steering wheel. This process is executed cyclically in each control cycle (e.g., 5 milliseconds), thereby providing the driver with continuous, smooth, and realistic road feel feedback.

[0072] It should be noted that SBW requires HWA to actively generate the target drag torque to achieve road feel simulation, and the overall target feedback torque T ref It is synthesized from the torque components calculated by multiple sub-modules, and its calculation formula is as follows:

[0073] Among them, T main The basic feedback torque is the core of road feel; T damp This is a damping compensation torque used to improve stability during high-speed driving; T fric Friction-compensating torque, used to create a sense of center; T return The active return torque is used to compensate for the return to center after the driver lets go; these components together constitute the total hand force that the driver ultimately perceives on the steering wheel.

[0074] To achieve a driving experience that is light at low speeds and stable at high speeds, the basic feedback torque is determined by the basic characteristic curve g(F). z ) and vehicle speed gain adjustment K v (v) Composed of two parts, the basic feedback torque design is as follows: Figure 3 As shown; basic characteristic curve g(F) z A linear-like assist method is adopted, namely rack force F. z When the output is very low (in the center zone), it is close to zero, filtering out minor disturbances; after exceeding the threshold in the center zone, it enters the linear growth zone, where the slope, i.e., the steering gradient, directly affects the clarity of road feel; when the rack force is at its maximum, it enters the saturation zone, where the torque growth slows down to saturation, preventing excessive hand force and indicating the control limits; vehicle speed gain adjustment K v (v), K at low speed v (v) Small, achieving easy steering; K at high speeds v (v) Large, amplifies the output of the basic characteristic curve, making it feel heavy.

[0075] Despite T main While it contains most of the self-centering torque, it still suffers from insufficient self-centering under specific conditions such as low speeds. This is addressed by adding an additional self-centering torque T, which is related to the steering wheel angle and vehicle speed. return Optimize low-speed self-alignment performance, such as Figure 4 As shown.

[0076] Traditional steering systems rely on mechanical friction for a sense of center, while SBW (Steering Wheel Steering) has low physical friction and uses algorithms to simulate the frictional torque T. fric To enhance the driving experience, this torque is used to simulate the characteristics of Coulomb friction and viscous friction, and its direction is opposite to steering; to avoid a step-like sensation near zero speed, an arctangent function is used for fitting.

[0077] Among them, T f,max For the maximum simulated frictional torque, c is the coefficient of the slope of the adjustment curve near zero; Damping torque T damp Used to suppress high-frequency vibrations of the steering wheel, its direction is opposite to the steering angle, its magnitude is proportional to the rotational angular velocity, and its damping coefficient K is variable. d Related to vehicle speed and steering wheel torque, it increases with vehicle speed, enhancing high-speed stability; when T sw When T is small (e.g., during straight-line driving), damping is applied to absorb disturbances; when T... sw When the resistance is high (e.g., during an emergency lane change), reduce the damping to ensure a rapid response:

[0078] Among them, K d is a variable damping coefficient.

[0079] Preferably, to verify the effectiveness of the proposed algorithm in the embodiments of the present invention, a CARSIM and MATLAB / Simulink co-simulation platform was also built. CARSIM (a C-class sedan model) was used to simulate the real vehicle dynamics and road environment, and output the real rack force as an evaluation benchmark; Simulink was used to build the SBW system dynamic model, the proposed ST-AKF observer, the standard KF observer for comparison, and the road feel simulation controller; the simulation principle is as follows. Figure 2 As shown.

[0080] To fully verify the performance of the algorithm proposed in this invention under scenarios with drastic changes in road feel information, the simulated road was set as the Trough Wander Road, a bumpy surface, to simulate continuous road impact. Based on this, the following two typical steering inputs were designed: Condition 1: The vehicle speed is set to 30 km / h, and the steering wheel input is a sinusoidal angle of 2 rad / s ±90° to simulate continuous cornering or obstacle avoidance scenarios. The simulation results are as follows: Figure 5 As shown; Condition 2: The vehicle speed is set to 40 km / h. A 60° step turn is input and maintained at 0.45 s to simulate transient response scenarios such as emergency obstacle avoidance. The simulation results are as follows: Figure 6 As shown; Simulation results from the sinusoidal turn show that the macroscopic trends of both estimated curves match the true values. However, at the peaks and troughs, the KF algorithm exhibits significant lag and amplitude deviation, especially at points with large rack force changes. This is because high-frequency disturbances from bumpy roads cause errors in the model prediction, and KF cannot adjust its Kalman gain in time, relying excessively on the inaccurate prediction model. In contrast, the ST-AKF curve almost coincides with the true values, attributed to its adaptive fading factor. When an increase in the deviation between the predicted and measured values ​​is detected (i.e., the residual sequence loses orthogonality), the fading factor λ... k It will increase rapidly, thereby increasing the Kalman gain and forcing the filter to rely more on the latest measurement information, thus achieving fast and accurate tracking of the force signal.

