Robust Transformer-Enhanced Tire Force Estimation: Combining Model-Based Observer and Deep Learning

Many vehicle state parameters will be used in the electronic stability control (ESC) system. ESC system must estimate some important parameters based on a few sensors and help to develop the control strategies. ESC can estimate the work state of vehicle with tire forces based on the estimating of adhesion coefficient and wheel slip ratio, then determine the control strategy on wheels. The tire forces to be estimated include vertical force, lateral force and longitudinal force on tires. In this article, a method of estimating vertical tire force based on vehicle geometric parameters and vehicle working state, a method of estimating tire lateral force based on vehicle yaw rate and the change of tire longitudinal forces caused by the change of vertical forces and lateral forces have researched. Co-simulation shows that the trend of the tire forces changing is correct and small error on high and low adhesive pavement. The ESC system is available on HIL bench and testing vehicle, so the estimates of tire forces are practical methods.

Road friction coefficient is a key factor for the stability control of the vehicle dynamics in the critical conditions. Obviously the vehicle dynamics stability control systems,including the anti-lock brake system(ABS),the traction control system(TCS),and the active yaw control(AYC) system,need the accurate tire and road friction information. However,the simplified method based on the linear tire and vehicle model could not obtain the accurate road friction coefficient for the complicated maneuver of the vehicle. Because the active braking control mode of AYC is different from that of ABS,the road friction coefficient cannot be estimated only with the dynamics states of the tire. With the related dynamics states measured by the sensors of AYC,a comprehensive strategy of the road friction estimation for the active yaw control is brought forward with the sensor fusion technique. Firstly,the variations of the dynamics characteristics of vehicle and tire,and the stability control mode in the steering process are considered,and then the proper road friction estimation methods are brought forward according to the vehicle maneuver process. In the steering maneuver without braking,the comprehensive road friction from the four wheels may be estimated based on the multi-sensor signal fusion method. The estimated values of the road friction reflect the road friction characteristic. When the active brake involved,the road friction coefficient of the braked wheel may be estimated based on the brake pressure and tire forces,the estimated values reflect the road friction between the braked wheel and the road. So the optimal control of the wheel slip rate may be obtained according to the road friction coefficient. The methods proposed in the paper are integrated into the real time controller of AYC,which is matched onto the test vehicle. The ground tests validate the accuracy of the proposed method under the complicated maneuver conditions.

This paper presents a motion control method of 6WD/6WS (6-wheel-drive and 6-wheel-steering) vehicle using optimum tire force distribution method. When we make used of independent driving and steering system, the motion control performance of the 6WD/6WS vehicle can be improved. For example, vehicle's maneuverability is more improved at low speed and vehicle's stability is also improved at high speed. Therefore, it needs to find a control strategy using optimum tire forces. A control strategy must satisfy the driver's command and minimize the energy consumption. When the driver commands vehicle (input steering angle and acceleration/brake pedal stroke), we can obtain the desired yaw moment, the desired lateral force and the desired longitudinal force. Those values are used in the optimum tire force distribution method. The optimum tire force distribution method can find the longitudinal tire force and lateral tire force. That force minimizes the cost function. The cost function is the sum of normalized tire force. The longitudinal tire forces and lateral tire force of each wheel are converted to the reference torque inputs and steering wheel angle inputs. The method was tested on the simulation and effectiveness was verified.

The knowledge of tire-ground interaction forces is valuable for intelligent vehicles technologies. However, tire forces transducers are expensive and not suitable for ordinary passenger cars. An alternative is to estimate these forces using less sophisticated sensors. This paper presents a new delayed interconnected cascade-observer structure to reconstruct the forces acting on each tire in all directions. The addition of delayed interconnections overcomes the mutual dependence problem in the cascade estimator. Observers are developed based on nonlinear vehicle dynamic models. The Extended Kalman Filter and the Unscented Kalman Filter algorithms are applied for comparison purposes. Two experimental data are used to validate the estimator: a driving in ordinary urban streets and a high-speed slalom maneuver in a banked road. The results are compared with force measurements obtained with transducers and with an existing estimator from the literature.

This paper proposes an automated vehicle trajectory following system that uses four constrained wheel torques to regulate a vehicle on a reference trajectory. The constrained wheel torques can be achieved using the two-wheel drive and differential brakes. The proposed control algorithm is developed using the following steps. First, the sliding-mode control is used to find stability constraints for trajectory following when the vehicle system is subjected to modeling errors. Second, these stability constraints, along with other actuator constraints, are particularly tuned for the proposed control distribution method. The proposed control distribution method determines four longitudinal tire forces and minimizes actuator control efforts. Finally, these tire forces are converted to traction/braking wheel torques. The proposed method has the following advantages: 1) It achieves both robust trajectory following and optimal control efforts, 2) the optimal control effort is obtained analytically instead of from a numerical search, and 3) the robust performance of this vehicle control system can be theoretically verified. The proposed method is evaluated using numerical simulations on two front-drive vehicle models: a full-state vehicle model and a sedan model from the Carsim commercial software. The simulation results indicate that, in both cases, the proposed method can regulate the vehicle to finish a “double-lane change” when the vehicle is moving at an initial speed of 90 km/h. The maximum lateral acceleration is 6.56 m/s <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> , and the regulated position error is less than 6.9 cm.

