aula02AndreUFJF.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "aula02AndreUFJF",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_2wNTPVNo935"
      },
      "source": [
        "# AULA 02 - PLANEJAMENTO DE SISTEMAS ELÉTRICOS\n",
        "\n",
        "## PPEE - UFJF\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3usgXHjesgLi"
      },
      "source": [
        "## Dados do Problema\n",
        "\n",
        "UHE, UTE e Dados Gerais"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o-8M0ziJpWSu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "c09095d5-e042-40da-bb58-a91fe163b9e5"
      },
      "source": [
        "lista_uhe = []\n",
        "lista_ute = []\n",
        "\n",
        "usina = {\n",
        "    \"Nome\": \"UHE DO MARCATO\",\n",
        "    \"Vmax\": 100.,\n",
        "    \"Vmin\": 20.,\n",
        "    \"Prod\": 0.95,\n",
        "    \"Engol\": 60.,\n",
        "    \"Afl\": [\n",
        "            [23,16],\n",
        "            [19,14],\n",
        "            [15,11]\n",
        "    ]\n",
        "}\n",
        "\n",
        "lista_uhe.append(usina)\n",
        "\n",
        "usina = {\n",
        "    \"Nome\": \"UHE DO MENGAO\",\n",
        "    \"Vmax\": 200.,\n",
        "    \"Vmin\": 40.,\n",
        "    \"Prod\": 0.85,\n",
        "    \"Engol\": 100.,\n",
        "    \"Afl\": [\n",
        "            [46,32],\n",
        "            [38,28],\n",
        "            [30,22]\n",
        "    ]\n",
        "}\n",
        "\n",
        "#lista_uhe.append(usina)\n",
        "\n",
        "usina = {\n",
        "    \"Nome\": \"GT_1\",\n",
        "    \"Capac\": 15.,\n",
        "    \"Custo\": 10.\n",
        "}\n",
        "\n",
        "lista_ute.append(usina)\n",
        "\n",
        "usina = {\n",
        "    \"Nome\": \"GT_2\",\n",
        "    \"Capac\": 10.,\n",
        "    \"Custo\": 25.\n",
        "}\n",
        "\n",
        "lista_ute.append(usina)\n",
        "\n",
        "dgerais = {\n",
        "    \"CDef\": 500.,\n",
        "    #\"Carga\": [100,100.,100.],\n",
        "    \"Carga\": [50,50.,50.],\n",
        "    \"Nr_Disc\": 12,\n",
        "    \"Nr_Est\": 3,\n",
        "    \"Nr_Cen\": 2\n",
        "}\n",
        "\n",
        "sistema = {\n",
        "    \"DGer\": dgerais,\n",
        "    \"UHE\": lista_uhe,\n",
        "    \"UTE\": lista_ute\n",
        "}\n",
        "\n",
        "print(sistema[\"UTE\"][0][\"Nome\"])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "GT_1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bqJzsQBbstZm"
      },
      "source": [
        "## Modelo de Otimização\n",
        "\n",
        "Criação de uma função para o despacho hidrotérmico\n",
        "\n",
        "$Min \\hspace{0.5 cm} C_1 \\cdot gt_1 + C_2 \\cdot gt_2 + CDEF \\cdot def + 0.01 \\cdot vvert$\n",
        "\n",
        "Sujeito a:\n",
        "\n",
        "$vf = VI + AFL - vtur - vvert$ (Equação de balanço hídrico)\n",
        "\n",
        "$\\rho \\cdot vtur + gt_1 + gt_2 + def = CARGA$ (Equação de atendimento à demanda)\n",
        "\n",
        "$20 \\le vf \\le 100$\n",
        "\n",
        "$0 \\le vtur \\le 60$\n",
        "\n",
        "$0 \\le vvert \\le \\infty$\n",
        "\n",
        "$0 \\le gt_1 \\le 15$\n",
        "\n",
        "$0 \\le gt_2 \\le 10$\n",
        "\n",
        "$0 \\le def \\le \\infty$\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xDCZdA1ezVRK"
      },
      "source": [
        "# RESOLVENDO O PROBLEMA DE OTIMIZACAO"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7vmiPjhUwb3S"