[0081] The step steering simulation results further highlight the advantages of ST-AKF in handling abrupt signals. After applying the steering input at 0.45s, the rack force surges instantaneously. The slope of the estimated curve of the KF algorithm is significantly smaller than the true value, showing a significant delay of about 0.043 seconds and a peak error of over 1000N. In contrast, the response curve of the ST-AKF algorithm closely follows the true value with minimal delay. This fully demonstrates that the strong tracking theory can effectively solve the tracking sluggishness problem of traditional Kalman filtering under abrupt conditions, and is more robust.

[0082] In summary, the ST-AKF algorithm proposed in this embodiment of the invention outperforms the traditional KF algorithm in both estimation accuracy and dynamic response speed under continuously changing and transient impact conditions.

[0083] To further evaluate the dynamic tracking capability of the ST-AKF algorithm under nonlinear friction interference, this embodiment also designed a unidirectional tension and return test. The test simulates the process of the steering rack moving unidirectionally from the center position (0 mm) to the maximum stroke (50 mm) and then smoothly returning to the initial position. This working condition is intended to simulate the typical operation of a driver turning the steering wheel on one side and returning it to the center. During this process, the rack is subjected to not only the elastic restoring force that varies linearly with the displacement, but also the dry friction force and viscous damping force that are unavoidable in the mechanical transmission system. This makes the system exhibit significant nonlinear mechanical characteristics.

[0084] Reference Figure 7The results show the time-domain response of torque tracking. The gray background curve represents the original sensor measurement value, and the red solid line represents the real-time estimation value of the ST-AKF algorithm. As can be seen from the magnified view, the algorithm greatly suppresses high-frequency measurement noise, and the output curve is smooth and stable. At the same time, at the rising and falling edges of the torque change rapidly, the estimated value shows only a very small phase lag in tracking the real signal. This indicates that the algorithm achieves a good balance between filtering smoothing and dynamic response, and can meet the stringent real-time requirements of the steer-by-wire system.

[0085] Reference Figure 8 The study clearly reveals the energy dissipation mechanism of the mechanical system—the hysteresis loop. Since the direction of friction is always opposite to the direction of motion velocity, during the stretching process (upward curve), the system needs to overcome the spring force plus friction, resulting in a large force on the rack. During the return process (downward curve), the friction force opposes the return to center, making the system behave as spring force minus friction, and the force on the rack is significantly reduced. The blue track generated by the ST-AKF algorithm depicts this closed loop, capturing the peak load at 50mm, and more completely reproducing the resistance difference during the stretching and return processes. This proves that the observer has excellent friction decoupling capability and can truly reflect the mechanical feel characteristics of the steering system.

[0086] Furthermore, in order to verify the performance of the proposed strong tracking adaptive Kalman filter (STAKF) algorithm under actual working conditions, this embodiment conducted loading tests with different road stiffness characteristics on a steering test bench; the test controlled the steering rack to move from the center position to 50mm, simulating two typical load scenarios: vehicle turning in place on a dry road surface (high stiffness) and driving on a wet and slippery road surface (low stiffness).

[0087] Reference Figure 9 The gray background curve represents the raw rack force signal directly acquired by the force sensor. Affected by electromagnetic interference from the test bench motor and mechanical transmission vibration, the raw measurement data contains significant high-frequency noise and signal glitches. Especially under high load conditions, the fluctuation amplitude of the sensor signal increases significantly with the increase of torque, which accurately reflects the poor signal quality under real vehicle conditions.

