Development of a Multi-Resolution Network to Support ...
Development of New Supply Models in Maryland Using Big Data Jonathan Avner, Scott Thompson-Graves and Ashley Tracy (WRA) Subrat Mahaptra, Mark Radovic (MDOT SHA) 15th TRB National Transportation Planning Applications Conference Raleigh, NC May 15, 2017 Agenda Challenge Project Options Evaluated Approach
Findings Challenge MDOT SHA is working on several fronts to improve the capabilities of its tools for addressing both traditional and non-traditional applications Trip Based Model Activity Based Model DTA Lite Agent Based Freight Model By having a suite of tools available, able to align the tool with the type of question being asked including Resolution or detail of the analysis Runtime Desired performance measures
Challenge Ability to develop accurate forecasts in highly congested areas The way the traditional volume delay functions treat congested condition is less than ideal Challenge in ability to model peak hour conditions that make sense Project Can volume delay functions be developed: Improve sensitivity in the model Preform better under high volume conditions providing more realistic speeds Provide better performance measures including delay and level of service that would more closely replicate observed system conditions Requirements: Be developed using data available from the Statewide Model, Centerline and Route datasets and count program which includes
speeds at select locations Project Create a methodology to clean data for use in development of free flow speeds and volume delay functions Use of the data to identify variations in behavior that warrant unique facility types in the model platforms Use of the detailed centerline attribute data to capture these behavioral differences in improved speed and capacity Approach Data Collection MSTM Statewide Model Network Model Facility Type Model Number of Lanes Free Flow Speed and Hourly Per Lane Capacity Centerline Data Urban / Rural Designation Roadway Functional Class Number of Lanes
ATR Data 15 minute counts by speed bin Approach Leverage MDOT SHAs own Big Data ATR Data Observed travel speed by direction including volumes at 53 locations across the state Data provided by 15 minute interval for the month of September, 2015 Stations distributed by statewide model facility type: Freeways = 21 Expressways = 2 Arterials = 30 Stations distributed by urban / rural: Urban = 18 Rural = 29 Approach Approach
Approach Approach Data Cleaning Aggregate of 15 minute data to hour by station by direction for each day Identify data that was inconsistent Stations that included support facilities Inconsistent data Data Preparation Calculation of weighted hourly volume and observed speed Assignment of Free Flow Speed Assignment of Capacity Validate Speed, Capacity Logic Validate Volume Delay Functions Approach Validate Speed Comparison of observed speed under low volume (uninterrupted)
conditions to input speeds used in MSTM Validate Capacity Comparison of observed volume under high volume (interrupted) conditions to hourly capacity used in MSTM Volume Delay Functions Comparison of observed speed relationship using observed volumes to model delay functions Data Preparation Observed Volume and Speed For each hour Estimate of Flow (Veh/ Hr/Ln) Estimate of Speed Data Volume by 5 mph Speed Bin
Data Preparation - Capacity For purposes of calculation volume to capacity ratio, assignment of capacity by station Relied upon MSTM Capacity by SWFT (Functional Class) and Area Type Future enhancement is to calculate capacity in MSTM using geometric data Compared observed flow to MSTM capacity SWFT Urban Suburban Rural Halo External Interstate
53 68 Validation Volume Delay Functions Plot of Speed vs. V/C Speed: because grouping of locations, normalized to percent of free flow (PFFSPD) Free Flow Speed based on observed speed by station Capacity: MSTM capacity Testing goodness of fit using %RMSE Observed vs Model VDF
Consider urban vs rural sections Testing typical values of VDF for goodness of fit Validation Volume Delay Functions Reviewed data to ensure meet expected relationships Speed drop due to traffic delay Identified conditions of lower speed during low volume conditions Validation Volume Delay Functions Observation of decreasing speed under low V/C Silver Springs, MD on
the Beltway Validation Volume Delay Functions Goodness of fit urban vs rural Locations of decreasing flow removed from curve testing Diminishing Flow Conditions (Freeway) Validation Volume Delay Functions Isolation of point observations where speed is decreasing consistently with volume (before reaching capacity) Validation Volume Delay Functions
Observed variation in response to V/C ratio Consistent area type Need to investigate location and roadway geometrics Classify model links to capture variation Validation Volume Delay Functions Arterial includes SWFT 4-6 (Arterial Collector) Urban Suburban: issue with capacities and speeds Rural: several patterns Validation Volume Delay Functions Urban / Suburban MSTM capacities are
low as observed conditions exceed capacity Rural Additional factors influencing volume delay Validation Volume Delay Functions Consistency with volume delay functions except at low volume conditions variability of speed under uninterrupted conditions Validation Volume Delay Functions Overall Urban Rural
Freeway All Points 10.708 11.575 7.0618 Freeway Pre Jam Density 7.0705 7.1023 6.9676 Expressway 9.5834
9.5834 Arterial All 6.93779 77.17075 6.00623 Major Arterial 6.84764 10.6736 6.34373 Minor Arterial 6.90378
58 Next Steps Volume Delay Functions Improved goodness of fit with refinement by functional class, area type and other geometric factors Diminishing Flow Conditions 2 phase volume delay function Planning for autonomous vehicles Questions Jonathan Avner
Get to know your Professional School Counselor. It is EXTREMELY important for you to know who your school counselor is and what the school counseling office has to offer.. Ask about: Scholarships. After school programs. Clubs and activities. Community activities....
Grid data structure for representing numerical and categorical data. Map algebra raster calculations . Interpolation . Calculate slope on a raster using. ArcGIS method based in finite differences. D8 steepest single flow direction. D steepest outward slope on grid centered...
Panel Adam Lindgren, Principal Jim Simon, Principal Ken Lee, Senior Associate Audience Who's in the room? What do spheres of influence mean to you in your role as a commissioner? How about in your role as an agency board member...
What is WHMIS and what role does it play in laboratory safety? Why is it important to have a common set of safety labels and icons for hazardous chemicals in all workplaces and schools? Summary: How does matter affect your...
To begin the process, however, it is important for the student to collect the critical information from each source as they read: the author's name, title, publisher, and date of publication, and page number for quotes.
Bread The king and rich people ate white bread. Poorer people would eat rye, or wholemeal bread Baking was dangerous in Tudor times, as ovens were heated by fire. If the fire spread, the whole house could burn down! Dairy...
Nat. Rev. Neurol. 9, 572-582 (2013), Macmillan Publishers Limited. The corneal innervation image in part b (left panel) is reproduced with permission from Tavakoli, M. et al. Corneal confocal microscopy: a novel noninvasive test to diagnose and stratify the severity...
Ready to download the document? Go ahead and hit continue!