{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "
\n", "\n", "# Radiation Source Localization\n", "Author(s): Paul Miles, Jason Hite | Date Created: July 19, 2019\n", "\n", "This example demonstrates how MCMC methods can be used to locate unknown sources of gamma radiation in an urban environment. For this demonstration we consider a simulated 250m x 178m block of downtown Washington D.C.\n", "\n", "We utilize a highly simplified radiation transport model which ignores scattering. The model consists of the following components. There are $N_d$ gamma detectors located at positions $\\mathbf{r}_i, i = 1, ..., N_d$. A ray tracing algorithm is implemented for a source located at $\\mathbf{r}_0 = (x_0, y_0)$ with intensity $I_0$. Buildings affect the readings of the gamma detectors, so we have modeled their impact using a exponential decay function that depends on the building dimensions and average cross-section. For the $i^{th}$ detector, we let $N_i$ denote the number of buildings between $\\mathbf{r}_0$ and $\\mathbf{r}_i$. For $n_i = 1,...,N_i$, we let $s_{n_i}$ and $\\sigma_{n_i}$ denote the length of the building segment and the average cross-section of the $n_i^{th}$ building, respectively. The surface area, efficiency, and measurement duration for the $i^{th}$ detector are denoted by $A_i, \\varepsilon_i,$ and $\\Delta t_i$, respectively. Finally, $B_i$ represents the expected background count rate at position $\\mathbf{r}_i$. \n", "$$\n", "\\Gamma_i = \\frac{\\Delta t_i\\varepsilon_iA_iI_0}{4\\pi|\\mathbf{r}_i - \\mathbf{r}_0|^2}\\exp\\Big(-\\sum_{n_i=1}^{N_i}\\sigma_{n_i}s_{n_i}\\Big) + E[B_i\\Delta t_i]\n", "$$\n", "\n", "This model is implemented in the open source Python package [gefry3](https://github.com/jasonmhite/gefry3). The input deck for this example can be found [here](https://github.com/jasonmhite/gefry3/blob/master/examples/g3_deck.yml). For additional reading on this area of research, please refer to the following articles:\n", "\n", "- Hite, J., & Mattingly, J. (2018). Bayesian Metropolis Methods for Source Localization in an Urban Environment. Radiation Physics and Chemistry. [https://doi.org/10.1016/j.radphyschem.2018.06.024](https://doi.org/10.1016/j.radphyschem.2018.06.024)\n", "- Ştefănescu, R., Schmidt, K., Hite, J., Smith, R. C., & Mattingly, J. (2017). Hybrid optimization and Bayesian inference techniques for a non‐smooth radiation detection problem. International Journal for Numerical Methods in Engineering, 111(10), 955-982. [https://doi.org/10.1002/nme.5491](https://doi.org/10.1002/nme.5491)\n", "- Hite, J. M., Mattingly, J. K., Schmidt, K. L., Ştefănescu, R., & Smith, R. (2016, September). Bayesian metropolis methods applied to sensor networks for radiation source localization. In Multisensor Fusion and Integration for Intelligent Systems (MFI), 2016 IEEE International Conference on (pp. 389-393). IEEE. [https://doi.org/10.1109/MFI.2016.7849519](https://doi.org/10.1109/MFI.2016.7849519)\n", "\n", "__Acknowledgment:__\n", "Funding for this research provided by the [Consortium for Nonproliferation Enabling Capabilities (CNEC)](https://cnec.ncsu.edu/).\n", "\n", "__Package Installation:__\n", "All packages can be install via\n", "```python\n", "pip install package_name\n", "```\n", "with the exception of [gefry3](https://github.com/jasonmhite/gefry3), which requires\n", "```python\n", "pip install git+https://github.com/jasonmhite/gefry3\n", "```\n", "Advanced statistical plots are generated using the [mcmcplot](https://prmiles.wordpress.ncsu.edu/codes/python-packages/mcmcplot/) package.\n", "\n", "To run this particular notebook, please execute the two cells below. The first cell will install [gefry3](https://github.com/jasonmhite/gefry3) and the second [descartes](https://pypi.org/project/descartes/)." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "pip install git+https://github.com/jasonmhite/gefry3" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "pip install descartes" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "# import required packages\n", "import os\n", "import gefry3\n", "import numpy as np\n", "from pymcmcstat.MCMC import MCMC\n", "from descartes.patch import PolygonPatch\n", "import seaborn as sns\n", "import matplotlib.pyplot as plt\n", "np.seterr(over = 'ignore')\n", "NOBS = 1 # number of observations for each detector\n", "NS = int(1e3) # number of MCMC simulations" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Define Radiation Model, Problem Domain, and Background Radiation" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[1522. 