[0088] The estimation results processed by the STAKF algorithm (shown by the solid line in the figure) demonstrate excellent filtering effect and dynamic tracking accuracy. Under high stiffness conditions (red curve), the algorithm successfully identified the equivalent stiffness characteristic of 166 N / mm and accurately tracked the peak load of up to 8300 N at full stroke. Under low stiffness conditions (blue curve), the algorithm also accurately reproduced the low adhesion characteristic of about 3000 N. It is worth noting that the estimation curve effectively filters out sensor measurement noise while retaining the slight dynamic trend that conforms to physical characteristics, without obvious phase lag. The experimental results show that even in real physical environments with low sensor signal-to-noise ratio and large load variations, the STAKF algorithm can still reconstruct the rack force signal stably and in real time, providing high-fidelity data support for the road feel feedback of the steer-by-wire system.

[0089] It should also be noted that, in order to verify the dynamic capture capability of the proposed STAKF algorithm for sudden road disturbances, a step load condition was also designed. During the test, the steering system was in a closed-loop position control state. At t=1.0s, a sudden load of 4000N was applied to the rack through the loading device to simulate the scenario where the front wheel of the vehicle suddenly hits the curb or drives into a deep pothole during driving. At this time, although the rack displacement was controlled and remained stable, the motor current changed drastically due to the anti-interference control.

[0090] Reference Figure 10 This demonstrates a performance comparison between the STAKF algorithm and the traditional Extended Kalman Filter (EKF) algorithm when dealing with abrupt load changes: Ordinary EKF (green curve): Due to its fixed gain matrix, it exhibits obvious hysteresis characteristics when faced with sudden signal changes; within 200ms after a load change, the estimated value can only rise slowly and cannot reflect the real road conditions in a timely manner. This will lead to false and delayed steering wheel force feedback, affecting the driver's emergency judgment.

[0091] STAKF (red curve): The strong tracking algorithm proposed in this embodiment introduces a fading factor to monitor the orthogonality of the residual sequence in real time; when a sudden change is detected at t=1.0s, the algorithm automatically increases the gain weight; the results show that the STAKF estimate completes the convergence of the 4000N step signal within 50ms, and the steady-state error is extremely small.

[0092] Therefore, this experiment strongly demonstrates that the STAKF algorithm can sensitively capture sudden changes in the external environment through current / torque signals under position control systems, solving the contradiction between noise reduction and fast tracking in traditional filtering methods, and ensuring the real-time performance and safety of road feel feedback in the steer-by-wire system under extreme conditions.

[0093] Reference Figure 11The high-frequency response range at the end of the frequency sweep test was extracted. The results showed that when faced with rapidly changing loads of 18-20Hz, the traditional low-pass filter (blue dashed line) exhibited significant phase lag and amplitude attenuation, and could not reflect changes in road force in a timely manner. In contrast, the ST-AKF algorithm (red solid line) maintained a high degree of synchronization with the real signal (black dashed line) with its strong tracking mechanism, proving that it could suppress noise without sacrificing the high-frequency dynamic characteristics of the signal.

[0094] Reference Figure 12 The Bode plot amplitude curve in the figure shows that the ST-AKF algorithm has a wider passband. With -3dB as the reference, the effective bandwidth of the traditional filter is only about 5Hz, which means that the fine texture of the road surface (usually located in 5-15Hz) will be filtered out. However, the amplitude gain of ST-AKF is maintained at around 0dB in the 15Hz range, effectively preserving the richness and fidelity of the road sense signal.

[0095] Reference Figure 13 Since phase delay is a key factor affecting the transparency of steering-by-wire feel, such as Figure 13 As shown, as the frequency increases, the phase lag of traditional filters increases rapidly, which can easily cause the driver to experience a false sense of position or induce human-machine coupling oscillations. In contrast, the phase curve of ST-AKF decreases gently, and the phase lag at 10Hz is only about 1 / 5 of that of traditional filters. This low lag characteristic ensures the real-time nature of road feel feedback, enabling the driver to perceive the contact state between the vehicle and the road surface at the first moment.

[0096] Preferably, this embodiment also includes a static bench test on the handwheel actuator of the automotive steer-by-wire system to evaluate its steering ease, steering stiffness, torque feedback characteristics of the limiting torque, and control algorithm performance. The test involves the driver rotating the steering wheel from zero to its left and right extreme positions at a relatively low, non-constant speed, repeating this cycle multiple times. Data acquisition is directly from the torque angle sensor on the steering column.