      },
      "source": [
        "from cvxopt.modeling import variable, solvers\n",
        "from cvxopt.modeling import op\n",
        "\n",
        "def despacho(sistema, VI, AFL, pote_de_corte, iest, imprime):\n",
        "\n",
        "  Num_UHE = len(sistema[\"UHE\"])\n",
        "\n",
        "  Num_UTE = len(sistema[\"UTE\"])\n",
        "\n",
        "  vf = variable(Num_UHE, \"Volume final da usina\")\n",
        "  vt = variable(Num_UHE, \"Volume turbinado da usina\")\n",
        "  vv = variable(Num_UHE, \"Volume vertido da usina\")\n",
        "\n",
        "  gt = variable(Num_UTE, \"Geracao na UTE\")\n",
        "  deficit = variable(1, \"Deficit de energia no sist\")\n",
        "\n",
        "  alpha = variable(1,\"Custo Futuro\")\n",
        "\n",
        "  # Construcao da f obj\n",
        "\n",
        "  fob = 0\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UTE\"]):\n",
        "    fob += iusi[\"Custo\"]*gt[i]\n",
        "\n",
        "  fob += sistema[\"DGer\"][\"CDef\"]*deficit[0]\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "    fob += 0.01*vv[i]\n",
        "\n",
        "  fob += 1.0*alpha[0]\n",
        "\n",
        "  #Def das restricoes\n",
        "\n",
        "  restricoes = []\n",
        "\n",
        "  #balanco hidrico\n",
        "  for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "    restricoes.append( vf[i] == float(VI[i]) + float(AFL[i]) - vt[i] - vv[i])\n",
        "\n",
        "  # atendimento a demanda\n",
        "\n",
        "  AD = 0\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "    AD += iusi[\"Prod\"]*vt[i]\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UTE\"]):\n",
        "    AD += gt[i]\n",
        "\n",
        "  AD += deficit[0]\n",
        "\n",
        "  restricoes.append(AD == sistema[\"DGer\"][\"Carga\"][iest-2])\n",
        "\n",
        "  # Restricoes de canalizacao\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "    restricoes.append( vf[i] >= iusi[\"Vmin\"])\n",
        "    restricoes.append( vf[i] <= iusi[\"Vmax\"])\n",
        "    restricoes.append( vt[i] >= 0)\n",
        "    restricoes.append( vt[i] <= iusi[\"Engol\"])\n",
        "    restricoes.append( vv[i] >= 0)\n",
        "\n",
        "  for i,iusi in enumerate(sistema[\"UTE\"]):\n",
        "    restricoes.append( gt[i] >= 0)\n",
        "    restricoes.append( gt[i] <= iusi[\"Capac\"])\n",
        "\n",
        "  restricoes.append(deficit[0] >= 0)\n",
        "\n",
        "  restricoes.append(alpha[0] >= 0)\n",
        "\n",
        "  for icorte in pote_de_corte:\n",
        "    if icorte[\"Estagio\"]==iest:\n",
        "      equacao = 0\n",
        "      for iusi in range(Num_UHE):\n",
        "        equacao += float(icorte[\"Coefs\"][iusi])*vf[iusi]\n",
        "      equacao += float(icorte[\"Termo_Indep\"])\n",
        "      restricoes.append(alpha[0] >= equacao)\n",
        "\n",
        "  problema = op(fob,restricoes)\n",
        "\n",
        "  problema.solve('dense','glpk')\n",
        "\n",
        "  Dger = {\n",
        "      \"Deficit\": deficit[0].value()[0],\n",
        "      \"CMO\": restricoes[Num_UHE].multiplier.value[0],\n",
        "      \"Custo total\": fob.value()[0],\n",
        "      \"Custo futuro\": alpha[0].value()[0]\n",
        "  }\n",
        "\n",
        "  lista_uhe = []\n",
        "\n",
        "  for i, iusi in enumerate(sistema[\"UHE\"]):\n",
        "    resultado = {\n",
        "        \"vf\": vf[i].value()[0],\n",
