Adaptive Control
Many traffic control systems manage the signals on a fixed-time basis, where a series of signal timing plans are scheduled by day of week and time of day. The time relationship between signals is pre-calculated; based on previously surveyed traffic conditions. Such fixed-time systems cannot be expected to cope with traffic conditions that differ from those prevailing when the intersection was surveyed.
Furthermore, as traffic patterns change with the passage of time, fixed time plans become outdated. This requires the area to be resurveyed, and new signal timing plans calculated every few years. Experience has shown this procedure to be expensive, and to require resources which are not always readily available. As a result, the development of new plans is either deferred beyond the useful life of the old plans, or improvised changes are made to the plans and timetables; either case results in sub-optimum performance.
The problems of most fixed-time systems make it clear that a more responsive approach to changing traffic conditions is needed. One cost-effective answer is the SCATS 6 Fixed Time Plan system. This is a great improvement on other fixed time systems because it has the benefit of improved decision making capabilities built-in.
The full answer, of course, is the Adaptive SCATS 6. Unlike most fixed-time or semi-responsive systems, it requires no costly pre-calculation of signal timing plans. Additionally, SCATS is self calibrating, automatically adjusting to changing traffic patterns over time. The SCATS 6 controllers and traffic control computer analyse real-time traffic data from vehicle detectors, and produce signal timings which are suitable for the traffic conditions as they really are. It offers a variable sequence of signal phases, and the option to omit phases or movements from the sequence on a cycle-by-cycle basis when there is no demand.
The implementation of a fully responsive system does not, however, mean that the careful design of each intersection can be avoided. The present state of technology only allows for the real-time variation of signal timings at intersections which have known or anticipated traffic requirements.
Adaptive control in SCATS
In the Masterlink mode of operation, SCATS control of traffic is effected at two levels which together determine the three principle signal timing parameters of traffic signal coordination; cycle time, phase split, and offset. These two levels are referred to as strategic and tactical. Strategic control is basically concerned with the determination of suitable signal timings for the areas and sub-areas based on average prevailing traffic conditions while tactical control refers to control at the individual intersection level within the constraints imposed by the regional computer's strategic control.
Traffic information for both strategic and tactical functions is measured using inductive loop vehicle detectors. All detectors are capable of performing the tactical function. All detectors are capable of being defined as strategic detectors and information from these is pre-processed in the local controller and sent to the regional computer for the strategic calculations.
The cycle time is the time taken to complete one sequence of all phases and must vary to meet the overall level of traffic demand because, in general, increased cycle time increases system capacity. All signals which are coordinated must share a common cycle time (or sub-multiple). Cycle time typically varies between 40 and 180 seconds. SCATS provides for operation of cycle times between 20 and 190 seconds, the actual limits being user defined. The system dynamically adjusts cycle time to maintain the highest degree of saturation in a coordinated group of signals within acceptable user defined limits.
Phase split refers to the division of the cycle into a sequence of green signals for the competing movements at each intersection and must reflect the relative demands for green time on each approach. SCATS determination of phase splits is essentially one of maintaining equal degrees of saturation on competing (representative) approaches. However, control may be biased to favour principal traffic movements when demand approaches saturation. Offset refers to the time relationship between the phase introduction points of adjacent signals. The pattern of offsets in a series of coordinated signals must be varied with traffic demand to minimise the stops and delay associated with travel through a network of signals. SCATS selects offsets, based on free flow travel time and degree of saturation, which provide minimum stops for the predominant traffic flows.
Strategic Control
SCATS strategic control refers to the top level of control which is impressed on a network of coordinated signals by the regional computer. Using flow and occupancy data collected from loop detectors in the road by the local controllers, the strategic algorithms determine, on an area basis, the optimum cycle time, phase splits and offsets to suit the prevailing average traffic conditions. This is carried out for adjacent groups of signals (usually one to ten in size) which are known as sub-systems. Provision is made for groups of sub-systems to link together to form larger systems. Up to 64 sub-systems may be controlled by each regional computer and these may group together to form one big system or several completely independent systems.
Each sub-system consists of one or more intersections and contains only one critical intersection which requires accurate and variable phase splits. The intersections in a sub-system form a discrete group which are always coordinated together and share a common cycle time and interrelated split and offset selection. The sub-system in SCATS is the basic unit of strategic control. Phase splits and cycle time are calculated for the critical intersection and offsets are determined by the amount of traffic flowing in each direction through the sub-system. Phase splits for minor intersections in the sub-system are, by definition, non critical and are therefore either non variable or selected by a matching process which selects splits which are compatible with the splits in operation at the critical intersection.
To give coordination over larger groups of signals, sub-systems can link together to form larger systems, operating on a common cycle time. These links, which determine the offsets between the sub-systems, may be permanent or may link and un-link according to varying traffic conditions. This ensures that where traffic flow between sub-systems is sufficient to warrant coordination the link is enforced but when one or more sub-systems can operate more efficiently at a lower cycle time, the link is broken.