[0097] Reference Figure 14From the perspective of torque establishment characteristics in the center area, the system exhibits high steering stiffness. Within a range of ±50° near zero, the torque gradient is established rapidly and clearly. This setting ensures that even in the light-assist mode with greater power assistance, the driver can still obtain a clear sense of center when making minor adjustments to the direction while driving straight, avoiding any play or vagueness in the steering feel. As the steering angle enters the linear operating range of 100~500°, the power assistance characteristics tend to stabilize. The outgoing torque (steering process) is strictly controlled between 2Nm and 2.7Nm, never exceeding the set threshold of 3.0Nm, and the curve shows an extremely weak linear growth trend. This successfully simulates the physical self-centering torque gain generated by the tire slip angle increasing with the steering angle, avoiding the electronic falseness brought by a flat curve.

[0098] In terms of return-to-center performance and limit protection, it exhibits significant asymmetric hysteresis characteristics and software limiting capabilities. The return torque (during the return-to-center process) fluctuates closely along the 0Nm axis with almost no residual resistance. This means that the system applies high-intensity friction compensation or assistance, allowing the steering wheel to return to center extremely lightly after the driver releases their hands, meeting the comfort requirements of low-speed and light-duty conditions. When the turning angle approaches the physical limit, the software limiting function intervenes in time, and the torque increases exponentially, reaching a peak of 12.5Nm at 540°. This virtual force wall not only effectively limits excessive operation by the driver, but also simulates a flexible buffer texture similar to the compression of a mechanical limit rubber sleeve through the damping design of the transition range, achieving a smooth and safe transition from light feel to limit resistance.

[0099] Reference Figure 15 As can be seen, the intercept of the curve on the horizontal axis (lateral acceleration) is small when the vertical axis (steering wheel torque) is zero, meaning that the SR value is distributed in a narrow range close to 0g (within ±0.03g). This means that the SR value reflects the friction level and self-centering ability of the system, and the smaller the SR value without active control, the better the self-centering ability. This test data shows that the mechanical friction and hysteresis of the steering system are small, and the vehicle has excellent passive self-centering performance. This means that after the driver releases the steering wheel, the vehicle can quickly and smoothly eliminate lateral acceleration and resume straight driving, and the steering feel will not show excessive resistance or heaviness.

[0100] Furthermore, the slope of the curve near the lateral acceleration of 0g reveals a significant gradient (SF) in the steering wheel torque as a function of lateral acceleration, exhibiting good linearity. The curve demonstrates high stiffness at the zero point, indicating that when the vehicle is traveling at high speed and subjected to minor lateral disturbances or requires fine-tuning, the steering wheel can quickly establish a clear torque feedback. This tight torque gradient in the center zone assumes a clear sense of the center position, helping the driver to accurately perceive vehicle dynamics, reduce the frequency of corrections at high speeds, and thus improve the vehicle's straight-line stability and driving confidence.

[0101] Reference Figure 16 In the region of the figure where the lateral acceleration reaches about 0.1g, the torque gradient (SOF) does not show a significant attenuation or abrupt change compared to the gradient in the central region, and the overall curve maintains good smoothness and linearity. This indicates that when the vehicle transitions from straight-line driving to small-angle steering (deviating from a straight line), the gain of steering torque is synchronized with the establishment of the vehicle's lateral dynamics. The consistent SOF characteristics show that the vehicle can provide predictable handling feedback in the early stage of cornering, and the driver can intuitively perceive the degree of deviation of the vehicle's trajectory through changes in hand force, demonstrating that the vehicle still has good handling linearity and controllability under non-zero lateral acceleration conditions.

[0102] In application of the above embodiments, other aspects of the present invention also provide a road feel torque simulation system for a vehicle steer-by-wire system, comprising: The parameter acquisition module is used to acquire vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and torque information of the steering actuator motor in real time. The core processing module is used to calculate and generate the target handwheel feedback torque; The torque execution module is used to drive the handwheel feedback motor to apply feedback force to the driver based on the target handwheel feedback torque.

[0103] The system also includes one or more processors and memory.

[0104] The memory is used to store operable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations, including the flow of the road feel torque simulation method for a vehicle steer-by-wire system as described in the foregoing embodiments, particularly... Figure 1 The flowchart of the method is shown.

[0105] Other aspects disclosed in the embodiments of the present invention also propose a computer-readable medium for storing software including instructions executable by one or more computers, which, upon execution, cause the one or more computers to perform operations including the flow of the road feel torque simulation method for a vehicle steer-by-wire system described in the foregoing embodiments, particularly... Figure 1 The flowchart of the method is shown.

