Transcript No Slide Title

Development of Capability-Enhanced
PARAMICS Simulation Environment
Lianyu Chu, Henry X. Liu, Will Recker
California PATH ATMS Center
University of California, Irvine
1
Introduction
• Microscopic simulation:
– a software tool to model traffic system, including
roads, drivers, and vehicles, in fine details.
• Micro-simulation models
– CORSIM, PARAMICS, VISSIM, AIMSUN2…
• Micro-simulation applications
– Model traffic network
– Model and Evaluate ITS
– Calibrate / optimize operational parameters of ITS
strategies
– Develop and test new traffic models, algorithms,
control strategies
2
Introduction
• Why capability enhancements are needed
– Complement missing functionalities of the current
model
– Modeling and evaluating ITS
• How to enhance
– Working on source codes
• proprietary of a software
– API Programming: a practical way for users
• API -- Application Programming Interface
• Requirement: an API library provided by micro-simulator
3
Role of API
User
Input Interface
Professional
Community
Oversight
Core Model
API
Output Interface
GUI
Tools
Developer
*source: from FHWA website
4
PARAMICS
• PARAMICS: PARAllel MICroscopic Simulation
• Developer: Quadstone, Scotland
• Features:
– Large network simulation capability
– Good graphical interface
– Modeling the emerging ITS infrastructures
– API Programming
• Purpose of this paper:
– Enhancing capabilities of PARAMICS via API in
order to better model and evaluate ITS
5
PARAMICS
6
How PARAMICS API works
Callback
Override
Overload
N
End of
simulation ?
Advanced API
modules
Update vehicles and
their environment at
every time step:
User-developed
basic API modules
Start
Other applications
e.g. database
Y
Stop
7
Aspects of PARAMICS need to be
enhanced
• Path-based Routing (a missing function)
– PARAMICS: Link-based routing
• Traffic controller:
– Actuated signal control (a missing function)
– Ramp metering control (functionality enhancement for
facilitating ITS modeling)
• Real-time traffic information (a missing function)
– including the collection and aggregation of loop data and
probe vehicle data
• Database connection (functionality enhancement)
• Performance measure (functionality enhancement)
8
Current framework of enhanced
PARAMICS
Basic Modules:
Actuated
Signal
Dynamic Linking
Path-based
Routing
Commercial
Paramics Model
Loop data
Aggregator
Interface
functions
MOE
Dynamic Linking
Probe
vehicle
MySQL Database
Ramp
Metering
Advanced Modules
9
PARAMICS API Development:
A Hierarchical Approach
Signal
Provided
API Library
Control
Modules
Ramp
Routing
Loop …
Basic
modules
CORBA
Data
Handling
Database
XML …
Adaptive Signal Control
Advanced
Modules
Ramp metering algorithms
Integrated Control
10
Basic modules:
Path-based routing
• Required for the simulation of driver responses to
the information supply and resulting route choice
• Logic: specified vehicles follow a given path
• Interface function:
void uci_vehicle_route_set (void *Vp, VROUTE route)
Parameters: Vp: the pointer of a vehicle, equivalent to vehicle ID
route is the initial address of the whole path
VROUTE is a link list storing whole path the vehicle
should follow, which is defined as:
type VROUTE
{
// link name along the route
char *linkName;
VROUTE *next;
};
11
Basic modules:
Full-actuated signal controller
•
NEMA Dual-ring, 8-phase logic:
8
Approach 2
1+5
7
6
1+6
5
2+5
Approach 1
4
1
7
2
3
4
6
2+6
1
5
3+7
2
12
3+8
4+7
11 10
3
9
8
NEMA Phase
Detector number
Approach 3
13
14
Detector
15
4+8
Approach 4
16
12
Basic modules:
Ramp metering controller
•
•
•
•
•
•
op
St
Operation: n-cars-per-green basis (with n >= 1)
Pre-timed control
Support multiple timing plans
Support HOV bypass
Support the use of ramp
detectors for metering
Mainline Traffic
control
Interface functions for
external algorithm
modules to control
or
metering
ct
e
t
ne
La
de
13
Basic modules:
Loop data aggregator
• Emulating the real-world loop data collection
– Aggregated loop data are used for real-time traffic
control and off-line analysis
• Inputs:
– loop detector name
– aggregation interval (typically with 30 second
interval)
• Outputs:
– Data broadcasting to shared memory, accessible
through interface functions
– Reported to text files or database for performance
analysis
14
Basic modules:
Probe vehicle
• Emulating section travel time data collection through
GPS equipped vehicles
• Inputs:
– Sections of travel time collection
– Sample rate of equipped vehicles
– Aggregation interval
• Outputs:
– Aggregated data broadcasting to shared memory,
accessible through interface functions
– Reported to text files or database for performance
analysis
15
Basic modules:
Database connection interface
• MYSQL: most popular and highly efficient Open
Source SQL database
• This interface is developed using API functions of
MYSQL. This interface allows us to operate the
database via our developed PARAMICS API codes.