        "        \"vt\": vt[i].value()[0],\n",
        "        \"vv\": vv[i].value()[0],\n",
        "        \"cma\": restricoes[i].multiplier.value[0]\n",
        "    }\n",
        "    lista_uhe.append(resultado)\n",
        "\n",
        "  lista_ute = []\n",
        "\n",
        "  for i, iusi in enumerate(sistema[\"UTE\"]):\n",
        "    resultado = {\n",
        "        \"gt\": gt[i].value()[0]\n",
        "    }\n",
        "    lista_ute.append(resultado)\n",
        "\n",
        "  resultado = {\n",
        "      \"DGer\": Dger,\n",
        "      \"UHE\": lista_uhe,\n",
        "      \"UTE\": lista_ute\n",
        "  }\n",
        "\n",
        "  if imprime:\n",
        "    print(\"Custo total:\",fob.value())\n",
        "\n",
        "    for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "      print(vf.name,i,\"é\",vf[i].value(),\"hm3\")\n",
        "      print(vt.name,i,\"é\",vt[i].value(),\"hm3\")\n",
        "      print(vv.name,i,\"é\",vv[i].value(),\"hm3\")\n",
        "\n",
        "    for i,iusi in enumerate(sistema[\"UTE\"]):\n",
        "      print(gt.name,i,\"é\",gt[i].value(),\"MWmed\")\n",
        "\n",
        "    print(deficit.name,\"é\",deficit[0].value(),\"MWmed\")\n",
        "\n",
        "    print(alpha.name,\"é\",alpha[0].value(),\"R$\")\n",
        "\n",
        "    for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "      print(\"O valor da água na usina\",i,\"é\",restricoes[i].multiplier.value)\n",
        "\n",
        "    print(\"O Custo Marginal de Operacao é\",restricoes[Num_UHE].multiplier.value, \"R$/MWh\")\n",
        "    print(\"\\n###########################################################\\n\")\n",
        "\n",
        "  return(resultado)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IXnlVI3axvDt",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "fa4772b0-4292-4c94-c2ee-41875530362b"
      },
      "source": [
        "from itertools import product\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from matplotlib import cm\n",
        "import time\n",
        "\n",
        "t = time.time()\n",
        "\n",
        "Num_UHE = len(sistema[\"UHE\"])\n",
        "\n",
        "Num_UTE = len(sistema[\"UTE\"])\n",
        "\n",
        "passo = 100/(sistema[\"DGer\"][\"Nr_Disc\"]-1)\n",
        "\n",
        "lista = np.arange(0,100+passo,passo)\n",
        "\n",
        "discretizacoes = product(lista,repeat=Num_UHE)\n",
        "\n",
        "discretizacoes = list(discretizacoes)\n",
        "\n",
        "pote_de_corte = []\n",
        "\n",
        "for iest in np.arange(sistema[\"DGer\"][\"Nr_Est\"],0,-1):\n",
        "\n",
        "  if Num_UHE == 1:\n",
        "    plt.figure(iest)\n",
        "    plt.title(\"Funcao de custo futuro - Estagio\"+str(iest))\n",
        "    plt.xlabel(\"Volume inicial (hm^3)\")\n",
        "    plt.ylabel(\"Custo total ($)\")\n",
        "\n",
        "    eixox = []\n",
        "    eixoy = []\n",
        "\n",
        "  if Num_UHE == 2:\n",
        "    fig = plt.figure(iest)\n",
        "    ax = fig.gca(projection=\"3d\")\n",
        "    ax.set_title(\"Funcao de Custo Futuro - Estagio\"+str(iest))\n",
        "    ax.set_xlabel(\"Vol Ini UHE 1\")\n",
        "    ax.set_ylabel(\"Vol Ini UHE 2\")\n",
        "    ax.set_zlabel(\"Custo ($)\")\n",
        "    Usi_1 = np.arange(0,100+passo,passo)\n",
        "    Usi_2 = np.arange(0,100+passo,passo)\n",
        "    Usi_1, Usi_2 = np.meshgrid(Usi_1, Usi_2)\n",
        "    Custos = np.zeros((sistema[\"DGer\"][\"Nr_Disc\"],sistema[\"DGer\"][\"Nr_Disc\"]))\n",
