In Figure 4.14, the estimated upstream flow is compared with the simulation input.
To give a better understanding, these flow rates have been transformed into flow ratio (r). The flow ratio (r) is then compared with the flow ratio derived from simulation (r*).
The comparison of estimated upstream speed (U) and detected speed (U*) is illustrated in Figure 4.15. The estimated upstream speed (U) is space mean speed but the detected speed is time mean speed; therefore, the detected speed is transformed to space mean
Red time Green time
Detector
Red time Green timeRed time
(a) (b) (c)
speed with the method proposed by Drake, Schofer, and May [47]. The MAPE and MAE of flow ration are 18% and 0.03 respectively. While those of space mean speed are 4% and 1.79ft/sec respectively. These results demonstrate that the proposed algorithm is capable of estimating flow and speed at upstream area.
Figure 4.14.The predicted traffic flow of state 3
Figure 4.15 The predicted traffic speed of state 3
4.5 Traffic signal control algorithm
To investigate the performance, the proposed model will adjust the maximal green time dynamically according to flow demands. The enhanced actuated control will be compared with traditional full-actuated control which‟s maximal green time is fixed generally.
0.0 0.1 0.2 0.3 0.4 0.5 0.6
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 cycle
flow (ratio of Qm)
r r*
0 10 20 30 40 50 60
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 cycle
feet/sec
U U*
This study employs a closely spaced intersections consisting of three intersections.
Basic layouts and phase configurations are given in Figure 3.16 and Figure 3.17(b). The spacing between intersections is set to be 200 feet. The numerical test includes 10 demand entries (A-J) and two volume levels (stable and unstable) designed to test the performance of proposed control model. Table 4.9 summarizes all experimental scenarios. The stable demand entries in all critical paths have the same volume. The unstable demand scenario has different peak periods for four critical paths.
Table 4.9. Experimental scenarios for model evaluation
Demand scenario Path Demand entries (in vph)
A B C D E F G H I J
Stable 1 700 700 700 700 700 700 700 700 700 700
2 700 700 700 700 700 700 700 700 700 700 3 700 700 700 700 700 700 700 700 700 700 4 700 700 700 700 700 700 700 700 700 700
Unstable 1 700 800 900 900 800 700 600 600 600 600
2 600 600 600 600 600 600 700 800 900 900 3 600 600 600 600 700 800 900 900 800 700 4 900 900 800 700 600 600 600 600 600 600
The proposed model was coded in C++ and tested under the runtime extension of CORSIM. The CORSIM is used as an evaluator. To overcome the stochastic nature of a microscopic simulation system, an average of 10 simulation runs has been used. For the measure of effectiveness (MOE) comparison, since CORSIM calculates total delays or average delays only for departed vehicles, it is not computationally convenient to use delay as the MOE for over-saturated conditions. Hence, in this study we use total queue time, maximal queue, queue delay and speed as the MOE.
The overall results of proposed model and full-actuated method under different traffic demands (as indicated in Table 4.9) are compared in Table 4.10. Compared to ordinary full-actuated scheme, this proposed methodology would improve the total queue time of 13.8% under stable demand situation; while under unstable demand, this value would increase to 31%. Other performance indexes are laid out in Table 4.10. It is clear that compare to ordinary full-actuated scheme, the proposed model works even better in the unstable-demand scenario. This improvement can be creditable to the path-based progression and the dynamic maximal green times among critical paths.
Table 4.10 Comparison of CORSIM simulation results.
Scenarios MOEs Simulation results from CORSIM (4h) Proposed Full-act
uated
Improvementa(%
Stable–demand Total queue ) time(veh-min)
8476.7 9829.4 13.8 Maximal queue(veh) 12.6 13.4 6.0 Queue delay(sec/veh) 45.4 52.7 13.8
Speed(MPH) 6.5 5.8 12.0
Unstable-demand Total queue time(veh-min)
9257.6 13407.8 31.0 Maximal queue(veh) 19.9 33.8 41.1 Queue delay(sec/veh) 49.4 71.2 30.6
Speed(MPH) 9.3 7.3 27.6
a Improvement is calculated by (MOEproposed-MOEactuacted)/MOEactuated
Figure 4.16 demonstrates the dynamic change of maximal green time according to its corresponding demand on different critical paths. To evaluate the performance of the proposed method, this study compares indexes on critical paths i.e., the queue time, queue delay and speed for each path under different demand scenarios. The comparison of indexes is demonstrated in Figure 4.17 through Figure 4.19. The proposed model outperforms the ordinary full-actuated control scheme among all paths. Under both stable and unstable demand scenarios (see Figure 4.17 and 4.18), the proposed model can deliver a more efficient control strategy than ordinary full-actuated control scheme.