• Purposes of this module:
– Storing intermediate data during simulation and
simulation results
– Exchange data with other API modules / outside
programs
16
Basic modules:
Performance measures
• PARAMICS has strong abilities on the collection of
statistics / measurement data
• Current difficulties are:
– With the increase of network size, the number of links, trips,
intersections, and detectors increases drastically
– Large amount of data are required to be processed after
simulation runs for obtaining expected MOE.
– Some MOE, such as on-ramp waiting time, cannot be
extracted from output measurement data.
– PARAMICS has a restriction on the number of output files to
be opened during simulation under WINDOWS version.
• Use API to collect some generic overall MOE for
evaluating ITS
17
Basic modules:
Performance measures
• System performance
– System efficiency: Vehicle miles traveled (VHT)
– System reliability: standard deviation of average
OD travel times
• Freeway performance
– Mainline: average freeway travel speed and its
variance
– Ramp: total on-ramp delay and on-ramp queue
spillback percentage
• Arterial performance
– Average travel time of an arterial and its variance
18
Advanced modules:
How to develop an advanced module
New rate
PARAMICS
simulation
Provided API
Ramp metering
Controller
New rate
Loop Data
Aggregator
Old metering rate
Basic API modules
Loop data
Advanced
ramp-metering algorithms
Advanced API modules
19
Advanced modules:
How to develop an advanced module
• Interfaces with ramp metering controller
– Set a new metering rate:
void ramp_set_parameters
(RAMP *ramp, Bool status)
– Get current metering rate:
RAMP *ramp_get_parameters
(char *rampNode)
• Interface with loop data aggregator:
– LOOPAGG loop_agg
(char *detectorName)
type Ramp
{
// ramp signal name and location
char *rampNode;
char *controllerLocation;
// ramp control types and parameters
int ControlType;
float meteringCycle;
};
type LOOPAGG
{
intdetectorIndex;
floatAggregationTime;
intlane;
intg_vol;
floatg_occ;
floatg_spd;
int*vol;
float*occ;
float*spd; };
20
Developed advanced modules
• Signal control
– Adaptive signal
– Actuated signal coordination
• Adaptive ramp metering algorithms
– ALINEA
– ZONE
– BOTTLENECK
– SWARM
• PARAMICS-DYNASMART
• Demand-responsive Transit
21
Sample applications
• Signal
– Hardware-in-loop, testing 170 controller
– Adaptive signal control based on real-time delay
estimation
• Ramp metering
– Evaluating adaptive ramp metering algorithms
– Caltrans Freeway improvement projects
• I-10 freeway
• I-680 freeway
• Caltrans’ Traffic Management System master plan
– Evaluating potential ITS strategies
22
Current API user groups
• UC Irvine
• Caltrans:
– Transportation planning
– Traffic operation
• California PATH program headquarter at Berkeley
• UC Berkeley
• UC Davis
• National University of Singapore
• Consultant companies:
– Dowling Associates
– Cambridge Systematics
23
Conclusions
• API can access the core models of micro-simulation.
• PARAMICS is a good simulation shell that users can
replace its algorithms through API.
• We enhanced capabilities of PARAMICS and used the
enhanced PARAMICS for modeling and evaluating
ITS.
• The developed APIs have been used by Caltrans and
our research & project partners in many projects.
• The same enhancement mechanism can be
applicable to other micro-simulators, such as
VISSIM, AIMSUN2, etc.
24
More information
• PCTSS website:
http://www.its.uci.edu/~paramics/
• PATH website:
http://www.path.berkeley.edu/
• Contact: PATH ATMS Center @ UC Irvine
– Lianyu Chu: [email protected]
– Henry Liu: [email protected]
– Will Recker: [email protected]
25