        "\n",
        "\n",
        "  for discretizacao in discretizacoes:\n",
        "    VI = []\n",
        "\n",
        "    for i, iusi in enumerate(sistema[\"UHE\"]):\n",
        "      VI.append(iusi[\"Vmin\"] + (iusi[\"Vmax\"] - iusi[\"Vmin\"])*discretizacao[i]/100)\n",
        "    \n",
        "    if Num_UHE == 1:\n",
        "      eixox.append(VI[0])\n",
        "    \n",
        "    media = 0.\n",
        "\n",
        "    media_cma = []\n",
        "\n",
        "    for i,iusi in enumerate(sistema[\"UHE\"]):\n",
        "      media_cma.append(0)\n",
        "\n",
        "    for icen in np.arange(0,sistema[\"DGer\"][\"Nr_Cen\"]):\n",
        "      \n",
        "      AFL = []\n",
        "\n",
        "      for i, iusi in enumerate(sistema[\"UHE\"]):\n",
        "        AFL.append(iusi[\"Afl\"][iest-1][icen])\n",
        "      \n",
        "      #print(iest,discretizacao,VI,icen,AFL)\n",
        "      resultado = despacho(sistema,VI,AFL,pote_de_corte,iest+1,imprime = False)\n",
        "      media += resultado[\"DGer\"][\"Custo total\"]\n",
        "\n",
        "      for i,iusi in enumerate(resultado[\"UHE\"]):\n",
        "        media_cma[i] += iusi[\"cma\"]\n",
        "\n",
        "\n",
        "    media = media/sistema[\"DGer\"][\"Nr_Cen\"]\n",
        "\n",
        "    if Num_UHE == 2:\n",
        "      for linha in range(sistema[\"DGer\"][\"Nr_Disc\"]):\n",
        "        for coluna in range(sistema[\"DGer\"][\"Nr_Disc\"]):\n",
        "          if Usi_1[linha][coluna] == discretizacao[0] and Usi_2[linha][coluna] == discretizacao[1]:\n",
        "            Custos[linha][coluna] = media\n",
        "\n",
        "          \n",
        "\n",
        "    term_indep = media\n",
        "\n",
        "    for i,iusi in enumerate(resultado[\"UHE\"]):\n",
        "        media_cma[i] = -media_cma[i]/sistema[\"DGer\"][\"Nr_Cen\"]\n",
        "        term_indep -= VI[i]*media_cma[i]\n",
        "\n",
        "    corte = {\n",
        "        \"Estagio\": iest,\n",
        "        \"Termo_Indep\": term_indep,\n",
        "        \"Coefs\": media_cma\n",
        "    }\n",
        "\n",
        "    pote_de_corte.append(corte)\n",
        "\n",
        "    if Num_UHE == 1:\n",
        "      eixoy.append(media)\n",
        "\n",
        "  if Num_UHE == 1:\n",
        "    plt.plot(eixox,eixoy,marker='o',color=\"blue\")\n",
        "\n",
        "  if Num_UHE == 2:\n",
        "    surf = ax.plot_surface(Usi_1,Usi_2,Custos, cmap = cm.coolwarm)\n",
        "\n",
        "\n",
        "despacho(sistema, [62], [16], pote_de_corte, 2, imprime=True)\n",
        "#despacho(sistema, [62,150], [16,30], pote_de_corte, 2, imprime=True)  \n",
        "print(\"Tempo decorrido na PDE Adaptada\",time.time()-t)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Custo total: [ 9.97e+02]\n",
            "\n",
            "Volume final da usina 0 é [ 5.17e+01]\n",
            " hm3\n",
            "Volume turbinado da usina 0 é [ 2.63e+01]\n",
            " hm3\n",
            "Volume vertido da usina 0 é [ 0.00e+00]\n",
            " hm3\n",
            "Geracao na UTE 0 é [ 1.50e+01]\n",
            " MWmed\n",
            "Geracao na UTE 1 é [ 1.00e+01]\n",
            " MWmed\n",
            "Deficit de energia no sist é [ 3.55e-15]\n",
            " MWmed\n",
            "Custo Futuro é [ 5.97e+02]\n",
            " R$\n",
            "O valor da água na usina 0 é [ 2.38e+01]\n",
            "\n",
