Using 'Loop Detectors' in Paramics

Loop Detectors are designed to replicate the exact data returned by physical inductive loop detectors in the carriageway. This was part of the detailed information gathering and comparison that was part of the Paramics validation as a microsimulation tool.

Each loop detector is separated into two components, the upstream edge and the downstream edge, representing the two sides of the inductive loop. Each edge operates independently in gathering data, which it relates to the other edge in terms of formatting the output data; this is the same process that physical loops follow.

As Paramics operates with discrete time steps vehicles only exist on the network in a series of hops being placed on the road every time step and removed between timesteps. In this way it can be imagined that vehicles hop their way down a link, only being in place when they are in contact with the road surface. Each timestep, a vehicle has precise information regarding its current speed, vehicle length, acceleration, headway etc. as well as its planned position in the next timestep, considering all the dynamic attributes held currently.

In this way when a vehicle is positioned on the edge of a link the vehicle data is gathered by the edge for use in reporting. With this information the loop detector is able to calculate the actual time (not the timestep) that the vehicle first crossed the edge, creating the inducted current in the loop i.e. the rising edge. This data is also true of the falling edge of the inducted current and allows the edge to calculate the same data that an actual loop detector would.

When both the upstream and downstream edges of a loop detector are considered it is possible to identify a vehicles speed and acceleration etc as well as other data synonyms with traffic flow, such as headway, occupancy, gap, flow and count.

The detailed way in which the data is gathered allows loop detector data to be used to match direct output of physical loop data for comparison. However it is important that because of the way that the data is gathered any areas where physical loop detectors are inaccurate this is also true in Paramics. For example, in slow moving congested situations physical loop detector data is unreliable as separate vehicles can occupy the upstream and downstream edges of a detector, causing confusion regarding the count, flow and occupancy as there may be less than 2m (6.6 ft) between vehicles in these conditions.

The length of a loop detector is defined as the distance between the upstream and downstream edges. The default loop length is defined in the 'Core Network Attributes > Configuration > Core > Loop Length' and can also be set when each detector is created or edited.

Loops occupy one lane completely regardless of the width of the lane. Loops also report data by lane and are not aggregated across all lanes.

It is always worth considering the number of Time Steps per second used in the simulation as a low number of timesteps on high speed links may result in occasional vehicles hopping over loop detectors which will result in no data being recorded for the vehicle. Paramics will be able to record vehicle data if any part of a vehicle covers either edge of a loop detector in a timestep. To ensure that all vehicles are recorded for detectors (2m loop length on links of up to 120 km/h) it is best to use 3 or more timesteps per second.
UPDATED:  October 3, 2019