[0106] It should be recognized that embodiments of the present invention may be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable storage medium.

[0107] The method can be implemented using standard programming techniques, including a non-transitory computer-readable storage medium configured with a computer program in the computer program, wherein the storage medium is configured such that the computer operates in a specific and predefined manner.

[0108] Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system; however, if required, the program can be implemented in assembly or machine language.

[0109] In any case, the language can be either compiled or interpreted.

[0110] Furthermore, for this purpose, the program can run on programmed application-specific integrated circuits.

[0111] The processes described herein (or variations and / or combinations thereof) can be executed under the control of one or more computer systems configured with executable instructions, and can be implemented by hardware or a combination thereof as code (e.g., executable instructions, one or more computer programs, or one or more applications) that commonly executes on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.

[0112] Furthermore, the method can be implemented in any suitable computing platform, including but not limited to personal computers, minicomputers, mainframes, workstations, networked or distributed computing environments, standalone or integrated computer platforms, or in communication with charged particle tools or other imaging devices.

[0113] Various aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether portable or integrated into a computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein.

[0114] Furthermore, machine-readable code, or parts thereof, can be transmitted via wired or wireless networks.

[0115] When such media includes instructions or programs that combine with a microprocessor or other data processor to implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media.

[0116] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for simulating road feel torque in a vehicle steer-by-wire system, characterized in that, include: Real-time acquisition of vehicle status parameters, which include at least vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and torque information of the steering actuator motor; A fusion estimation strategy is used to calculate the basic rack force; Based on the vehicle state parameters, multiple compensation torque components are calculated in parallel, wherein the compensation torque components include homing compensation torque, friction compensation torque and inertia compensation torque; Calculate the soft stop limit torque based on the steering wheel angle; The basic rack force, return torque, friction compensation torque, inertia compensation torque, and soft stop limit torque are weighted and fused to generate the final target handwheel feedback torque.

2. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 1, characterized in that, The fusion estimation strategy includes: Based on the torque information of the steering actuator motor and the rack displacement, the first rack force is estimated by a strong tracking adaptive Kalman filter observer; Based on the vehicle speed and steering wheel angle, the second rack force is calculated using a vehicle dynamics model. The first rack force and the second rack force are dynamically weighted and fused to obtain the basic rack force.

3. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 2, characterized in that, The weighting coefficients of the dynamic weighted fusion are dynamically adjusted based on the vehicle speed and steering wheel angular velocity.

4. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 3, characterized in that, The dynamic adjustment of the weighting coefficients includes: Increase the weight of the first rack force when driving at low speeds or when the steering wheel is turned quickly. When traveling at high speed, the weight of the second rack force is increased.

5. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 2, characterized in that, The fusion estimation strategy also includes security redundancy judgment, including: Preset a rack force safety threshold; Before weighted fusion, it is determined whether the absolute value of the first rack force exceeds the safety threshold. If it does, the second rack force is directly used as the final base rack force, and fusion is no longer performed.

6. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 1, characterized in that, The return-center compensation torque is proportional to the return-center torque coefficient and the steering wheel angle, and the return-center torque coefficient increases with the increase of vehicle speed.

7. The method for simulating road feel torque for a vehicle steer-by-wire system according to claim 1, characterized in that, The friction compensation torque includes Coulomb friction torque which is opposite to the direction of the steering wheel angular velocity and viscous friction torque which is proportional to the steering wheel angular velocity.

8. A road feel torque simulation system for automotive steer-by-wire systems, characterized in that, include: The parameter acquisition module is used to acquire vehicle speed, steering wheel angle, steering wheel angular velocity, rack displacement, and torque information of the steering actuator motor in real time. The core processing module is used to calculate and generate the target handwheel feedback torque; The torque execution module is used to drive the handwheel feedback motor to apply feedback force to the driver based on the target handwheel feedback torque.

9. The road feel torque simulation system for a vehicle steer-by-wire system according to claim 8, characterized in that, The system also includes one or more processors; The memory stores operable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations, including the flow of the road feel torque simulation method for a vehicle steer-by-wire system as described in any one of claims 1 to 7.

10. A computer-readable medium for storing software, characterized in that: The software includes instructions executable by one or more computers, which, upon execution, cause the one or more computers to perform operations including the flow of a road feel torque simulation method for a vehicle steer-by-wire system as described in any one of claims 1 to 7.

Images