1544. 1475. 1682. 1602. 1517. 1546. 1769. 1664. 1668.]]\n", "S0 = [158. 98.] m, I0 = 3214000000.0 Bq\n" ] } ], "source": [ "# Setup Radiation Model - Downtown Urban Environment Simulation\n", "P = gefry3.read_input_problem('data_files' + os.sep + 'g3_deck.yml', 'Simple_Problem')\n", "# Setup background\n", "nominal_background = 300\n", "dwell = np.array([detector.dwell for detector in P.detectors])\n", "B0 = nominal_background * dwell\n", "# True source and intensity\n", "S0 = P.source.R # m\n", "I0 = P.source.I0 # Bq\n", "nominal = P(S0, I0)\n", "ndet = len(P.detectors)\n", "# Define observations\n", "observations = np.random.poisson(\n", " nominal + nominal_background * dwell,\n", " size=(\n", " NOBS,\n", " ndet,\n", " ),\n", ")\n", "observations = observations.astype(np.float64)\n", "# Setup parameter bounds\n", "XMIN, YMIN, XMAX, YMAX = P.domain.all.bounds\n", "IMIN, IMAX = I0 * 0.1, I0 * 10\n", "XMIN, YMIN, XMAX, YMAX = P.domain.all.bounds\n", "print(observations)\n", "print('S0 = {} m, I0 = {} Bq'.format(S0, I0))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Define Sum-of-Squares Error Function." ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "# Radiation Sum of Squares Function\n", "def radiation_ssfun(theta, data):\n", " x, y, I = theta\n", " model, background = data.user_defined_object[0]\n", " output = model((x, y), I) + background\n", "\n", " res = data.ydata[0] - output\n", " ss = (res ** 2).sum(axis = 0)\n", " return ss" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Initialize MCMC Object and Setup Simulation\n", "Note, we pass the radiation model into our residual function via the data structure. Also note, the background is included as a parameter, but it is not sampled during the simulation." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "# Initialize MCMC object\n", "mcstat = MCMC()\n", "# setup data structure for dram\n", "mcstat.data.add_data_set(\n", " x=np.zeros(observations.shape),\n", " y=observations,\n", " user_defined_object=[\n", " P,\n", " B0,\n", " ],\n", ")\n", "mcstat.parameters.add_model_parameter(\n", " name='$x_0$',\n", " theta0=140.,\n", " minimum=XMIN,\n", " maximum=XMAX,\n", ")\n", "mcstat.parameters.add_model_parameter(\n", " name='$y_0$',\n", " theta0=85.,\n", " minimum=YMIN,\n", " maximum=YMAX,\n", ")\n", "mcstat.parameters.add_model_parameter(\n", " name='$I_0$',\n", " theta0=3.0e9,\n", " minimum=IMIN,\n", " maximum=IMAX\n", ")\n", "\n", "mcstat.simulation_options.define_simulation_options(\n", " nsimu=NS,\n", " updatesigma=False,\n", " method='dram',\n", " adaptint=100,\n", " verbosity=1,\n", " waitbar=1,\n", " save_to_json=False,\n", " save_to_bin=False,\n", ")\n", "\n", "mcstat.model_settings.define_model_settings(\n", " sos_function=radiation_ssfun,\n", " sigma2=observations.sum(axis=0),\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Run Simulation" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "Sampling these parameters:\n", " name start [ min, max] N( mu, sigma^2)\n", " $x_0$: 140.00 [ 0.00e+00, 246.62] N( 0.00e+00, inf)\n", " $y_0$: 85.00 [ 0.00e+00, 176.33] N( 0.00e+00, inf)\n", " $I_0$: 3.00e+09 [ 3.21e+08, 3.21e+10] N( 0.00e+00, inf)\n", " [-----------------100%-----------------] 1000 of 1000 complete in 66.6 sec" ] } ], "source": [ "# Run mcmcrun\n", "mcstat.run_simulation()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Process Results and Display Statistics" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "\n", "------------------------------\n", "name : mean std MC_err tau geweke\n", "$x_0$ : 160.6402 2.7195 0.3393 10.0607 0.9957\n", "$y_0$ : 98.9279 1.8359 0.1848 6.0494 0.9936\n", "$I_0$ : 2.406e+09 3.638e+08 48017969.7110 11.4787 0.9392\n", "------------------------------\n", "==============================\n", "Acceptance rate information\n", "---------------\n", "Results dictionary:\n", "Stage 1: 20.30%\n", "Stage 2: 56.10%\n", "Net : 76.40% -> 764/1000\n", "---------------\n", "Chain provided:\n", "Net : 80.00% -> 400/500\n", "---------------\n", "Note, the net acceptance rate from the results dictionary\n", "may be different if you only provided a subset of the chain,\n", "e.g., removed the first part for burnin-in.\n", "------------------------------\n", "Acceptance rate: 76.4%\n", "Model Evaluations: 1797\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "