Research Of Fast Algorithm For Real-time Cross-correlation Based On Control Unit Of Traffic Flow

1. Significance of studyWith the economic swift and rapid development of the urbanization process, people's traffic demand runs high continuously. With the increase which has caused by the city traffic demand and supply contradiction in a large amount of private car, the present traffic jam phenomenon in many cities has already become a " bottleneck " of city development.Many practice indicate: the effective outlet of solving the city traffic problem does not increase only through the road construction and the function of traffic administration is more outstanding after the city develops into certain scale. The intellectual traffic transportation system (ITS) appeared at the beginning of the 1990s. Through the real-time collection, transmitting and dealing with the relevant traffic information, we can hold the traffic operation conditions at present and predict the future traffic. With the support from many kinds of means and equipment, various of traffic situations were solved. The intensive development of the communications and transportation has already been acknowledged as the best way to utilize the space-time resource of traffic.At present, many road-information Monitor have different shortcomings. For example, it is easy to be influenced by the weather factor and inability to supply a closed-loop control. This thesis puts forward the concept of the traffic electronic control unit in order to carry on a method with effective economy to gathere and dealt with the flow of traffic data. And it affords a method based on Fast Algorithm for Real-time Cross-correlation.2. Universal Project of the driving systemCross-correlation function can be used for confirming transmission time when a signal was processed in the system in the project practice (Fig.1). Utilize this principle to realize the measurement to the speed of fluid. Suppose the fluid flowsin sealing the pipeline, can think briefly , the fluid in every place is the same in speed on the section. Suppose average speed as Vc, which flow with the fluid separately two vertical directions in the section. The distance between them is L. It is T that the fluid passes two sectional time intervals. Thus, we can get the conclusion from formula: Vc=L/T. The electronic controller unit of the traffic actually requires the information assurance offered.Main content of the article include a research of Fast Algorithm for Real-time Cross-correlation and a realization with the hardware on this base. The hardware is made up of two parts: DSP and data acquisition. (Fig.2).3. The software and hardware realization of the whole systemThis text has been discussed from two main respects: Feasible principle of making operation fast of cross-correlation algorithms, Namely the realization of: Software; the speed question of he circuit reflects, Namely the realization of: hardware.The realization of: Software: At first, estimate the cross-correlation peak value position roughly, then carry on accurate peak value to calculate and search for according to its position. Thus we can reduce the number of cross-correlation coefficient greatly.Secondly the consideration of the real-time performance on the hardware, there are many multiplication operation and addition operation in operation. In order to complete many operations in the interval sampling, The processor must have the higher speed of operation. In the paper, there is a processor (P89C668): Speed up to 20 MHz with 6 clock cycles per machine cycle(40 MHz equivalent performance); up to 33 MHz with 12 clocks In-System Programmable. RAM expandable externally to 8kbyte. Two sample-holding circuits are controlled by the same signal, (one order control all the sample-holding circuits). Thus we can get the sampling with the same time. A/D574 is selected for it provide 35 us Maximum Conversion Time and provide four calibrated ranges: 0 volts to +10 volts and 0 volts to +20 volts unbipolar. -5 volts to +5 volts and -10 volts to +10 volts bipolar, the precision, laser-trimmed scaling and bipolar offset resistors.4. Conclusions(1) This text has set up a real-time cross-correlation based on "conditional cross-correlation". And design program according the hardware circuit. The rang of measuring is designed from 15km/h to 225km/h and the precision is 0.5%. It provides a method to measure the traffic flow using simple electronic device.