In order to prevent severity traffic accident on freeway caused by tire blowout, it is necessary to monitor the tire pressure in real time. This paper established a linear model between the force and the slip to estimate the tire radius through least square ,where only ABS wheel speed sensors were used. The result of the simulation about this method is perfect ,it is a effective method for indirect tire pressure monitoring.

The effect of vehicle active safety systems is subject to the friction force arising from the contact of tires and the road surface. Therefore, an adequate knowledge of the tire-road friction coefficient is of great importance to achieve a good performance of these control systems. This paper presents a tire-road friction coefficient estimation method for an advanced vehicle configuration, four-motorized-wheel electric vehicles, in which the longitudinal tire force is easily obtained. A hierarchical structure is adopted for the proposed estimation design. An upper estimator is developed based on unscented Kalman filter to estimate vehicle state information, while a hybrid estimation method is applied as the lower estimator to identify the tire-road friction coefficient using general regression neural network (GRNN) and Bayes' theorem. GRNN aims at detecting road friction coefficient under small excitations, which are the most common situations in daily driving. GRNN is able to accurately create a mapping from input parameters to the friction coefficient, avoiding storing an entire complex tire model. As for large excitations, the estimation algorithm is based on Bayes' theorem and a simplified “magic formula” tire model. The integrated estimation method is established by the combination of the above-mentioned estimators. Finally, the simulations based on a high-fidelity CarSim vehicle model are carried out on different road surfaces and driving maneuvers to verify the effectiveness of the proposed estimation method.

This paper presents a predictive vehicle directional stability control structure that has integrated energy-loss reduction benefits during transient handling maneuvers. The method is based on the idea of balancing longitudinal and lateral tire force saturation levels using a cascade model predictive structure for the optimal distribution of tractive or braking torques. Balancing saturation levels also has the added benefit of reducing and evening-out tire wear. To demonstrate the energy-loss reduction benefits, we consider nonlinear simulations of a nominally unstable truck featuring an independent drive system. Comparisons against a commonly cited brake-based yaw stability control strategy with similar directional control performance shows that the proposed predictive saturation management approach provides energy-loss reductions of more than 60%. This energy efficiency benefits are retained whether or not the drive system has regenerative/energy recovery capabilities.Copyright © 2012 by ASME

Reeds–Shepp (RS) curves are widely employed for path smoothing in path planning algorithms such as Hybrid A* and Rapidly-Exploring Random Trees (RRT). To ensure the stability of autonomous vehicles when traversing RS-curve paths, this paper proposes a velocity prediction and planning algorithm based on vehicle lateral dynamics. Specifically, given the path curvature and initial velocity, the yaw rate and side-slip angle of the vehicle during steady-state steering are derived using the lateral dynamics model. Considering load transfer effects, the lateral tire force is calculated from the rate of change of vertical loads, and the safe driving velocity is predicted through gradient descent with the tire force saturation coefficient as the optimization objective. Subsequently, a velocity planning method that integrates the Gaussian pseudospectral method (GPM) with sequential quadratic programming (SQP) is developed. The problem is formulated as a nonlinear programming model, in which the predicted safe velocity serves as the upper velocity constraint. Joint simulations in MATLAB/Simulink and CarSim are then conducted to evaluate the proposed framework. The results show that the prediction accuracy of the lateral tire force exceeds 94% under different road friction coefficients, confirming the reliability of the proposed prediction model. Furthermore, comparative experiments demonstrate that the velocity planning method can effectively reduce vehicle velocity from unsafe to safe levels, thereby ensuring driving stability and safety when negotiating curves.

The knowledge of individual torques at each wheel are important intermediate variables for the design of advanced driver assistance systems. A nonlinear high-gain unknown inputs state observer is designed to compute simultaneously longitudinal and vertical tyre forces, resultant torques applied to wheels, and the vehicle speeds with a very simple four wheels model. The necessary measurements are the usual rotation speed of each wheel. Successful simulated experimentations of the method are shown as well as its present limits.

