Modeling,Design And Implementation Based On Self-adaptive Traffic Signal Scheduling Algorithm

With the fast development of economy, vehicles are increasing in a rapid way.The problem of heavy traffic is becoming more and more grievous.The technology of traffic signal control is one of the effective ways to improve transitable capacity in the intersection.Currently, timing distribution strategy is adopted in traffic light control, however, this strategy can-not do real-time adjustment of control signals based on roadway information. Traffic signal control based on self-adaptive traffic signal scheduling algorithm can solve this problem to some extent.This dissertation mainly studies modeling,design and implementation based on self-adaptive traffic signal scheduling algorithm. The main research work is as follows:1.Self-adaptive Traffic Signal Scheduling Algorithm.This dissertation proposes an improved self-adaptive traffic signal scheduling algorithm to solve the problem of timing distribution strategy,describes the algorithm’s execution flow and corresponding pseudo-code and gives an example to illustrate the difference of the vehicles passage rate before and after the improved algorithm.2.Research on Modeling Strategy of Traffic Signal Scheduling Algorithm. Ptolemy Ⅱ platform is choosed after comparing current modeling platforms based on the characteristic of self-adaptive traffic signal control system, Ptolemy Ⅱ actor-oriented modeling method of hierarchically heterogeneous is also expounded, characteristic and operational principle are introduced about actor,channel and models of computation.Finally,theoretical guidance is provided to achieve traffic signal control system’s modeling.3.Research on Real-time Event Storage Strategy of Traffic Signal Scheduling Algorithm.Real-time events of traffic signal control system are stored in dynamic calendar queue, and dynamic calendar queue is often adjusted based on the amount and distribution of events. However, this leads to frequent reconstruction of calendar queue and degradation of system’s performance. This dissertation proposes a new method of dynamic calendar queue based on future event list(FELDCQ) to solve those problems mentioned above. Events with current model time are stored in dynamic calendar queue, while events with future model time are stored in future event list. Therefore, total stored events in dynamic calendar queue are greatly reduced,and system’s performance is improved based on reduced rescontruction.The strategy of event storage and implementation is outlined and illustrated as an example in the end.4.Research on Real-time Concurrency Processing Strategy of Traffic Signal Scheduling Algorithm. This dissertation proposes a strategy based on correlation matrix,model delay and network delay to set node’s time window, so that the determinism problem of real-time event concurrency processing can be solved to some extent.Finally, logical consistency and execution efficiency are analyzed.5.Modeling Processing and Simulation Design of Traffic Signal Scheduling Algorithm.The system-level behavior of traffic signal control system is described based on self-adaptive traffic signal scheduling algorithm and is divided into independent system modules.Module’s function and simulation principle are illustrated respectively,then the system’s overall modeling is completed and we get expected results.6.Implementation and Validation of Traffic Signal Scheduling Algorithm.UCOS-Ⅱ system is choosed after comparing current embedded platforms based on the characteristic of self-adaptive traffic signal control system.The characteristic of UCOS-Ⅱ operating system is summarized.Finally,traffic signal control system’s functions are implemented on UCOS-Ⅱ system.