Abstract— Traditionally, only basic signal timings have been optimized in order to minimize delays and stops of private vehicles. Transit Signal Priority parameters and subsequent beneficiaries, such as transit vehicles and passengers, are usually neglected in the signal timing optimization process. Little research has been done to reveal specific benefits of optimizing Transit Signal Priorities and whether consideration should be given to both private vehicles and others when optimizing signal timings. Research presented here tests optimization of three performance measures (auto delay, transit delay, and person delay) by adjusting signal timings in different ways. A Genetic Algorithm formulation, coupled with a high-fidelity microsimulation model, is used to investigate benefits of each optimization on a large urban traffic corridor. The results show that basic signal timings are the most important measure to optimize when transit and private cars share a corridor. Also, the findings show that personal delay represents a suitable objective function for optimization of signal timings. I. INTRODUCTION istorically, macroscopic optimization programs, which are primarily deterministic, such as TRANSYT-7F [1] and SYNCHRO [2], have been used to optimize traffic signal timings. In the last decade a new approach has emerged to stochastically optimize signal timings by using Genetic Algorithms (GA) [3]-[6]. Although GAs can be coupled with deterministic models their full advantage is obtained when interfaced with high-fidelity microsimulation tools. Consequently, a seemingly infinite number of objective functions and parameters may be modified. Researchers have previously used GA-based optimizations to adjust signal timings that were traditionally neglected (e.g. actuation settings) in the optimization process [7], [8]. Researchers also showed that GA stochastic optimizations can be used to optimize objective functions based on vehicular emissions and transit performance measures, which were also traditionally neglected in the traffic control optimization practice [9]. Manuscript received January 20, 2011. *A. Stevanovic, PhD, PE is Assistant Professor of Civil Engineering at Florida Atlantic University, Boca Raton, FL 33431, USA. (* - Corresponding Author - phone: 561-297-3743; fax: 561-297-3740; e-mail: astevano@fau.edu). J. Stevanovic, PhD, is an independent consultant who resides in Boca Raton, FL 33431, USA (email: jelkastev@yahoo.com). C. Kergaye, PhD, PMP, PE, is Director of Research at the Utah Department of Transportation, P.O. Box 148410, Salt Lake City, Utah 84114-8410 USA (e-mail: ckergaye@utah.gov). P.T. Martin, PhD, PE, is a Professor of Civil Engineering at University of Utah, Salt Lake City, UT 84112, USA, (email: peter@trafficlab.utah.edu). Transit Signal Priority (TSP), where transit vehicles get priority from the traffic controller at signalized intersections, is a functionality that can help make transit service more reliable, faster, and cost effective. Although TSP has been extensively investigated in urban traffic control, most of the research has focused on adaptive TSP strategies [10]. Little has been done on optimization of TSP settings (e.g. values for green extension and early green) within the broader optimization of basic signal timings (i.e. cycle lengths, phases, offsets, and splits). The current state of practice does not include optimization of TSP settings. Table 1 shows nine optimization scenarios and the frequency of each scenario in traffic control optimization practice. Basic signal timings are adjusted to improve operations of vehicular traffic. Sometimes they are also adjusted to improve transit operations although such actions are rare and not based on any standardized procedures [11]. Once basic signal timings are set, TSP settings are manually ‘fine-tuned’ based on the judgment of traffic signal practitioners (i.e without optimization). Furthermore, the signal timing optimization rarely goes beyond the need to minimize negative impacts (delays, stops, etc.) of traffic signals on private automobiles. Experiences of transit passengers and overall person-based performances on multi- modal transportation system are often neglected. Table 1 Current practice in optimization of signal timings Minimizing Delay of: Signal Timings to Optimize: Private Cars Person* Transit Vehicles Basic Very common Uncommon Common TSP Uncommon** Uncommon Very common** Basic + TSP Uncommon Uncommon Uncommon * Delay per person in the system regardless of travel mode used. ** TSP parameters not optimized (but manually adjusted) since commercial optimization tools do not adjust these parameters. The goal of this study is to address the relative importance of optimizing various signal timings using non-traditional (transit-inclusive) objective functions in the optimization process. Optimization of signal timings for all nine scenarios presented in Table 1are investigated. A large-scale urban network that represents a suitable environment to test scalability of the proposed approach is used. The study also employs a special GA to model a TSP emulator (EPAC-300) Traffic Control Optimization for Multi-Modal Operations in a Large-Scale Urban Network Aleksandar Stevanovic*, Jelka Stevanovic, Cameron Kergaye, and Peter Martin H