This paper describes a unified chassis control (UCC) strategy to improve the lateral stability andmanoeuvrability of vehicles by integrating individual chassis control modules such as electronic stability control (ESC), active front steering (AFS) and continuous damping control (CDC). In order to achieve a target lateral vehicle response, an integrated AFS and four individual wheel braking controls have been used for an optimum distribution of longitudinal and lateral tyre forces the desired yaw moment for lateral stability has been designed by the sliding control method using a planar bicycle model and taking into consideration cornering stiffness uncertainties. The desired yaw moment is generated by the coordinated control of AFS and ESC. Optimal coordination of the control authority for the AFS and the ESC has been determined to minimise longitudinal deceleration. Estimated vertical tyre forces have been used for the optimum distribution of the longitudinal and lateral tyre forces. For the improved performance of the lateral stability control system, the damping forces at the four corners have been controlled to minimise roll angle by the CDC system. The response of the vehicle to the UCC system has been evaluated via computer simulations using vehicle dynamic software CarSim and a UCC controller coded with Matlab/Simulink. Computer simulations of a closed-loop driver–vehicle–controller system subjected to double lane change have been carried out to prove the improved performance of the proposed UCC strategy over a conventional ESC.

The motion of a vehicle is governed by the forces generated between the tyres and the road. Knowledge of these vehicle dynamic variables is important for vehicle control systems that aim to enhance vehicle stability and passenger safety. This study introduces a new estimation process for tyre/road forces. It presents many benefits over the existing state-of-art works, within the dynamic estimation framework. One of these major contributions consists of discussing in detail the vertical and lateral tyre forces at each tyre. The proposed method is based on the dynamic response of a vehicle instrumented with potentially integrated sensors. The estimation process is separated into two principal blocks. The role of the first block is to estimate vertical tyre forces, whereas in the second block two observers are proposed and compared for the estimation of lateral tyre/road forces. The different observers are based on a prediction/estimation Kalman filter. The performance of this concept is tested and compared with real experimental data using a laboratory car. Experimental results show that the proposed approach is a promising technique to provide accurate estimation. Thus, it can be considered as a practical low-cost solution for calculating vertical and lateral tyre/road forces.

Longitudinal and lateral forces acting on tyres are known to be closely related to tractive ability, braking characteristics, handling stability and manoeuvrability of ground vehicles. However, it is not feasible in the operating vehicles to measure the tyre forces directly because of the high cost of sensors, the limitations in sensor technology, interference with tyre rotation and the harsh environment. Another approach is the indirect monitoring technique that can estimate the tyre forces based on remote sensor outputs and vehicle dynamics models. In this paper, in order to develop tyre force-monitoring systems, a monitoring model is proposed utilizing not only the yawing motion but also the roll motion. Based on the monitoring model, a monitoring system is designed to estimate the lateral tyre force acting on each tyre. The monitoring system is constructed on the basis of a new scaled Kalman filter with model error compensator (SKFMEC) technique that is developed in this study to improve the robustness performance of Kalman filter methods. The SKFMEC technique adopts both the well-conditioned observer and the model error compensator concepts. Tyre force estimation performance of the monitoring system is evaluated in the MATLAB simulations where true tyre force data are generated from a 14-degree-of-freedom vehicle model with the combined-slip ‘magic formula’ tyre model.

Road friction information is very important for vehicle active braking control systems such as ABS, ASR, or ESP. It is not easy to estimate the tire/road friction forces and coefficient accurately because of the nonlinear system, parameters uncertainties, and signal noises. In this paper, a robust and effective tire/road friction estimation algorithm for ABS is proposed, and its performance is further discussed by simulation and experiment. The tire forces were observed by the discrete Kalman filter, and the road friction coefficient was estimated by the recursive least square method consequently. Then, the proposed algorithm was analysed and verified by simulation and road test. A sliding mode based ABS with smooth wheel slip ratio control and a threshold based ABS by pulse pressure control with significant fluctuations were used for the simulation. Finally, road tests were carried out in both winter and summer by the car equipped with the same threshold based ABS, and the algorithm was evaluated on different road surfaces. The results show that the proposed algorithm can identify the variation of road conditions with considerable accuracy and response speed.

Internet of Things (IoT) based healthcare monitoring system is becoming the present and the future of the medical field around the world. Here the monitoring system acquires the regular health details of hospital discharged patients like elderly patients, patients out of critical operations, and patients from remote areas, etc., and transmits it to the doctors. But the system is highly susceptible to sensor faults. Hence a data-driven hybrid approach of Hidden Markov Model (HMM) based on baum-welch algorithm with Support Vector Machine (SVM) is proposed to predict the abnormality caused by the medical sensors. The proposed work first perform the abnormality detection on the sensor data using the HMM based on baum-welch algorithm in which the normal data is separated from abnormal data followed by classifying the abnormal data as critical patient data or sensor fault data using the SVM. Here the proposed work efficiently performs fault diagnosis with an overall accuracy of 99.94% which is 0.59% better than the existing SVM model. And also a comparison is made between the hybrid approach and the existing ML algorithms in terms of recall and F1-score where the proposed approach outperforms the other algorithms with a recall value of 100% and F1-score of 99.7%.