Estimation And Experimental Research On The Steer-by-wire Vehicle States And Road Parameters Based On Information Fusion

Information fusion, an automated comprehensive information processing technology, has had significant development in the last decade. The technology provides more comprehensive, reliable and accurate signal source through the integration of real-time, redundant or complementary signals provided by multiple sensors, so as to guarantee the enhancement of system performance. On the other hand, along with the rapid development of electrical-electronic technology and the micro-processing technology,Steer–by-wire, as a new generation of automotive mechatronics control technology, has been increasingly attracted attentions from the world's major automobile manufacturers. One of the key technologies for practical real-time access of Steer–by-wire vehicle to motor sport is the motion state and road status information. As the number of vehicle states and road condition information (such as side slip Angle, road tire friction coefficient, etc.) are difficult and costly to obtain accurately. Researches on how to obtain accurate state of vehicle and road state estimation algorithm and put it into practical use for steer–by-wire vehicle have important theoretical and engineering significance.Based on the steering control information fusion theory, this paper proposed to use a decoupled extended Kalman filtering technique to achieve steering control of steer–by-wire vehicle and accurate estimation the state of motion and road tire friction coefficient, which has been broadly tested by a test-bed. The main contribution of this paper are as follows:1,steer–by-wire vehicle steering control Control StrategySteering performance is an important evaluation criterion of steer–by-wire vehicle performance, it directly affects the handling and stability capability of steer–by-wire vehicle. To ensure safe driving of vehicles, reducing traffic accidents plays an important role. Due to the cancellation of the fixed connection between the steering wheel and front wheel steering mechanism (ie, steering ratio is no longer a fixed value), steering ratio is changed to ensure the yaw rate gain G^r_δsw on the steering wheel angle to be always a fixed value. The steer–by-wire simulation test-bed test verification shows that when G^r_δsw is 0.5, the driver has the best subjective feeling, and can adapt to a variety of speed and driving conditions. Based on this theory, the yaw rate and lateral acceleration feedback steering control strategy are integrated to verify different road adhesion coefficient respectively. Testing results show that the use of fixed gain of yaw rate and yaw rate and lateral acceleration feedback control can improve the steer–by-wire vehicle control and stability, and reduce driver's burden.2,The application of information fusion on estimation algorithm of steer–by-wire vehicle status parameters and road tire friction coefficientOn steer–by-wire vehicle steering control strategy, various types of state parameters of vehicles and roads are needed to be estimated. This paper adopts information fusion technology,in using of Kalman filter theory for rapid simulation and estimation of these parameters. Using different vehicles model and tire model, the classical Kalman filter, extended Kalman filter, the decouple extended Kalman filter recursive estimation models are established and verified. Experimental results show that dynamic model based on three degrees of freedom, using the HSRI tire model of a decouple extended Kalman filter, not only accurately estimates the vehicle state parameters, but also estimates the road tire friction coefficient in real-time. Road tire friction coefficient is difficult to measure in practice, but it is essential to control steer–by-wire vehicle. When there are changes on road, steer–by-wire vehicle control strategies must be changed accordingly. In the decouple extended Kalman filter, two recursive state and parameter estimation model exists in parallel, while they are dependent on each other, and has real-time interaction correction to forecast information, which quickly converges towards estimated true value for simulation. The accurate estimation of decouple extended Kalman filter theory in the vehicle state and road information make it possible that some of the parameters can be estimated, which is proved to be difficult to obtain, and it also provides the necessary conditions for steer–by-wire vehicle steering control strategy. In the meantime, the validity and feasibility of this algorithm have been verified by steer–by-wire simulation and the real vehicle test-bed venue.3,Steer–by-wire simulation test-bed software and hardware designIn this paper, a steer–by-wire simulation test-bed is proposed, which is composed of two parts, namely test-bed hardware and software. For software designation, real-time computing power steering models, data acquisition system and control the caller are proposed. For hardware designation, real steer–by-wire system, drivers operating systems, instrument show the simulated system are proposed. Steer–by-wire simulation test-bed is developed for the driver in order that a real operating environment is provided, in which the estimation results of parameters and state is more similar with the actual vehicle condition. This paper focuses on the simulation test-bed signal acquisition and control methods and drive control algorithms, including stalls, pedal displacement, ignition, steering, steering wheel angle etc; analog, digital acquisition card and timing card driver control algorithm preparation. From this research, a real driver, steer–by-wire system hardware and real-time dynamic model of the organic integration are achieved. The entire test-bed provides a person - Vehicle Hardware-in-loop simulation platform, on which experiments can be carried out without danger to any limits of environmental conditions. Sufficient preparatory works have also been done for the real vehicle field test4. steer–by-wire vehicle status parameters and the road tire friction coefficient estimation algorithm validationUsing the theory of information fusion technology to model the simulation, this paper respectively controlled steer–by-wire simulation test-bed and give real vehicle field test validation. During the steer–by-wire simulation test-bed, we propose to use Carsim vehicle dynamics model and the matlab built under the Simulink simulation model state estimation, real-time interactive vehicle status information, the vehicle and road condition are then estimated. Test-bed experimental results show that the algorithm is able to precisely estimate the speed, centroid lateral angle, road tire friction coefficient and other state parameters. The validity and accuracy of the proposed estimation algorithms are also verified by field test using real vehicle data.The novelty of this paper is that we propose to use decoupled extended Kalman filter theory to accurately estimate the vehicle state parameters and road tire friction coefficient, the proposed estimation algorithm is able to estimate the state parameters, including vehicle speed, lateral acceleration, yaw rate and the road tire friction coefficient etc. Estimation results are applied to steer–by-wire vehicle steering control feedback strategy to improve the active operation safety of steer–by-wire vehicle. This paper also designed a steer–by-wire simulation test-bed software and hardware equipment, which is proved to be a good platform for fugure research on steer–by-wire vehicle theory. We design a set of wheel speed signal acquisition methods, which is proved to able to accurately capture the vehicle speed signal. The proposed research achievements lay the foundation for the theoretical and practical development of steer-by-wire vehicle in the future.