With the proposed model, all paths have improved queue time, maximal queue, queue delay, and speed compared to ordinary full-actuated method.
Figure 4.16 Relation between the maximal green time and traffic flow for proposed model.
(a) (b)
Figure 4.17 Queue times of paths under (a) stable demand (b) unstable demand.
(a) (b)
Figure 4.18 Queue delay of paths under (a) stable demand (b) unstable demand.
18
0 2000 4000 6000 8000 10000 12000 14000
Time(sec)
Max green (sec) Flow (veh)
Path1-Flow Path2-Flow Path3-Flow Path4-Flow
Path1-Max green Path2-Max green Path3-Max green Path4-Max green
0
(a) (b)
Figure 4.19 Speed of paths under (a) stable demand (b) unstable demand.
0 4 8 12
path1 path2 path3 path4
Speed(MPH)
Proposed Actuated
0 4 8 12
path1 path2 path3 path4
Speed(MPH)
Proposed Actuated
V. Conclusions
In this study, a CMOS based transceiver with STC antenna has been successfully implemented for advanced traffic signal processing. The collected radar signals from the CMOS radar system have been processed with developed optimization algorithms for vehicle-type classification and speed determination. The high recognition rate optimization algorithms are mainly based upon the information of short setup time and different environmental installation of each sensor. The algorithm includes four phases, namely signal processing, calibration, learning and „classification and speed estimation‟.
In the calibration and learning phases, a video recognition module has been further adopted as a supervisor of SVM and SVR. SVM has successfully classified vehicles into four categories: motorcycles, small, medium and large vehicle in the classification phase. SVR has estimated vehicle lengths and determined their speeds accurately in the speed estimation phase. Specially, the proposed algorithm can detect motorcycles and estimate their speeds precisely. Compared with conventional circuit-based detector systems, the developed CMOS radar integrates submicron semiconductor devices and thus not only possesses low stand-by power but also is ready for production. In the meanwhile, the algorithm has successfully provided a high recognition rate in a grey area which traditional unsupervised classifiers have low recognition rates and supervised classifiers are hard to prepare training data. Furthermore, the developed algorithm of this study simultaneously optimizes the vehicle-type classification and speed determination in a computationally cost-effective manner, which benefits real-time intelligent transportation system. In the future, the enhanced vehicle length and speed accuracy can be obtained by applying SVR to each category of vehicles.
Another direction for future research could be to apply the SVM model to vehicle signals of each lane.
In this study, we also proposed an innovative approach to estimate the upstream traffic information at intersection under oversaturated situation using shockwave analysis. A key methodological contribution of the approach is that it estimates shockwaves by combining traffic parameters, dynamic traffic signal timing and traffic flow models. By utilizing parameters of stopped duration, moving duration, and empty duration, that are estimated form the presence of the radar detector, we are able to calculate shockwaves including 1) forward recovery, 2) ideal backward forming, 3) ideal forward recovery, 4) backward forming, and 5) forward recovery shockwave.
To the best of authors‟ knowledge, this is the first study that utilizes real time shockwave by stopped duration to estimate upstream traffic flow and speed far beyond detection zones of vehicle detectors. With the shockwaves, upstream traffic flow and speed information can be estimated accordingly. These models are evaluated by traffic simulation and demonstrate a significant result. The proposed model has some pre-conditions for traffic flow state. These assumptions can be solved by combining linear regression and the information derived from multi-zone sensors to capture the variation of shockwaves.
Traditional full-actuated control scheme would work well only on an intersection or arterials. In closely spaced intersections, it might suffer from capacity loss, poor coordination and long congestion time due to considering only gap-out criterion, non-path-based progression and fixed maximal green time, respectively. With the above, a novel actuated critical path control model for designing traffic signal timings in closely spaced intersections had been presented.
effectiveness of the proposed method. The numerical example demonstrates a satisfying result compare to ordinary full-actuated scheme. Compared to ordinary full-actuated scheme, under stable demand this novel methodology would improve the total queue time, maximal queue, queue delay, and speed of 13.8%, 6%, 13.8%, and 12%, respectively; while under unstable demand, these values would increase to 31%, 41.1%, 30.6%, and 27.6%.
