| In recent years, "Telematics" and "Advanced Driver Assistant System" had been the two hot points of vehicle industry. The former was focusing on the development model of internet combining with vehicles, while the latter was focusing on the intelligence of driver assistant function. The vehicle mobile application was one of the typical researches of internet combining with vehicles. Although relating vehicle Android applications had been marketing, the application which could achieve the driver assistant functions still had not had the mature product. This paper took some researches on vehicle roll-over and yaw danger precaution with Android mobile phone, which expected to have some explorations for the relating researches.(1)This paper introduced the overall design scheme of Android application based on driver assistant function, which was under the background of driver assistant systematization achievement. The researches were took from two design layers, which were the sensor layer and programming layer. It could achieve two main functions which were online monitoring and posture danger precaution.(2)The theoretical models of roll-over and yaw danger were established, and then the auto-regressive predicting model was studied from three aspects, which were predicting principle, order selection principle and predicting process. In roll-over precaution part, it chose the 4-DOF vehicle model and lateral-load transfer ratio as the judgment index. In yaw danger precaution part, it chose the 2-DOF vehicle model and the ratio of difference and threshold value of actual and ideal yaw rate as the judgment index.(3)In the sensor layer design part, the sensor network was proposed, which was to gather the vehicle running information through arranging the sensors in multiple places on the vehicle. This sensor network had two kinds of sensor sources which were the vehicle motion sensor and OBD model. They were arranged at four positions on the vehicle which were the centroid position of vehicle, the steering wheel position,the front wheel position and the OBD slot. It could collect three kinds of information which were the vehicle posture information, engine running information and driverbehavior information. The vehicle motion sensor was designed by ourselves which contained the MPU-6050 motion model and the ARM kernel chip, integrated the algorithm of four elements and Kalman filtering. It could output the acceleration,angular speed and posture angles.(4)The Android programming part had finished under the Eclipse development environment. The wireless communication of Bluetooth, data analysis of sensors, AR predicting model, roll-over and yaw danger precaution, data storage, SOS message remote sending, UI design of chart style and other functions had achieved in this programming part. The Android thread mechanism was used in the programming,because it could achieve the background operation. According to the object-oriented mechanism, different class files were programming, which could provide convenience for afterwards extension.(5)In order to improve the accuracy of data acquisition and posture danger precaution, calibration test for sensors and real vehicle experiment were designed in this paper. The angular velocity and acceleration could be calibrated by "Turntable Test", which could get the error compensation model. The real vehicle experiment contained the roll-over and yaw danger precaution experiments, which were tested under different typical conditions. The error analysis was concluded through the contrast of precaution results. Through the test verification, this application could not only run stably, but also have good precaution effect, which had generally achieved the expected effect. |
PDF Full Text Request