hi,

Does SimTraffic effectively model weaving aspects (based on distance, speed), of traffic?

Specifcally, do the default values for Positioning Distance, Mandatory distance, and the other lance change distances produce results in SimTraffic that are comparative to standard weaving analysis techniques (ie, Highway Capacity Manual)?

Essentially, I have a specific issue where I'm trying to model a highway on-ramp/off-ramp section of road where cars must weave. What are the SimTraffic vehicle weaving characteristics been based/tested on?

The manual says that the default values for Lane Change Distances are based on "casual observations and there is no formal research backing up these parameters.". What exactly were they based on?

ANy information would be most helpful!

Thanks!
D

Hi D,

Well, I don't think you're go to get the types of outputs that you want. SimTraffic is going to give you a speed, delay, etc. for a LINK. Often, for this type of analysis you want measures such as the speed of just the weaving vehicles. This type of output can't be extracted from SimTraffic.

The behavior of vehicle merging was originally based on the NETSIM model. The driver behavior has been modified over the years to calibrate it to achieve certain saturation flows. Specificlly for surface streets. However, as we suggest we reducing the headway factor to get more reasonable freeway saturation flows. Next, you'll also want to watch the turning speeds.

Thanks,

John.

Okay, but I'm more concerned with whether the visual output that SimTraffic displays (ie, cars moving around) is adequate for modelling a section of road where vehcile weaving is a concern?

Was vehicle weaving characteristics ever compared with any other standardized Engineering methods of analyzing the weaving behaviour? (like Highway Capacity methods)

Can I show a region the SimTraffic demonstration and tell them that what they are seeing is a ____(good, fair, poor) representation of what will take place in a weaving section of road? (ie, highway on-ramp/off-ramp)

When they ask me as a Traffic Engineer: "How is the weaving modelled in this software?", I'd like to respond with an answer that can be backed up with some standard engineering theory.

Any clarification would be great!

cheers,
D

Hi D,

Directly to the question, no SimTraffic results have not been directly compared to macroscopic methods of the HCM for weaving behavior. SimTraffic is based on sound engineering principles for driver behavior. That is, weaving results are not determined by equations, such as in the HCM, but by logic that the vehicles make based on lane change algorithms.

Now, the question becomes whether one can feel comfortable with the lane changing logic in SimTraffic. Is it valid for weaving? It has been shown to be valid for other types of lane changing. The question becomes if weaving is that different then other lane changing behavior? The lane change logic is based on the models used to develop the NETSIM model.

I could say that I am comfortable with the lane change logic in SimTraffic. In addition, the end user does have some ability to calibrate the model for local conditions. For instance, the Positioning Distances and Mandatory Distances can be adjusted for a specific situation.

Thanks,

John.