This study has several key contributions, including 1) developing a radar vehicle detection algorithm to simultaneously optimize the vehicle-type classification and speed determination, 2) using the presence of radar sensor to compute the stopped, moving and empty durations; and combining them to estimate the shockwaves, 3) introducing shockwave detection theory to dynamically adjust maximal green time for critical path with unstable traffic demands, 4) designing a path-based progression scheme that suitable for closely spaced intersections, 5) providing a traffic signal control method which can use fewer detectors than traditional traffic signal control scheme.
Reference
[1] A. Stove, “Linear FMCW Radar technigues,” in IEE Proceedings-F, 1992, pp.
343–350.
[2] D. E. Barrick, “FMCW Radar signals and digital processing,” National Oceanic and Atmospheric Administration, Tech. Rep., 1973.
[3] D. V. Arnold, J. B. D. JR., and B. C. Files, “Systems and methods for monitoring speed,” in US Paten No. 7426450 B2, 2008.
[4] H. H. W.-H Lin, J. Dahlgren, “An enhancement to speed estimation using single loop detectors,” in Proceeding of Intelligent Transportation Systems, 2003.
[5] J. J. Reijmers, “On-line vehicle classification,” IEEE Transactions on Vehicular Technology, vol. 29, pp. 156–161, 1980.
[6] H.-S. Lai and H.-C. Yung, “Vehicle-type identification through automated virtual loop assignment and block-based direction-biased motion estimation,” IEEE Transaction on Intelligent Transportation Systems, vol. 1, pp. 86–97, 2000.
[7] H. Roe and G. Hobson, “Improved discrimination of microwave vehicle profiles,”
Microwave Symposium Digest, IEEE MTT-S International, vol. 2, pp. 717–720, 1992.
[8] S. J. Park, T. Y. Kim, S. M. Kang, and K. H. Koo, “A novel signal processing technique for vehicle detection Radar,” Microwave Symposium Digest, IEEE MTT-S International, vol. 1, pp. 607–610, 2003.
[9] P.-F. Pai, “System reliability forecasting by support vector machines with genetic algorithms,” Mathematical and Computer Modelling, vol. 43, pp. 262–274., 2006.
[10] Q. He, Z.-Z. Shi, and L.-A. Ren, “A novel classification method based on hypersurface,” Mathematical and Computer Modelling, vol. 38, pp. 395–407, 2003.
[11] L. Edler, J. Grassmann, and S. Suhai, “Role and results of statistical methods in protein fold class prediction,” Mathematical and Computer Modelling, vol. 33, pp.
1401–1417, 2001.
[12] J. Duchene and S. Leclercq, “An optimal transformation for discriminant and principal component analysis,” IEEE Transactions on PAMI, vol. 10, pp. 978–983, 1988.
[13] R. A. Fisher, “The use of multiple measures in taxonomic problems,” Ann.
Eugenics, vol. 7, pp. 179–188, 1936.
[14] M. J. Lighthill and G. B. Whitham., “On kinematic waves: a theory of traffic flow on long crowded roads,” Proceedings of the Royal Society of London, vol. 229A, pp. 317–345, 1955.
[15] P. G. Michalopoulos, “Shock waves in traffic signal analysis and control,” in American Control Conference, 1982.
[16] A. D. May, Traffic Flow Fundamentals. Prentice Hall, 1990.
[17] H. X. Liu, X. Wu, W. Ma, and H. Hu, “Real-time queue length estimation for congested signalized intersections,” Transportation Research Part C-Emerging Technologies, vol. 17, pp. 412–427, 2009.
[18] D. C. Gazis, “The origins of traffic theory,” Operations Research, vol. 50, pp.
69–77, 2002.
[19] D. C. Gazis, Traffic Science. John Wiley Inc., 1974.
[20] D. L. Gerlough and M. J. Huber, Traffic flow theory: a monograph. Transportation Research Board, 1975.
[21] T. Kim and H. Zhang, “A stochastic wave propagation model,” Transportation Research Part B-Methodological, vol. 42, pp. 619–634, 2008.
[22] H. M. Zhang, “A theory of nonequilibrium traffic flow,” Transportation Research
Part B-Methodological, vol. 32, pp. 485–498, 1998.
[23] G. Abu-Lebdeh and R. F. Benekohal, “Development of traffic control and queue management procedures for oversaturated arterials,” Transportation Research Record, vol. 1603, pp. 119–127, 1997.
