#############################################################
Note that, this version is for the reproducibility of
our preprint, not for that of the newest submission.
#############################################################
This is an instruction for reproducing the experiments described in our preprint.
A video demonstration of our simulation is available at https://youtu.be/f2FpVc17SBA.
===== Part A =====
What the experimental environments we need? ->
Operation system: Ubuntu 18.04
Ruby version: 2.5.1
GCC version: 7.3.0
SUMO version: 0.32
Install SUMO in Ubuntu:
$ sudo apt-get install sumo sumo-tools sumo-doc
Setup QMaxSAT solver with incremental approach
$ cd ./source_code/sumoutil/QMaxSAT_ncc/code/
$ make clean && make
$ cp ./qmaxsatNcc_g3 ../../encodedProblem/qmaxsatNcc_g3
===== Part B =====
How to run experiments (described in Section 5)? ->
Firstly,
$ cd ./source_code/sumoutil/sample/001.Tsukuba
we can simulate an experiment for SBI allocation (existing method)
$ ./runSBI.sh
or, we can simulate an experiment for SAT-based allocation (proposed approach)
$ ./runSAT.sh
All output files are exported to the following directories:
./source_code/sumoutil/expDir/genrdMaxInst/
./source_code/sumoutil/expDir/maxSATime/
./source_code/sumoutil/expDir/seqOpTime/
All log files are exported to the directory: ./source_code/sumoutil/sample/001.Tsukuba/,Log/
Do not forget to remove the generated files (including encode files, log files, answer files)
before the next running of ./runSBI.sh or ./runSAT.sh
$ ./reset.sh
The default parameter settings:
the number of taxis = 20
the demand occurrence frequency = 3600/100 (interval = 100)
...
We can change any test value (integer) in the following files:
./source_code/sumoutil/sample/001.Tsukuba/tsukuba.00.savSimConf.json (taxi's information/parameters)
./source_code/sumoutil/sample/001.Tsukuba/tsukuba.00.demandConf.json (demand's information/parameters)
===== Part C =====
How to simulate for the real-world data (mentioned in Appendix D)? ->
Firstly,
$ cd ./source_code/sumoutil/sample
we need to build the symbolic links
$ ln -s ../Savs ./Savs && ln -s ../Tools ./Tools && ln -s ../Traci ./Traci
Then,
$ cd ./2018.1005.Yokohama
we can simulate an experiment for SBI allocation (existing method)
$ ./runSBI.sh
or, we can simulate an experiment for SAT-based allocation (proposed approach)
$ ./runSAT.sh
All output files are exported to the following directories:
./source_code/sumoutil/expDir/genrdMaxInst/
./source_code/sumoutil/expDir/maxSATime/
./source_code/sumoutil/expDir/seqOpTime/
All log files are exported to the directory: ./source_code/sumoutil/sample/2018.1005.Yokohama/,Log/
Do not forget to remove the generated files (including encode files, log files, answer files)
before the next running of ./runSBI.sh or ./runSAT.sh
$ ./reset.sh
The default parameter settings:
the date of the imported real-world data = 2018-12-01
...
We can change any test value (integer) in the following files:
./source_code/sumoutil/sample/2018.1005.Yokohama/yokohamaNedo.02.savSimConf.json (taxi's information/parameters)
./source_code/sumoutil/sample/2018.1005.Yokohama/yokohamaNedo.02.demandConf.json (demand's information/parameters)
===== Part D =====
A detected algorithm core with a sample initialization is given in directory
./source_code/sat-based_rtss_v1.1/