Research On Data Updates And Optimization In Vehicle Control System

Currently, automotive extensively use embedded chip, control bus, control software to control and implement some functional parts of the car. Control systems in automotive are complex distributed systems. In modern car, there have been installed many intelligent sensors. At the same time, the number of information, which needs to be handled, is predicted to increase with7%every year. Vehicle control system is also a typical real-time system and needs to manage a large number of real time data, so these make how to maintain data consistency in real time database as a key issue of real time database.A large amount of data and transactions in real time database of the control system have the real time requires, and the data also have relationships of calculating a new data item value from several other data items. When control systems use special real time database to store and manage data, the real time database needs to maintain data freshness and keep applications effective. This thesis focuses on issues more systematically and in-depth study, analysis and survey the requirement of data management in vehicle control system. An improved dynamic on-demand update approach for maintaining data freshness in real time database of vehicle control system is proposed. At the same time under the extending of data dependency graph in multiple electronic control units, the thesis also surveys how to reduce the transfer data among several electronic control units, and maintain data freshness. The main contributions of the thesis are as follows:(1) The vehicle control systems require real-time database technology to meet the needs of real-time data and consistency for data management. Current research in this field can trigger many data updates transactions, and that will increase database overhead. To solve the above problem, this paper proposes an improved on-demand updates algorithm.(2) The paper also extends data dependency graph to multiple electronic control units. A practical On-Demand updates approach based on extended data dependency graph for maintaining data freshness is proposed. It can convert periodic data transmission to on-demand updates through evaluating the workload of real time database. The approach can reduce the number of transfer data in CAN network. (3) We have analyzed the architecture and functionality of the body control system, and study the requirement of using real-time database for data management. And we have developed a simulated real time database to implement the functionality of data management in the simulation of body control system. We also show that these two proposed on-demand updates methods are effective in the simulation test.