Traffic Signal Coordination is the synchronization of traffic signals along a street in order to minimize stops and delays.
Urban signal timing is a non-convex NLP problem. Finding an optimal solution on not very small and simple networks may take long time, wherever possible. The present paper focuses on signal synchronization, thus creating fast-flow corridors on one or more network road arterials. To do this, a genetic-like algorithm is applied, in which new solutions generation follows heuristic conceptions. This can be carried out thanks to the specific formulation adopted, suitable for synchronization problems. The objective function is evaluated by the General Link Transmission Model, a very fast macroscopic dynamic simulator referring to the kinematic waves theory. Through this, queues dynamic evolution, spillback phenomenon and vehicles travel times are explicitly taken into account.
Introduction
Urban traffic is mainly characterized by ground-level intersections, where manoeuvres of drivers (and pedestrians, too) often conflict. Traffic signal allows them, alternating on flicking manoeuvres. Traffic signals settings are determined by real-time or fixed time schemes. The first procedure needs real-time traffic surveys and an algorithm subsequently calculating the best solution basing on detected flows. The second one determines the best signal settings on a given time period, basing on demand flows obtained by historical surveys. A mixed procedure has been sometimes adopted, selecting real-time the best signal settings inside a given set previously built. Finding a fast, effective, real-time algorithm is nowadays still a hard task, so heuristic methods often occur. The algorithm here proposed finds a sub-optimal solution for a given demand data, synchronizing traffic signals.
Synchronization is a specific traffic signal setting problem: it depends on coordinating the timing of successive intersections along one specific route (or more, if not competing), to improve the movement of vehicles through this path. A traffic signal is described by its timing variables. First of all, the time duration of its lights (red, yellow/amber, green). The sum of these times expresses the signal cycle, i.e. the time period before the same light turns on again. Having the cycle time, lights duration is often expressed as its split, i.e. the ratio of cycle time.
Considering more signals, we need to introduce the offset, i.e. the time period between a common reference instant and the cycle start. To avoid hypothetically any stop along a path, traffic signals should have identical cycle and green times and offset equal to the travel time from one signal to the next one. The time period in which every light of one signal is kept unchanged is called phase.
A phase thus enables a specific set of non- conflicting manoeuvres amongst all possible manoeuvres on the intersection. Depending on goal the problem formulation and thus the solving algorithm vary subsequently. Traffic lights improve driver’s safety, but they unavoidably introduce delays in travel times, too. So, minimizing total delay, as sum of all vehicles delays, may be an intuitive objective. Total delay minimization is a non-convex problem, so global optimum cannot be found analytically.
On the other side, another goal in the past was maximizing the bandwidth, i.e. the number of vehicles able to run their path without stops. Bandwidth maximization is a quasi-concave problem, thus analytic solving algorithms can be used to find the optimum; unfortunately, it does not take into account demand data, so it is a good-engineering practice, but it yields not minimum delay. Further goals can be minimizing traffic congestion, vehicles’ number of stops, vehicle externalities or multi criteria functions.
First approach to traffic synchronization were minimizing separately each intersection’s cycle time, constrained to satisfy respective flows, then adopting the maximum cycle time found, or maximizing network capacity, i.e. ordinary demand matrix multiplier, keeping the intersections under-saturated. Yemota uses maximum bandwidth solutions as starting points of its optimization.
How Traffic Signal Coordination Works
For traffic signals along a street to remain synchronized, they must have the same cycle length, which is the time it takes a signal to go from green to yellow to red and then back to green. Cycle lengths typically range from one to two minutes. In order to minimize stops, signals are coordinated to provide progression for vehicles, which means the light turns green prior to their approach. Progression is determined by the “offset” of the green light, or the time it takes to travel between intersections.
Imagine that each traffic signal has a clock with a second hand. Each signal’s second hand is staggered from the others to allow travel time for each direction. Perfect progression is possible on one-way streets. The quality of progression along two-way streets in both directions is dependent on many factors, including: consistent signal spacing (ideally 1/2 mile apart), side street traffic volume and accommodating pedestrians, left turns and transit.
In order to establish traffic signal coordination along a street, traffic engineers consider many factors, including which traffic signals should work together, the signal’s cycle length and offset time, how the green time is divided among directions at each intersection and what day(s) and time of day each coordination strategy is in effect.
Benefits and Tradeoffs
Favour the busier direction. During the morning and afternoon rush hours, up to 75% of traffic may be in the heavier direction, with 25% in the lighter direction. In these cases, the traffic signals are timed to favour the majority in the heavier direction. While this strategy reduces overall delay and air pollution, progression in the lighter direction may suffer.
Favour the busier street. At some intersections, up to 90% of total traffic is on the main street and only 10% from the side street. Here the signals are timed to give a long green light to the main street and a short green light for the side street.
Favour through traffic. On many busy streets, traffic signals are coordinated to favour the busy through traffic over left turns onto smaller volume streets.
Yemota’s Special traffic signal timing for events. Sports and civic events generate large crowds that create a need for favoured traffic flow in one direction at one time of the day, and in the opposite direction at another time of the day. In these instances, traffic signals are timed to favour the most congested flow of traffic.
