Transcript Traffic Estimation and Prediction Based On Real Time

1
Proceedings of the 11th International IEEE
Conference on Intelligent Transportation Systems, October2008
TRAFFIC ESTIMATION AND
PREDICTION BASED ON REAL
TIME FLOATING CAR DATA
Octo Telematics srl
ENEA
Corrado de Fabritiis, Roberto Ragona, Gaetano Valenti
Outline
2

Introduction




The OCTOTelematics Floating Car Data System
Traffic Speed Estimation
Preliminary Analysis of Estimated Link Speeds


Related Work
The Rome Ring Road Case Study
Approaches For Short-Term Travel Speed Prediction
Pattern Matching based approach
 Artificial Neural Networks based approach


Conclusions & Comments
Introduction
3

The wide scale deployment of ATIS and ATMS relies
significantly on the capability to perform
 Accurate
estimates of the current traffic status and
 Reliable predictions of its short-term evolution on the
entire road network

Real-time Floating-Car Data (FCD), based on traces
of GPS positions, is emerging as a reliable and
cost-effective way
ATIS – Advanced Traveler Information Systems
ATMS – Advanced Traffic Management Systems
Introduction
4



The FCD technique is based on the exchange of
information between floating cars traveling on a
road network and a central data system
The floating cars periodically send the recent
accumulated data on their positions, whereas the
central data system tracks the received data along
the traveled routes
The frequency of sending/reporting is usually
determined by the resolution of data required and
the method of communication
Related Work
5




The most common and useful information that FCD
technique provides is average travel times and
speeds along road links or paths [8], [13], [14]
Deploy FCD in order to predict short-term travel
conditions, to automatically detect incident or critical
situations [6], [7], [10]
Determine Origin-Destination traffic flow pattern [12]
The reliability of travel time estimates based on
FCD highly depends on the percentage of floating
cars participating in the traffic flow [3], [5], [11]
Introduction
6


This paper presents an evolution of an operating
FCD system, integrating short-term traffic
forecasting based on current and historical FCD
This system exploits data from a large number of
privately owned cars to deliver real-time traffic
speed information throughout Italian motorway
network and along some important arterial streets
The OCTOTelematics Floating-Car
Data System
7

OCTOTelematics is the European leader for
development and deployment of Telematics for
Insurance application
 With
approximately 600000 On Board Units (OBU)
installed (market penetration is 1.7%)
 Position,
 Provides
Heading, Speed
complete solutions for Pay As You Drive, Pay
How You Drive, Pay Per Use insurance
 Currently, OCTOTelematics is providing services to 32
insurance companies in Europe
The OCTOTelematics Floating-Car
Data System
8


Due to the large amount of real time data received for
insurance profiling purpose and due to the high market
penetration, several ITS application can be and have
been developed by OCTOTelematics
Large Scale Floating-Car Data System (LSFCD)
Tracks the received data along the traveled routes by
matching the related trajectories data to the road/street
network in order to
 Estimate link travel speeds and then, freely disseminates
them through WEB pages
 http://traffico.Octotelematics.it/index.html

ITS – Intelligent Transport Systems
Traffic Speed Estimation
9


LSFCD system monitors the entire Italian motorway
network (>6000Km) and some important arterial
streets located in major metropolitan areas
The proprietary LSFCD algorithm is divided in three
steps:
A) map matching (using Latitude, Longitude and Heading
from the GPS) for each positions
 B) routing (between subsequent positions) to determine the
average speed along the tacks
 C) then the link travel speed is estimated base on the GPS
position’s speed and the track average speed weighted
exponentially with the GPS time ‘distance’

Preliminary Analysis of the Estimated
Link Speeds
10
A preliminary analysis was undertaken to select the
appropriate prediction model and to identify the
candidate input variables
 The Rome Ring Road
case study

 FCD
travel speeds,
aggregated at 3-minute
periods (480 values per
link per day)
 January to April 2008
(penetration level: 2.4%)
Preliminary Analysis of the Estimated
Link Speeds
11

Spatio-temporal
traffic patterns can
be observed such
as
Morning peak hour
 Occurrence,
propagation and
dissipation of traffic
congestion

Preliminary Analysis of the Estimated
Link Speeds
12


The observable relationship among the FCD link travel
speeds of the neighboring links was further investigated
The cross-correlation
coefficient function ρk

Measuring the degree of
linear relationship
between random
variables at various time
lags
link 21
link 18
link 23
Approaches for Short-term Travel
Speed Predictions
13


