The Research On The Estimation Method Of Vehicle Speed Based On FPGA

There is a sustained growth in several aspects of the automotive electronics field today, including the type update of the cars, the gradually shorten average life of the motorcycle type, as well as that the reasons for change is not in a performance degradation, but consumer preferences . The reasons to promote the development of automotive electronics are: (1) Technology. The cost of components decreases with the advances in semiconductor technology. (2) Market competition. More and more car manufacturers will take electronic devices as part of its competitive edge or weapons. (3) Performance. The electronic products can be used to optimize the fuel consumption and improve engine performance. (4) Regulatory requirements. It is required that electronic devices used in the ignition and engine control systems must be able to contribute to reduce emissions. (5) Security. Security features such as airbags, ABS systems and emergency call system have now been tools to open up the market.On traditionally, automotive Engineers has been dependent on microcontrollers (MCU) and customized ASIC products to achieve and control automotive electronic systems, as well as expand the functions of automotive electronics for each generation. However, the engineers are forced to find other technologies such as the FPGA with low-cost, low power consumption and high reliability, as the increasing number of components, the increasing pressure for product introduction to the market, and the increasingly higher performance requirements. FPGA delivers higher performance and more features (such as I / O, programmable logic, etc.) to automotive designers compared wit MCU. Similarly, FPGA provides lower cost and higher flexibility compared with the ASIC products. Different from ASIC, FPGA can also be used in a variety of programs or projects to help designers for the largest time and resources related to the vehicle qualifications once completed a detailed qualification process. Based on all kinds of reasons, analysts of Gartner Dataquest market research firm believe that FPGA will be the fastest growing segments of the semiconductor in the automotive electronics industry, and the annual compound growth rate will be over 70% in 2007. Designers have been aware that there is a more positive advantage to use the standard FPGA than ASIC. For example, FPGA designers can make changes after the design is completed. In fact, the products put into use can also be upgraded, and will not cause serious accreditation problem. Manufacturers are reluctant to take risks in market environment with increasing pressure to the product development, so the FPGA is the ideal solution. This article is a preliminary study about vehicle speed estimation method based on FPGA in this environment.Main research contents are as follows:1. Introduce the structure, working principle and configuration model of the FPGA, as well as FPGA design software DSP Builder, Modelsim simulation software and Quartus II software briefly.2. Establish a eight freedom car model under the tire attached limits. This paper refers to the model derivation of the non-linear non-steady-state tire mechanical model under the vertical slide - side and the joint condition and eight degrees of freedom vehicle model in State Key Laboratory "vehicle dynamic simulation" of Jilin University. We build a unified tire model and eight degrees of freedom vehicle model (does not include the drive and braking system) in Simulink, which provides a good foundation for the estimated speed of the model in the basic conditions.3. The basic theory and the realization of the speed estimation algorithm. In this chapter we first introduces the simple classical Kalman filter theory and algorithms, and then use Kalman algorithm to estimate speed on the basis of previous work. Kalman filter algorithm is a state equation which introduces Kalman filter through the motor vehicles dynamic model, based on the establishment of the eight degrees of freedom vehicle model in Chapter III.4. Verify the basic processes and simulation algorithm of speed estimate based on FPGA. In the beginning of this chapter, we build a Kalman filter model using Altera DSP Builder Blockset in the Simulink, and do the model simulation of Simulink algorithms. After the simulation, we get VHDL code for specific FPGA chips with SignalCompiler, and make the design turn to the hardware. Then do functional simulation using Modelsim to verify the correctness of VHDL code; Input sinusoidal, step and two-shift angle lines on the steering wheel angle for hybrid simulation. From the simulation results, it's well in most conditions, but the estimated error is too big in certain conditions which can not meet the requirements of accuracy of the stability control system.This paper has some innovations, where we propose speed estimation method based on FPGA that change the test platform for study, though the speed estimation method is based on previous models and algorithms. FPGA has the built-in programmable flexibility and ability to repeat the editing, so that it doesn't take us huge cost for the hardware system be upgraded to meet system requirements. Then extend the life of vehicle electronic devices platform, when System requirements change dramatically.This article has only considered pure shift condition (shift process of vehicle did not drive and brake control). But steering and braking is at the same time in reality, so further research needs to be done.The study of the speed estimate with FPGA is only preliminary, as a result of the limited learning time. And I could only build linear and low-dimensional models in Simulink with limited knowledge of Altera DSP Builder Blockset module. Finally the accuracy of the simulation is not very good because of some issues such as parameter settings. So I will to try to use the new algorithm and optimize the parameter settings to achieve the desired results in the future.