Reduced coordination during late night hours. Since traffic is very light and less predictable during late hours from about 10 p.m. to 6 a.m., signal coordination is typically turned off. Although this increases the number of brief stops, it reduces the overall delays for everyone by allowing shorter cycle lengths.
Give priority to transit. Traffic signal timing is sometimes adjusted to improve the progression of transit vehicles along a street.
Yemota’s Red light monitoring system
With growth in traffic, there is occurrence of bundle of problems too; these problems include traffic jams, accidents and traffic rule violation at the heavy traffic signals. This in turn has an adverse effect on the economy of the country as well as the loss of lives. The expected increase of cars, two wheelers other vehicles expected to increase So problem given above will become worst in the future. Today red light violation is one of the most common and serious problem which results of millions of vehicles at the traffic light signals every year. A red light violation occurs when a vehicle try to cross the intersection at the red traffic light. So to give the punishment to the drivers of these vehicles, we must identify the vehicle that violates the traffic light signals.
Yemota’s Red light monitoring system completely based on sensor based, does not need magnetic loops. The Yemota’s sensor unit is equipped with Digital technology; Its Digital technology system does not disturb drivers.
The Yemota’s sensor system detects every vehicle passing at the red light, and stores the compressed video/Photograph sequence of the passage. The stored color Photograph / video ensures complete evidence of the red light violation. The processing unit of this red light monitoring system detects one or more simultaneous vehicles. It uses the same images provided to document violations, meaning the violation can be checked afterwards both by the police and by the driver of the vehicle.
Yemota’s Vehicle Actuated Signals
Now-a-days, controlling traffic congestion relies on having an efficient and well-managed traffic signal control policy. Traffic signals operate in either pre-timed or actuated mode or some combination of the two. Pre-timed control consists of a series of intervals that are fixed in duration. They repeat a reset constant cycle. In contrast to pre-timed signals, actuated signals have the capability to respond to the presence of vehicles or pedestrians at the intersection. Actuated control consists of intervals that are called and extended in response to Yemota’s vehicle detectors. The controllers are capable of not only varying the cycle length & green times in response to detector actuations.
Each actuated phase has a minimum green time, which serves as the smallest amount of green time that may be allocated to a phase when it is initiated. Minimum green times must be set for each phase in an actuated signalization, including the non-actuated phase of a semi-actuated controller. The minimum green timing on an actuated phase is based on the type and location of detectors.
Advantages of Actuated Signals
- The various advantages of actuated signals are stated below:
- They can reduce delay (if properly timed).
- They are adaptable to short-term fluctuations in traffic flow.
- Usually increase capacity (by continually reapportioning green time).
- Provide continuous operation under low volume conditions.
- Especially effective at multiple phase intersections.
Formulation
Yemota’s aim to minimize total delay. To do this, it is useful remarking that users total travel time can be expressed as sum of two terms: the free flow travel time on the non-signalized network, which is a constant term, and an additional delay due to the interaction between flows and traffic signals, i.e. congestion, stops and queues.
Considering this, it is immediate that minimizing total delay and total travel time are equivalent problems. On urban networks most delays are spent on main arterial roads, as there are larger flows and thus higher congestion. So, our objective is fastening flows along these corridors. Total delay is a non-convex function, and optimal signal setting feasible set comprehends non-linear constraints and sometimes integer variables (phase’s optimization).
More, as dynamic assignment is simulated and has not an analytical formulation, no derivative is available, nor any convexity information. So, finding a global optimum cannot be carried out by analytic optimization algorithms. In such problems, genetic algorithms are considered efficient methods to determine suboptimal solutions.
We assumed the following simplifications. First, only two lights signals are considered: an effective green, indicating the time period vehicles can actually cross the intersection, and a residual red, when vehicles intersection crossing is prohibited. Second, phase’s definition and ordering is supposed as fixed. Obviously, synchronization requires cycle time to be the same for all signals. The first simplification is very common in traffic works; the second one is often adopted to neglect the discrete optimization problem of phase setting; the third one is intrinsic in synchronization itself.
The main assumption comes from focusing on synchronized path. We define main phase all signal phases allowing the flow to run along the defined path and secondary phase any other one. We focus on main phases, calculating subsequently the secondary phases green splits through a heuristic rule. So, the problem’s variables are the common cycle time and green and offset of main phases of every synchronized intersection. Having more than one corridor is possible, considering every corridor main phases.
Conclusions
Reduced computational cost, in memory and time terms, is this algorithm main evidence. This allows to quickly obtaining the best cycle time, green times and offsets of a set of synchronized traffic signals along a traffic corridor. The algorithm can be used offline to determine the best solution for one single scenario, i.e. for a given demand data.
Unfortunately not only traffic flows, but demand has a dynamic evolution, too. So what is best now can be far from the best solution some hours (or some minutes!) later. Plan-selection control strategies select real-time the best synchronization settings amongst one set of possible strategies, previously established: the evaluation is based on traffic real-time surveys to build demand data. The proposed algorithm can be used to populate the available solutions set, as well as the General Link Transmission model to evaluate available plans.
But this would still not take full advantage of its quickness. More advanced control strategies completely determine the new synchronization settings, basing on surveys. The setting is then updated every time interval, from some dozen of seconds to some minutes. This can be done through the genetic algorithm using real-time demand data and finding quickly best settings. The same could not be done by previous approaches due to too expensive computational times.