[24] H.-J. Cho and M.-T. Tseng, “A novel computational algorithm for traffic control soc,” WSEAS Transactions on Mathematics, vol. 5, pp. 123–128, 2006.
[25] F. Diona, H. Rakhab, and Y.-S. Kangc, “Comparison of delay estimates at under-saturated and over-saturated pre-timed signalized intersections,” Transportation Research Part B-Methodological, vol. 38, pp. 99–122, 2004.
[26]H.-J. Cho and S.-C. Lo, “Modeling self-consistent multi-class dynamic traffic flow,”
Physica A, vol. 312, pp. 342–362, 2002.
[27] H.-J. Cho and Y.-T. Wu, “Microscopic analysis of desired-speed car-following stability,” Applied Mathematics and Computation, vol. 196, pp. 638–645, 2008.
[28] B. D. Greenshields, “A study of traffic capacity,” Proceedings of the Highway Research Board, vol. 14, pp. 448–477, 1934.
[29] P. G. Michalopoulos, G. Stephanopoulos, and G. Stephanopoulos, “An application of shock wave theory to traffic signal control,” Transportation Research Part B-Methodological, vol. 15, pp. 35–51, 1981.
[30] X. Wu, H. X. Liu, and D. Gettman, “Identification of oversaturated intersections using high-resolution traffic signal data,” Transportation Research Part C-Emerging Technologies, vol. 18, pp. 626–638, 2010.
[31] A. Skabardonis, “Determination of timings in signal systems with traffic-actuated controllers,” Transportation Research Record, vol. 1554, pp. 18–26, 1996.
[32]F. Webster, “Traffic signal settings, road research laboratory technical paper no. 39,”
[33] R. L. Gordon and W. Tighe, “Traffic control system handbook, report no.fhwa-hop-06-006,” Federal Highway Administration, Tech. Rep., 2005.
[34] D. Husch and J. Albeck, Synchro 5 User Guide. Trafficware 1009B Solano Avenue Albany, CA 94706, 2001.
[35] TRANSYT-7F Release 11 Users Guide. University of Florida, 2012.
[36] J. D. C. Little, “Maximal bandwidth for arterial traffic signals: Theory and interactive computation,” Working Paper WP 970-78, Alfred P. Sloan School of Management, 1977.
[37] J. D. C. Little, “The synchronization of traffic signals by mixed-integer linear programming,” Operations Research, vol. 14, pp. 568–594, 1966.
[38] J. D. C. Little, M. D. Kelson, and N. H. Gartner, “Maxband: A program for setting signals on arteries and triangular networks,” Transportation Research Record, vol. 795, pp. 40–46, 1981.
[39] T. H. Chang and G. Y. Sun, “Modeling and optimization of an oversaturated signalized network,” Transportation Research Part B-Methodological, vol. 38, pp.
687–707, 2004.
[40] Z. Tian, T. Urbanik, and R. Gibby, “Application of diamond interchange control strategies at closely spaced intersections,” Transportation Research Record, vol. 2035, pp. 32–39, 2007.
[41] C. J. Messer, “Simulation studies of traffic operations at oversaturated, closely spaced signalized intersections,” Transportation Research Record, vol. 1646, pp.
115–123, 1998.
[42] Y. Liu and G. L. Chang, “An arterial signal optimization model for intersections experiencing queue spillback and lane blockage,” Transportation Research Part C-Emerging Technologies, vol. article in press, 2010.
[43] X. Zheng and L. Chu, “Optimization of control parameters for adaptive traffic-actuated signal control,” Journal of Intelligent Transportation Systems, vol. 14, pp. 95–108, 2010.
[44] E. Conte and M. Lops, “Clutter-map CFAR detection for range-spread targets in non-gaussian clutter,” IEEE Transactions on AES, vol. 33, pp. 432–442., 1997.
[45] S. Wang, H.-S. Wu, and C.-H. Chang, “Modeling and suppressing substrate coupling of RF CMOS FMCW sensor incorporating synthetic quasi-tem transmission lines,” in IEEE MTT-S International Microwave Symposium, 2007, pp. 1939–1942.
[46] C.-K. Tzuang, C.-H. Chang, and H.-S. Wu, “An x-band CMOS multifunction-chip FMCW radar,” in IEEE MTT-S International Microwave Symposium, 2006, pp.
2011–2014.
[47]H.-J. Cho and M.-T. Tseng, “Shockwave detection for electronic vehicle detectors,”
Lecture Notes in Computer Science, vol. 4490, pp. 275–282, 2007.