These computer codes are free software provided by SSRG under the terms of the GNU General Public License (http://www.gnu.org/licenses/).
Estimating Post-disaster Traffic Conditions Using Real-time Data Streams
- Developer: Reece P. Otsuka (formerly a graduate student at the University of Illinois at Urbana-Champaign)
- What it does: Ensemble Kalman filter that can estimate post-disaster traffic conditions using real-time data streams even when traffic estimation model is subject to significant change caused by an earthquake event
- Reference: Otsuka, R.P., D.B. Work, and J. Song, Estimating post-disaster traffic conditions using real-time data streams. Structure and Infrastructure Engineering: Maintenance, Management, Life-Cycle Design and Performance. Vol. 12(8), 904-917, 2016.
- Download: HERE
- How to use: Download the package from the GitHub page and follow the instructions.
Accelerated Monte Carlo System Reliability Analysis through Machine-Learning-based Surrogate Models of Network Connectivity
- Developer: Raphael Stern (currently a graduate student at the University of Illinois at Urbana-Champaign)
- What it does: Develop surrogate models for network connectivity using support vector machine and logistic regression for accelerated Monte Carlo simulations.
- Reference: Stern, R.E., J. Song, and D.B. Work, Accelerated Monte Carlo system reliability analysis through machine learning-based surrogate models of network connectivity. Reliability Engineering & System Safety. Vol. 164, 1-9, 2017.
- Download: HERE
- How to use: Download the package from the GitHub page and follow the instructions.
Cross-entropy-based Adaptive Importance Sampling Using von Mises-Fisher Mixture for High Dimensional Reliability Analysis
- Developer: Ziqi Wang (currently at Earthquake Engineering Research & Test Center, Guangzhou University, Guanzhou, China)
- What it does: Perform adaptive importance sampling for component and system reliability problems using the “Cross-entropy-based Adaptive Importance Sampling Using von Mises-Fisher Mixture for High Dimensional reliability analysis” (Wang and Song 2016).
- Purpose of development: International collaborative research on high dimensional reliability problem
- Reference: Wang, Z., and J. Song, Cross-entropy-based adaptive importance sampling using von Mises–Fisher mixture for high dimensional reliability analysis. Structural Safety. Vol. 59, 42-52, 2016.
- Download: HERE
- How to use: Unzip the file into a local folder and set paths. Run the “TestExample.m” to reproduce the results of 5.2 in the reference.
Read “ReadMe.txt” for more details.
Cross-entropy-based Adaptive Importance Sampling Using Gaussian Mixture
- Developer: Ryan H.Y. Wong (currently working at AECOM in Hong Kong)
- What it does: Perform adaptive importance sampling for component and system reliability problems using the “Cross-entropy-based Adaptive Importance Sampling Using Gaussian Mixture” (Kurtz and Song 2013).
- Purpose of development: Undergraduate research to develop free computer codes for state-of-the-art algorithms (sponsored by the Department of Civil and Environmental Engineering at UIUC).
- Reference: N. Kurtz and J. Song, “Cross-entropy-based Adaptive Importance Sampling Using Gaussian Mixture,” Structural Safety, Vol. 42, pp. 35-44, 2013.
- Download: HERE (The “accepted author manuscript” of the paper is included)
- How to use: Unzip the file into a local folder and set paths. Run an input file (see the examples and input file template). Then, run “ceaisgm.m”. Read “CEAISGM.txt” for more details.
Sequent
ial Conditioned Importance Sampling
- Developers: Young Joo Lee (currently at Ulsan National Institute of Science and Technology) and Junho Song
- What it does: Evaluates multinormal probabilities using the “Sequential Conditioned Importance Sampling” (Ambartzumian et al. 1998) through a vectorized (i.e. improved efficiency) Matlab code.
- Purpose of development: Research on identifying critical sequence of failures induced by fatigue crack growth and quantifying the risk of system collapse.
- Reference: R. Ambartzumian, A. Der Kiureghian, V.Ohanian, and H.Sukiasian, “Multinormal Probability by Sequential Conditioned Importance Sampling,” Probabilistic Engineering Mechanics, Vol.13, No.4, pp.299-308, 1998.
- Download: HERE
FERUM toolbox for Matrix-based System Reliability (MSR) Analysis
- Developer: Bora Gencturk (currently at University of Houston)
- What it does: Computes the system failure probability and its parameter sensitivities (with respect to parameters that do not affect the correlation coefficients between components) using the MSR method (Song and Kang, 2009)
- How to use: (a) Standalone mode: the user provides the component failure probabilities (and sensitivities) and correlation coefficient matrix for direct MSR analysis; and (b) FERUM toolbox mode: the user provides input data for first order component reliability analyses by FERUM. The code collects the calculated component failure probabilities (and sensitivities) and calculate the correlation coefficient matrix to perform MSR analysis. A detailed manual written by the developer (Bora Gencturk) is enclosed.
- Purpose of development: Final term project of CEE491: Decision and Risk Analysis at the University of Illinois, Urbana-Champaign (Spring 2008).
- Disclaimer: This is an open-source code provided to promote education and research, and may contain methods under further development.
- Reference: J. Song, and W.H. Kang, “System Reliability and Sensitivity under Statistical Dependence by Matrix-based System Reliability Method,” Structural Safety, Vol.31, No.2, pp.148-156, 2009.
- Download: HERE
More free computer codes are coming.
