Real-time Estimation Method Of Turning Movement Counts At Signalized Intersection

Turning movement counts at signalized intersections is origin-destination(OD)matrix in the micro sense,which reflects the characteristics of the local signalized intersection traffic flow.Estimation of intersection origin-destination(OD)matrices is a frequently-encountered problem in transportation management operations.What is more,modern traffic signal control system need not only accurate and reliable information to make effective control measures,but also to adapt to the traffic demand under the condition of changing circumstances.The more accurate traffic volume estimation,the more effective system control scheme is.In recent years,more and more scholars to focus on the real-time estimation of turning movement counts at signalized intersections.A variety of sensor technologies,such as loop detectors,traffic cameras,and radar have been developed for real-time traffic monitoring at intersections most of which are limited to providing link traffic information with few being capable of detecting turning movements.Based on this condition,first using intersection access and exit amount as input to establish unconstrained least squares estimation model and estimate the signal intersection turning movement counts.Due to no constraints about turning proportion,so the estimated value have larger error.Then using signal intersection exit counts as input to establish recursive least squares recursive model and with forgetting factor model,the error of intersection of the turning proportion estimation significantly reduces.The recursive least squares method with forgetting factor can weaken the old estimates of the new parameters.The new data indirectly improve the role of the parameter estimation results can be better.Secondly,the genetic algorithm is used to estimate the number of vehicles at signalized intersections,because it is a heuristic algorithm,so the efficiency and accuracy of the proposed method are better than those of the previous two methods.Finally,considering the signal phase information as a constraint condition is introduced to estimate the model and put forward the estimation model based on least squares signal phase information,in order to improve the signal control intersection steering vehicle number estimation accuracy and stability.The new algorithm makes full use of signal phase,minimizing the potential fuzzy estimation,to signal phase schemes and obtain data shift signal intersection by Vissim micro simulation,the data processing into the algorithm and actual values were compared,the estimation model and algorithm proposed in this dissertation are verified.This evaluation method is proposed to estimate the vehicle steering stability and accuracy of the signal intersection,and the other four existing estimation results were compared to estimate method.Finally,the case analysis shows that the constrained least squares estimation model of signal phase information stably and accurately on signal intersection turning vehicle number estimation based on real time,and better than the other methods described above.However,the stability of the traffic volume data is higher and the algorithm is easier to implement.