Two algorithms, designed to on-line perform shortterm (15 to 30 minutes) predictions of link travel
speeds from FCD are presented
Pattern Matching & Artificial Neural Networks
 Take
into account spatial and temporal average speed
information simultaneously
Pattern Matching
14

Only categorical data are available to describe
the traffic speed (4 levels)
 Free (90 km/h up),
Conditioned (50-90 km/h), Slowed (3050 km/h), Congested (0-30 km/h)

Speed patterns for a specified link can be
constructed by lining up
 present
and past categorical speed values
 of the target and of the spatial correlated
upstream/downstream links
Pattern Matching – Speed Pattern ex.
15



Time step k represents the actual time
Ls the target link, Ls-1 and Ls+1 adjacent upstream and
downstream links
p, n1, n2 regression parameters
Pattern Matching
16

The assumption of time recurrence of traffic patterns
can enable a computational reduced searching
procedure
Scanning of all previous days in the historical database
within a time frame of ±x minutes from current time step k
 Evaluate in terms of their similarity to the current pattern
and chosen for the subsequent steps


Euclidean distance alone is not able to fully represents
similarity between two categorical time series

Similar trends/shapes could be better represented by
measures of rank correlation (Spearman, Kendall, Gini etc)
Pattern Matching
17


Use jointly the Euclidean distance En (0≦En≦1) and
the Spearman coefficient S (-1≦S≦1) to drive the
process of similarity-based selection among the
candidate pattern
Only the past speed patterns having both
 0≦Eni≦lEf
and 0≦(1-Si)≦lSf
 lEf = lSf ≒0.1
 Selection is also associated to a weighting procedure
 Weight
wni inversely proportional to
 0≦wni≦1 and Σiwni=1
Pattern Matching
18

Estimating the future

 si(k+1)
being the categorical speed value at time (k+1)
of the i-th selected past pattern

Free parameters all needing a careful tuning
 Upstream/downstream
links
 regression order p, n1, n2
 Interval limits lEf and lSf
 Trial and error procedure
Pattern Matching – Result
19



15 min ahead
prediction for
link 19
When past
examples are
absent or
insufficient, the
estimation
process fails or
can degrade
On all GRA
links reports
that average
misclassification
error: 18.7%
Artificial Neural Networks
20


Learn to associate input and output patterns
adaptively with the use of learning algorithms
without understanding the fundamental or physical
relationships between them
Feed forward ANNs comprise
an input layer, one or more
hidden layers and an output
layer, furthermore each layer
contains different number of
units (neurons)
Artificial Neural Networks
21



As a rule, the type and # of units in the input layer and
hidden layers, are chosen through a preliminary
analysis of the data or by empirically comparing the
results from different ANN architectures
The relations between neighboring layers’ units are
defined by the weight given to the connections during
the training process
The output of each unit is given by a transfer function
fed up with the weighted sum of the incoming units
values and then transmitted to all of the units in the next
layer
Artificial Neural Networks
22



The two ANN models, aimed at prediction the link travel
speed respectively at 5 and 10 steps into the future
Incorporated as input the current and the near-past 10
and 15 FCD travel speeds of the target link,
respectively
Moreover the two ANN models considered as input the
current and at most the near-past 10 FCD travel speeds
of the immediate neighboring upstream and
downstream links

Correlation degree higher than 0.6
Artificial Neural Networks
23

FCD link travel speeds stored from 7:00 am to 9:00 pm
were used in the learning-testing process (forty-seven
working days)
Mean absolute percentage error (MAPE) and the root
mean square error (RMSE) were calculated for
investigating the accuracy of the model in the testing
process

N is the total # of testing case

Artificial Neural Networks – Results
24
MAPE
RMSE
15-min prediction
2% ~ 8%
2km ~ 7 km
30-min prediction
3% ~ 16%
3.5km ~ 9.5km
Conclusions
25


The use of large scale real-time FCD is gaining an
important role as component of ATIS applications
because of its cost-effectiveness
A fundamental requirement is its statistical consistency
that can be assured only when
# of monitored cars achieves a significant penetration level
 Car transmission rate is adequate



According to field tests, the accuracy of the LSFCD
System in estimating current link travel speeds is about
90%
Two method were developed for short-term speed
predictions
Comments
26


Take into account spatial and temporal information is
good and intuitive
Condition of motorway and road (or street) is
different
#
of lanes
 Crossroads and traffic lights