The basic traffic measurement used by SCATS for strategic control is the degree of saturation on each approach or, more accurately, a measure analogous to degree of saturation. Inductive loop vehicle detectors placed in all important approach lanes at the stop line of the critical intersections (and some detectors at other intersections) are defined in the regional computer data base as strategic detectors. The local controller collects flow and occupancy data from these detectors during the green of the approach. After pre-processing, the data is sent to the regional computer and used (together with automatically self-calibrated saturation flow data for each detector) to calculate the SCATS degree of saturation (DS).
DS is defined as the ratio of the effectively used green time to the total available green time on the approach. The effectively used green time is the length of green which would be just sufficient to pass the same platoon of vehicles had they been travelling at optimum headways as in saturation flow conditions. The difference between the effectively used green and the available green can be thought of as wasted green and this is easily measured by summing the periods of non-occupancy of the detector during the green period and from this subtracting the spaces which must necessarily accompany each vehicle under saturation flow conditions. The value of saturation flow space to subtract is automatically calibrated by SCATS for each lane of strategic detection.
The measure DS is essentially independent of vehicle length and therefore independent of vehicle mix (eg cars, trucks, buses). The algorithm is capable of producing values of DS greater than unity in congested conditions, enabling SCATS to deal effectively with oversaturated traffic. The SCATS DS algorithm will produce reliable and useful values for bicycles when a suitable detector, one metre in length, is installed in a bicycle lane. The calculation of DS relies on the detector being of sufficient length in the direction of traffic flow to ensure that large values of space are not measured under conditions of slow moving closely spaced traffic (which would appear to be the same as light traffic widely spaced). The detector must not, however, be too long as it would not measure any spaces when traffic moves freely. Research has shown the optimum length of the detection zone to be 4.5 metres.
From the DS measured for each lane of strategic detection, a normalised flow rate is calculated which is analogous to pcu (passenger car unit) flow. This is simply obtained by multiplying the value of DS by the automatically calibrated saturation flow rate.
Cycle time is increased or decreased to maintain the degree of saturation around 0.9 (user definable) on the lane with the greatest degree of saturation. Other lanes or approaches may have lower degrees of saturation. A lower limit for cycle time (usually 30 to 40 seconds) and an upper limit (usually 100 to 150 seconds) are specified by the user. Cycle time can normally vary by up to 6 seconds each cycle but this limit increases to 9 seconds when a trend is recognised.
Phase splits are varied by up to four percent of cycle time each cycle so as to maintain equal degrees of saturation on the competing approaches, thus minimising delay. The minimum split which can be allocated to a phase is either a user definable minimum or, more usually, a value determined from the local controller's minimum phase length which may vary according to whether pedestrians are using the walk signals associated with the phase. The maximum split which can be allocated to a phase is limited by the current cycle time and the minimum requirements of the other phases.
Offsets are selected for each sub-system (ie, the offsets between intersections within the sub-system) and between sub-systems which are linked together on the basis of the pcu traffic flow obtained from DS. In this way the best offsets are selected for the high flow movements. Other links carrying lower flows may not receive optimum coordination when the cycle time is inappropriate. However, when traffic conditions permit, the system maintains a cycle time which can provide good offsets on a majority of links even though a smaller cycle time could provide sufficient capacity. Good offsets on the heavy flow links minimise the total number of stops in the system, reducing fuel consumption and increasing the capacity of the system.
Tactical Control
SCATS tactical control refers to the lower level of control which is undertaken by the local controllers at each intersection. Tactical control operates under the strategic umbrella provided by the regional computer but provides local flexibility to meet the cyclic variation in demand at each intersection. Tactics essentially provide for green phases to be terminated early when the demand for the phase is less than the average demand and for phases to be omitted entirely from the sequence if there is no demand. The local controller bases its tactical decisions on information from the vehicle detector loops at the intersection, some of which may also be strategic detectors.
The tactical level of control is carried out in the local controller using exactly the same operational techniques as for isolated operation. The degree to which tactical control is able to modify the signal operation is entirely under the strategic control of the regional computer. Briefly, any phase may be omitted from the sequence if not demanded, may terminate early (before expiry of the time allocated by the strategic operations in the regional computer) under control of the gap timers or waste timers or the phase may continue to its maximum value. The action of these timers and vehicle actuated control in general is treated in the section on local controllers.
A basic difference from isolated operation is that one phase, usually the main road phase, cannot skip and cannot terminate early by action of gap and waste timers. This is because all controllers in a system must share a common cycle time to give coordination. Any time saved during the cycle as a result of other phases terminating early or being skipped may be used by subsequent phases or is added on to the main phase to maintain each local controller at the system cycle length.
The combination of strategic control which varies the split, cycle time and offsets in response to gradual changes in traffic demand patterns together with tactical control which handles the rapid but smaller changes in demand cycle by cycle results in a very efficient operation of the signals on the street.