            "O Custo Marginal de Operacao é [-2.50e+01]\n",
            " R$/MWh\n",
            "\n",
            "###########################################################\n",
            "\n",
            "Tempo decorrido na PDE Adaptada 0.3746304512023926\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deZhcZZn38e8vBEICmgQIIQskKBl5AZGlZZFFBNJhE9Bh6UxG4ojmdcQZ3EZRVHbHFZVRURQUNLIKJjK8YNgERJYOOwRMwEASQhIIhCVsIff7x/MUqXS6urqTrj7dVb/PdZ2rTj3nVJ27qqvrrvM859xHEYGZmVlH+hUdgJmZ9X5OFmZmVpWThZmZVeVkYWZmVTlZmJlZVU4WZmZWlZOF9RmSxkoKSf2LjqU7SPqIpHmSXpa0c9Hx9FX5/XtX0XHUOyeLBiBprqRX8z9VaRpZdFx9XU5c26zDU3wf+GxEbBwR91bZ1n6S5q/DtrqdpJslvdbmc/WnTjxurqQDuyuO/P490YntfkjSg5JekPScpKskjequOOqdk0Xj+HD+pypNTxcdkDEGeLgnNlTDvbHPtvlcfbhG2+kOjwATImIIMBKYDZxbbEh9h5NFA2v7C0/SqZJ+l+dLXT6TJT0l6VlJJ5etu56kr0l6XNJLkmZK2jIv+3HuXnkxt+9T9rgBkn4k6ek8/UjSgArxrSfp+3nbTwCHtlk+WNL5khZKWiDpTEnrdfBca8TbXtdW/sX8yTy/jaS/SFqW47g0t9+SV78//6I+Nrd/StIcSUslTW9vDy6/By8D6+XHP57bV9tTkfSb/Jo2Av4fMLJ8z7C0vGz91fY+8t/3K5IeAF6R1F/S4ZIezr+ub5b0f9p7v9aVpM0kXZ23s1TSrZL6SfotsBXwp/w6vpzXv1zSM/l9vkXS9mXPtamkP+XP0935PbmtbPnb71v+TFwkaYmkJyV9XVI/gIhY1OZH0lvAuuwZNhQnC6tmb+A9wAHAN8u+XL4ATAQOAd4JfAJYnpfdDewEbAL8Hrhc0oZ52cnAHnn5+4DdgK9X2PangMOAnYEm4Kg2y38DrCD9w+8MNAOfrPBcHcXbkTOAPwNDgdHA/wBExL55+fvyL+pLJe0P/DdwDDACeBK4pO0TRsTrEbFx2ePf3VEAEfEKcDDw9FrsGU4kJdkhwLuAi4HPAcOAa0hf2ht08rm64ovA/Lyd4cDXgIiIjwFPsWpP97t5/f8HjAM2B+4BppY910+BV4AtgMl5quR/gMGk1/pB4Djg30oLJW0l6QXgVeBLwHfbexJrR0R4qvMJmAu8DLyQpz+WtR9Ytt6pwO/y/FgggNFly+8CWvL8Y8ARndz+86QvRYDHgUPKlk0A5lZ43I3Ap8vuN+eY+pO+gF4HBpYtnwjcVOG52o237HX2L2u7Gfhknr8IOK/8fShbL4Btyu6fD3y37P7GwJvA2AoxtX182/u/Ac7M8/sB89s8/u3l7a2T/76fKLv/DeCysvv9gAXAfmv5ubqZlHBfKJvOyMtOB6aVv542cR3YwfMOye/FYNLe15vAe8qWnwnc1vZ9y+u+AWxXtuz/Aje3s41NgK8Ae9T6/69eJu9ZNI4jI2JIno7swuOeKZtfTvoCBNiS9MW/BklfkjQrdym8QPqn3ywvHkn6xV3yZG5rz0hgXpt1S8YA6wMLc1fHC8AvSL9M21Mx3iq+DAi4K3fffKKDdVd7bRHxMvAcUOQgavn71za+lXn5GvHlLrtSl9fPO3j+/yz7XA2JiG/k9u8Bc4A/S3pC0kmVniB3EX47dxG+SEomkD4zw0g/DspfxzzatxnpM9H287XG64uIpcCFwDTVydF1teZk0dheAQaV3d+iC4+dB6zRfZLHJ75M6ooZGmkwcRnpCxfgadIXfclWua09C0lf8uXrlm//dWCzsi+qd0bE9rSv3XhJ7wFUeB8i4pmI+FREjCT9Sv2ZKh8Btdpry2MNm5J+vXfG8kpxkH49txd7tb9f+ePaxifS+7tGfBHxrVjV5fXpTsTe9vEvRcQXI+JdwOHAFyQd0E5MAP8CHAEcSPphMbYUIrCE1NU4umz98s9EuWdJeyFtP1+V3v/+pB8X76z2eszJotHdB7RIWl9Se2MCHfkVcIakcUp2lLQp8A7SP/cSoL+kb7L6P+PFwNclDZO0GfBN4HcVtnEZ8J+SRksaCrz96zQiFpLGEn4g6Z158PTdkj7YlXgjYgnpy+Rf8y/cT1CWVCQdLan0RfU86YtuZb6/iNQ3Xv7a/k3STkqD9t8C7oyIuZXexDbuA/4lx3EQqc+9ZBGwqaTBbdY/RNImkrYgjUV05DLgUEkHSFqfNK7wOnB7J+PrNEmHKR0cINKPhbeo/L69I8fxHCn5fau0ICLeAq4ETpU0SNK2pHGINeR1LwPOkvQOSWNIY1WlgzY+Kuk9+bMyDDgbuDfvZVgVThaN7RukL8bngdNIg9GddTbpH/PPwIuk/vqBwHXAtcDfSV0Ar7F6t8GZQCvwAPAgaTDzTNr3y/x89+f1rmyz/DhgA9Ihkc8DV5AGlrsSL6SB9P8ifVltz+pfnu8H7lQ6emk6cGKsOqb/VODC3A12TERcT3pP/0DaK3o30FIhnvacCHyY1Pc/CfhjaUFEPEpKRk/k7Y0Efkt6b+bm13VpR08eEY8B/0oaBH42b+vDEfFGF2Js6yda/TyLmbl9HHA9aazsb8DPIuKmvOy/ST8YXpD0JdK40JOkpP0IcEebbXyWtMfxDOk1X0xKLu35D9Ie1xPAbaTP9AV52SjSZ/Ml0mdvJfCRtX3hjUZ5sMfMrE+Q9B1gi4jo6Kgo62beszCzXk3StrnbUJJ2A44Hrio6rkbjowDMrLd7B6nraSRpvOMHpMNyrQe5G8rMzKqqWTdUPurgvrLpRUmfy0duzJA0O98OzetL0jlKpRIekLRL2XNNzuvPluR+SjOzHtYjexZK9XoWALsDJwBLI+Lb+USdoRHxFUmHkI5kOCSv9+OI2F3SJqSjZ5pIhy3OBHaNiOcrbW+zzTaLsWPH1vQ1mZnVm5kzZz4bEcPaW9ZTYxYHAI9HxJOSjiCVJYB0BuXNpNPujwAuipS97pA0RNKIvO6M0rHQkmYAB5H6MNs1duxYWltba/RSzMzqk6QnKy3rqaOhWlj15T48n1AF6bjp4Xl+FKsfjz8/t1VqX42kKZJaJbUuWbKkO2M3M2t4NU8WShUtDwcub7ss70V0Sz9YRJwXEU0R0TRsWLt7UWZmtpZ6Ys/iYOCeiFiU7y/K3Uvk28W5fQGr13wZndsqtZuZWQ/piWQxkdXHF6azqh79ZFYdLz0dOC4fFbUHsCx3V10HNEsamo+cas5tZmbWQ2o6wJ2rbo4nVess+TZwmaTjSfVgjsnt15COhJpDqr75b5BKCUs6g3RBHYDTXfjLzKxn1XTPIiJeiYhNI2JZWdtzEXFARIyLiANLX/yRnBAR746I90ZEa9ljLoiIbfL061rFO3UqjB0L/fql26lTqz3CzKwxuNxHNnUqTJkCy/OFNp98Mt0HmDSpuLjMzHoDFxLMTj55VaIoWb48tZuZNToni+ypp7rWbmbWSJwssq226lq7mVkjcbLIzjoLBg1avW3QoNRuZtbonCyySZPgvPNgs83S/S22SPc9uG1m5mSxmkmT4OGH0/xnP+tEYWZW4mTRxuabw847w4wZRUdiZtZ7OFm0Y/x4uP12eOmloiMxM+sdnCza0dwMb74Jf/lL0ZGYmfUOThbt2GsvGDgQ/vznoiMxM+sdnCzaseGG8MEPetzCzKzEyaKC5mZ49FGfwW1mBk4WFY0fn269d2Fm5mRR0fbbw4gRHrcwMwMni4qk1BV1/fXw1ltFR2NmViwniw40N8PSpXDvvUVHYmZWLCeLDhx4YLp1V5SZNToniw5svjnstJOThZmZk0UVzc2p9MfLLxcdiZlZcWqaLCQNkXSFpEclzZK0p6RNJM2QNDvfDs3rStI5kuZIekDSLmXPMzmvP1vS5FrG3JZLf5iZ1X7P4sfAtRGxLfA+YBZwEnBDRIwDbsj3AQ4GxuVpCnAugKRNgFOA3YHdgFNKCaYn7LVXOqPbXVFm1shqliwkDQb2Bc4HiIg3IuIF4AjgwrzahcCRef4I4KJI7gCGSBoBTABmRMTSiHgemAEcVKu42yqV/nCyMLNGVss9i62BJcCvJd0r6VeSNgKGR8TCvM4zwPA8PwqYV/b4+bmtUnuPKZX+mDev+rpmZvWolsmiP7ALcG5E7Ay8wqouJwAiIoDojo1JmiKpVVLrkiVLuuMp39bcnG5d+sPMGlUtk8V8YH5E3JnvX0FKHoty9xL5dnFevgDYsuzxo3NbpfbVRMR5EdEUEU3Dhg3r1hfi0h9m1uhqliwi4hlgnqT35KYDgEeA6UDpiKbJwLQ8Px04Lh8VtQewLHdXXQc0SxqaB7abc1uPkVJhwRkzXPrDzBpT/xo//38AUyVtADwB/BspQV0m6XjgSeCYvO41wCHAHGB5XpeIWCrpDODuvN7pEbG0xnGvobkZLroolf5oaurprZuZFaumySIi7gPa+2o9oJ11AzihwvNcAFzQvdF1Tan0x4wZThZm1nh8BncnDR8O73ufxy3MrDE5WXRBczP89a8u/WFmjcfJogtc+sPMGpWTRRfsvXc6o9vnW5hZo3Gy6AKX/jCzRuVk0UXjx8OsWS79YWaNxcmii1z6w8wakZNFF+2wA2yxhZOFmTUWJ4suktLexYwZsHJl0dGYmfUMJ4u1MH48PPdcKv1hZtYInCzWQqn0h4+KMrNG4WSxFrbYIpX+8LiFmTUKJ4u11NwMt90Gr7xSdCRmZrXnZLGWxo936Q8zaxxOFmupVPrD4xZm1gicLNbSwIGw775OFmbWGJws1kFzcyr9MX9+0ZGYmdWWk8U6GD8+3fqoKDOrd04W6+C9701X0HNXlJnVOyeLdVAq/XH99S79YWb1zcliHTU3w7PPwn33FR2JmVnt1DRZSJor6UFJ90lqzW2bSJohaXa+HZrbJekcSXMkPSBpl7LnmZzXny1pci1j7iqX/jCzRtATexYfioidIqIp3z8JuCEixgE35PsABwPj8jQFOBdScgFOAXYHdgNOKSWY3mCLLWDHHZ0szKy+FdENdQRwYZ6/EDiyrP2iSO4AhkgaAUwAZkTE0oh4HpgBHNTTQXfEpT/MrN7VOlkE8GdJMyVNyW3DI2Jhnn8GGJ7nRwHlFyudn9sqta9G0hRJrZJalyxZ0p2voarm5lT645ZbenSzZmY9ptbJYu+I2IXUxXSCpH3LF0ZEkBLKOouI8yKiKSKahg0b1h1P2Wku/WFm9a6mySIiFuTbxcBVpDGHRbl7iXy7OK++ANiy7OGjc1ul9l5j4EDYZx8nCzOrXzVLFpI2kvSO0jzQDDwETAdKRzRNBqbl+enAcfmoqD2AZbm76jqgWdLQPLDdnNt6leZmeOQRl/4ws/pUyz2L4cBtku4H7gL+NyKuBb4NjJc0Gzgw3we4BngCmAP8EvgMQEQsBc4A7s7T6bmtV2luTrfXX19sHGZmtaA0bFBfmpqaorW1tUe3GQEjRsD++8Pvf9+jmzYz6xaSZpad5rAan8HdTaRUWHDGDJf+MLP642TRjVz6w8zqlZNFNyqV/nDJcjOrN04W3WjECJf+MLP65GTRzcaPd+kPM6s/ThbdrLkZ3njDpT/MrL44WXSzffaBAQM8bmFm9cXJopsNHAj77utxCzOrL04WNTB+PDz8MCzoVRWszMzWXtVkIWm0pC9Jmibpbkm3SPqZpEMlOdm0o1T6w11RZlYvOvyyl/Rr4ALgDeA7wERSzabrSRcguq1t2XGD974Xhg93sjCz+tG/yvIfRMRD7bQ/BFwpaQNgq+4Pq2/r1y91RV13XSr90c/7X2bWx3X4NVYhUZQvfyMi5nRvSPWhuRmWLIH77y86EjOzddeZMYsPSRqd58dIul7SHe5+6lip9IePijKzetCZDpJvA8vy/LeAK4ATgR/VKqh6MGJEGrtwsjCzelBtgPsU0iVNP5/nJwCjSIPbm0n6pvcwKmtuTqU/li8vOhIzs3VTbcziNOAp4CbSoPbtEfGN3D4vIk6PCBe2qMClP8ysXnSmG+qLwPdJXU9fBpC0PeCrNlRRKv3hrigz6+uqHTpLRPwV2L1N28PACbUKql4MHJgShpOFmfV11cYs9q6y/J2SdujekOpLc3Mq/fH000VHYma29qp1Q/2zpNvzQPahknaTtK+kT0j6LXA1MLCjJ5C0nqR7JV2d728t6U5JcyRdmk/sQ9KAfH9OXj627Dm+mtsfkzRhnV5xD3PpDzOrB9UGuD8PHAYsBI4GzgC+AIwDfhER+0bE3VW2cSIwq+z+d4AfRsQ2wPPA8bn9eOD53P7DvB6StgNagO1JR2H9TNJ6nX6FBXvve2Hzzd0VZWZ9W9UB7ohYGhG/jIiPR8SEiDgyIr4aEbdVe2w+me9Q4Ff5voD9SedqAFwIHJnnj8j3ycsPyOsfAVwSEa9HxD+AOcBunX+JxSqV/pgxI5X+MDPri2pdtehHpCOoSl+TmwIvRMSKfH8+6bwN8u08gLx8WV7/7fZ2HvM2SVMktUpqXbJkSXe/jnVSKv3xwANFR2JmtnZqliwkHQYsjoiZtdpGuYg4LyKaIqJp2LBhPbHJThs/Pt26K8rM+qpa7lnsBRwuaS5wCan76cfAEEmlQ3ZHA6VLBC0gnS1OXj4YeK68vZ3H9AkjRsAOOzhZmFnf1eF5FpI+2tHyiLiyg2VfBb6an2c/4EsRMUnS5cBRpAQyGZiWHzI93/9bXn5jRISk6cDvJZ0NjCQNrt9V/aX1Ls3N8JOfpNIfgwYVHY2ZWddUOynvwx0sC6BisujAV4BLJJ0J3Aucn9vPB34raQ6wlHQEFBHxsKTLgEeAFcAJEfHWWmy3UM3NcPbZcOutMKFPHfxrZgaKiKJj6HZNTU3R2tpadBirWb4cNtkETjgBfvCDoqMxM1uTpJkR0dTesqrlPsqe5FDSuQ4bltoi4vR1D68xDBrk0h9m1nd1aoBb0s+BY4H/AEQ6QW9MDeOqS+PHw0MPufSHmfU9nT0a6gMRcRzpDOvTgD2Bf6pdWPWpVPrj+uuLjcPMrKs6myxezbfLJY0E3gRG1Cak+rXjji79YWZ9U2fHLK6WNAT4HnAP6UioX9UsqjrVtvRHv1qfP29m1k06+3X13Yh4ISL+QBqr2BY4s3Zh1a/x42HxYpf+MLO+pbPJ4m+lmVzQb1l5m3VeqfSHS5abWV9S7eJHW0jaFRgoaWdJu+RpP8DnIa+FkSNd+sPM+p5qYxYTgI+T6jGdXdb+EvC1GsVU95qb4ac/dekPM+s7ql386MKI+BDw8Yj4UNl0eEd1oaxj48fD66+n0h9mZn1BZ8csbpB0dul6EZJ+IGlwTSOrY/vuCxts4HELM+s7Opsszid1PR2TpxeBX9cqqHrn0h9m1td0Nlm8OyJOiYgn8nQa8K5aBlbvmpvhwQdh4cKiIzEzq67TZ3BL2rt0R9JerDqr29bCm2+m25EjYexYmDq10HDMzDrU2TO4Pw1cVDZO8TzpQkW2FqZOhW99a9X9J5+EKVPS/KRJxcRkZtaRzu5ZvBgR7wN2BHaMiJ1JYxi2Fk4+OR02W2758tRuZtYbdTZZ/AEgIl6MiBdz2xW1Can+PfVU19rNzIpW7Rrc25IueDS4zfW430nZRZCsa7baKnU9tdduZtYbVduzeA9wGDCEdD3u0rQL8Knahla/zjprzTO3Bw1K7WZmvVGHexYRMQ2YJmnPiHDhwG5SGsQ++eTU9RQBn/mMB7fNrPfq1JiFE0X3mzQJ5s6FV1+FIUNg0aKiIzIzq6xml9+RtKGkuyTdL+lhSafl9q0l3SlpjqRLJW2Q2wfk+3Py8rFlz/XV3P6YpAm1irkIAwbAP/8zXHVVShxmZr1RLa/V9jqwfz7kdifgIEl7AN8BfhgR25DO1zg+r3886Rrf2wA/zOshaTughTTQfhDwM0nr1TDuHtfSAi+/DNdcU3QkZmbt61SykDRY0g+7Ukgwkpfz3fXzFMD+rDrs9kLgyDx/RL5PXn6AJOX2S/JFl/4BzAF26+Tr6xM+9CEYPhwuuaToSMzM2tfZPYsLSMUDu1RIUNJ6ku4DFgMzgMeBFyJiRV5lPjAqz48C5gHk5cuATcvb23lM+bamlJLZkiVLOvmyeof11oOjj4arr4YXX6y+vplZT6tpIcGIeCsidiJdPGk30rW7ayIizouIpohoGjZsWK02UzMtLfDaazB9etGRmJmtqUcKCUbEC8BNwJ7AEEmlQ3ZHAwvy/AJgy/z8/YHBwHPl7e08pm7suWc6Ke/ii4uOxMxsTZ1NFp8GfipprqS5wE+A/9vRAyQNkzQkzw8ExgOzSEnjqLzaZGBanp/OquKERwE3RkTk9pZ8tNTWwDjgrk7G3Wf06wfHHpuucfHcc0VHY2a2uloWEhwB3CTpAeBuYEZEXA18BfiCpDmkMYnz8/rnA5vm9i8AJwFExMPAZcAjwLXACRHxVmdfYF8ycSKsWAFX+oK1ZtbLKP14r7KSdE9E7NKmbWZE7FqzyNZBU1NTtLa2Fh1Gl0XAttvCqFFw441FR2NmjSZ/rze1t8yFBHsRKQ10n3FGuoLeiBFFR2RmlriQYC/T0pL2MC67rOhIzMxW6Ww3VJ8qJNhXu6FKdtoJBg6Ev/WZd9zM6kFH3VCdHeD+iKR3Slpf0g2Slkj6126M0cpMnAh33AH/+EfRkZiZJZ1NFs35CnmHAXOBbYD/qlVQje7YY9PtpZcWG4eZWUlnk8X6+fZQ4PKIWFajeAwYOxb22MO1osys9+hssviTpEeBXYEbJA0DXqtdWDZxItx/P8yaVXQkZmadv/jRScAHgKaIeBN4hVQN1mrk6KPTWd3euzCz3qDD8yxKJB1XNl++6KLuDsiSESNgv/1SrahTT03nYJiZFaWz3VDvL5v2AU4FDq9RTJa1tMDs2XDvvUVHYmaNrrPdUP9RNn2KdFLexrUNzT76Uejf311RZla8tb2s6ivA1t0ZiK1p001hwoSULFauLDoaM2tknR2z+BPpkqiQEsx2pEqwVmMtLfC//5vO5t5rr6KjMbNG1alkAXy/bH4F8GREzK9BPNbGEUfAhhumgW4nCzMrSofdUJK2kbRXRPylbPorMEbSu3soxob2jnfAYYfB5Zena12YmRWh2pjFj4AX22l/MS+zHtDSAosXw803Fx2JmTWqaslieEQ82LYxt42tSUS2hkMOSXsYvj63mRWlWrIY0sGygd0ZiFU2cCAceWS63OrrrxcdjZk1omrJolXSGhc5kvRJYGZtQrL2TJwIL7wA111XdCRm1oiqHQ31OeAqSZNYlRyagA2Aj9QyMFvdgQem8y4uuQQO97nzZtbDOtyziIhFEfEB4DTSdSzmAqdFxJ4R8UxHj5W0paSbJD0i6WFJJ+b2TSTNkDQ73w7N7ZJ0jqQ5kh6QtEvZc03O68+WNHndXnLftP76cNRRMG0avPJK0dGYWaPpbLmPmyLif/J0YyefewXwxYjYDtgDOEHSdsBJwA0RMQ64Id8HOBgYl6cpwLmQkgtwCrA7sBtwSinBNJqWFli+HK6+uuhIzKzRrG25j6oiYmFE3JPnXwJmAaNIpc0vzKtdCByZ548ALorkDmCIpBHABGBGRCyNiOeBGcBBtYq7N9tnn1SN1rWizKyn1SxZlJM0FtgZuJN0OO7CvOgZYHieHwXMK3vY/NxWqb3tNqZIapXUumTJkm6Nv7dYb710ydVrrkmD3WZmPaXmyULSxsAfgM/l63i/LSKCVTWn1klEnBcRTRHRNGzYsO54yl6ppQXeeAP++MeiIzGzRlLTZCFpfVKimBoRV+bmRbl7iXy7OLcvALYse/jo3FapvSHtthtsvbVP0DOznlWzZKF0Sb3zgVkRcXbZoulA6YimycC0svbj8lFRewDLcnfVdUCzpKF5YLs5tzUkKe1d3HBDKgFiZtYTarlnsRfwMWB/Sffl6RDg28B4SbOBA/N9gGuAJ4A5wC+BzwBExFLgDODuPJ2e2xpWSwu89Rb84Q9FR2JmjUJp2KC+NDU1RWtra9Fh1EwE7LBDOknvlluKjsbM6oWkmRHR1N6yHjkayrpXqSvq1lthvq8qYmY9wMmij2ppSbeXXlpsHGbWGJws+qhx42DXXX2Cnpn1DCeLPqylBVpbYc6coiMxs3rnZNGHHXtsuvXehZnVmpNFH7bllrD33k4WZlZ7ThZ93MSJ8PDD8OAaF781M+s+ThZ93FFHQb9+3rsws9pysujjNt8cDjggJYs6PL/SzHoJJ4s6MHEiPPEE3H130ZGYWb1ysqgDH/kIbLCBu6LMrHacLOrAkCFw8MHpbO633io6GjOrR04WdaKlBZ5+Gm67rehIzKweOVnUiQ9/GAYNcleUmdWGk0Wd2GgjOPxwuPxyePPNoqMxs3rjZFFHWlrguefSVfTMzLqTk0UdOeggGDzY1+c2s+7nZFFHBgyAj34UrroKXnut6GjMrJ44WdSZlhZ46SW45pqiIzGzeuJkUWf23x+GDfNRUWbWvWqWLCRdIGmxpIfK2jaRNEPS7Hw7NLdL0jmS5kh6QNIuZY+ZnNefLWlyreKtF/37w9FHw9VXpz0MM7PuUMs9i98AB7VpOwm4ISLGATfk+wAHA+PyNAU4F1JyAU4Bdgd2A04pJRirbOJEePVVmD696EjMrF7ULFlExC3A0jbNRwAX5vkLgSPL2i+K5A5giKQRwARgRkQsjYjngRmsmYCsjQ98AEaPdleUmXWfnh6zGB4RC/P8M8DwPD8KmFe23vzcVqndOtCvX7rk6nXXwdK26drMbC0UNsAdEQF02xUYJE2R1CqpdcmSJd31tH3WxInpTO4rryw6EjOrBz2dLBbl7iXy7eLcvgDYsmy90bmtUvsaIuK8iGiKiKZhw4Z1e+B9zS67wDbbuCvKzLpHTyeL6UDpiKbJwLSy9uPyUVF7AMtyd9V1QLOkoXlguzm3WRVS2ru46SZ45pmiozGzvq6Wh85eDPwNeI+k+ZKOB74NjJc0Gzgw3we4BngCmAP8EvgMQEQsBc4A7s7T6fVV7a8AAA58SURBVLnNOqGlBVauTMUFzczWhaIOL9zc1NQUra2tRYfRK+y4I2y8Mdx+e9GRmFlvJ2lmRDS1t8xncNe5iRPhb3+DuXOLjsTM+jInizp37LHp9rLLio3DzPo2J4s69653we67u2y5ma0bJ4sG0NIC990Hjz5adCRm1lc5WTSAY45Jh9L6nAszW1tOFg1g5EjYdls466xUCmTsWJg6teiozKwv6V90AFZ7U6fC44/DihXp/pNPwpQpaX7SpOLiMrO+w3sWDeDkk+GNN1ZvW748tZuZdYaTRQN46qn22598Eu69F+rwvEwz62ZOFg1gq60qL9tll3SW9/e+B08/3XMxmVnf4mTRAM46CwYNWr1t0CD4xS/g3HNTOZAvfxm23BIOOgh+//vUTWVmVuJk0QAmTYLzzoMxY9IhtGPGpPtTpsCnP53KgTz2GHztazBrVlp/iy3gE5+Av/wlFSM0s8bmQoK2mpUr4ZZb4KKLUrXal19OyeVjH4PjjoNx44qO0MxqxYUErdP69YP99oMLLoBFi9Jht9tuC9/6FvzTP6Xre//8575cq1mjcbKwigYNgn/5F7j2Wpg3D777XXjxRfj3f4cRI+Coo+BPf0qXbzWz+uZkYZ0yciT813/Bgw/CPffAZz6TuqsOPxxGjYITT0ztpV7NqVPTmeI+Y9ysPnjMwtbam2/Cddel8Y1p09KJf9tvDzvsANOnw6uvrlp30KA0qO4zxs16L49ZWE2svz4cdli6VsYzz6RDcQcPhksvXT1RQDoU94tfhL//HV55pZh4zWztec/Cul2/ftXPCh88OHVtjRqVptJ8edvw4dC/E9XLpk5NpUueeiqdgHjWWd6DMVsbHe1ZuJCgdbuttkqlRNrafHP4/vfTmeILFqTp6afhxhth4cJVhQ5L+vVLCaNSMhk5Ml1b/POfX3USoYskmtWGk4V1u7POSl/Y5WeBDxoEZ59d+Qt85UpYsmRVAilPJgsWpGuI//Wv8Nxz1be/fHkagJ83DzbddNW0ySar5gcMWPvX5z0Za0gR0Scm4CDgMWAOcFJH6+66665hxfrd7yLGjImQ0u3vftc9z/vqqxFPPBFx660Rl1wSkTq8uj5ttFHElltG7LRTxAEHRBxzTMS//3vEySdHnH12xIUXRlx9dcTtt0c89ljEs89GrFiRXsegQas/16BB3ff62lOr97LobdX79vriawNao8L3ap8Ys5C0HvB3YDwwH7gbmBgRj7S3vscsGsfYse13eY0ZA488kk4efO65Naf22pcuTVOl8iZSmtpbvtFGae9iww1h4MCu3bZtGzAgdcFB2otpby+tFkeW9eS26n17ffW1dTRm0VeSxZ7AqRExId//KkBE/Hd76ztZNI7u/qdcuRKWLaucVM44o/Jjhw+H115LR4K1vX5IVw0YkJLHSy+1n5z6909n1Henv/99zXGjWm2r3rfXW17bmDGpC7ez6mGAexQwr+z+fGD38hUkTQGmAGzVUU1uqyulhNBdYwj9+sHQoWnaZps1l190UeU9mfJ/ypUrU+IoJY/2bjtaVro955z241yxArbbbu1eYyWPtLufXptt1fv2estrq3Qtm7VSqX+qN03AUcCvyu5/DPhJpfU9ZmG10tNjFmPGtD/mMmZM395WvW+vr742Ohiz6Csn5S0Atiy7Pzq3mfWoSuXea3U0VKVrkZx1Vt/eVr1vry5fW6Us0psmUnfZE8DWwAbA/cD2ldb3noXVEx/B0ze31xdfG339aCgASYcAPwLWAy6IiIo50wPcZmZdVw8D3ETENcA1RcdhZtaI+sqYhZmZFcjJwszMqnKyMDOzqpwszMysqj5zNFRXSFoCtHOebadtBjzbTeF0J8fVNY6raxxX19RjXGMiYlh7C+oyWawrSa2VDh8rkuPqGsfVNY6raxotLndDmZlZVU4WZmZWlZNF+84rOoAKHFfXOK6ucVxd01BxeczCzMyq8p6FmZlV5WRhZmZVNXSykLSlpJskPSLpYUkn5vZNJM2QNDvfDu3huDaUdJek+3Ncp+X2rSXdKWmOpEslbdCTcZXFt56keyVd3VvikjRX0oOS7pPUmtsK/TvmGIZIukLSo5JmSdqz6LgkvSe/T6XpRUmfKzquHNvn82f+IUkX5/+F3vD5OjHH9LCkz+W2Qt4vSRdIWizpobK2dmNRck5+7x6QtMvabrehkwWwAvhiRGwH7AGcIGk74CTghogYB9yQ7/ek14H9I+J9wE7AQZL2AL4D/DAitgGeB47v4bhKTgRmld3vLXF9KCJ2KjvGvOi/I8CPgWsjYlvgfaT3rdC4IuKx/D7tBOwKLAeuKjouSaOA/wSaImIH0uUIWij48yVpB+BTwG6kv+FhkrahuPfrN8BBbdoqxXIwMC5PU4Bz13qrlS500YgTMA0YDzwGjMhtI4DHCoxpEHAP6ZrjzwL9c/uewHUFxDM6fxj3B64G1Evimgts1qat0L8jMBj4B/lAkt4SV5tYmoG/9oa4gFHAPGAT0uUTrgYmFP35Ao4Gzi+7/w3gy0W+X8BY4KFqnyngF8DE9tbr6tToexZvkzQW2Bm4ExgeEQvzomeA4QXEs56k+4DFwAzgceCFiFiRV5lP+ufqaT8i/aOszPc37SVxBfBnSTMlTcltRf8dtwaWAL/O3Xa/krRRL4irXAtwcZ4vNK6IWAB8H3gKWAgsA2ZS/OfrIWAfSZtKGgQcQrrMc2/6O1aKpZSAS9b6/XOyACRtDPwB+FxEvFi+LFI67vHjiyPirUjdBKNJu7/b9nQMbUk6DFgcETOLjqUde0fELqTd7hMk7Vu+sKC/Y39gF+DciNgZeIU2XRVFfb4Act//4cDlbZcVEVfuZz+ClGRHAhuxZndLj4uIWaSusD8D1wL3AW+1Waewv2NbtYql4ZOFpPVJiWJqRFyZmxdJGpGXjyD9ui9ERLwA3ETa/R4iqXR1w9HAgh4OZy/gcElzgUtIXVE/7gVxlX6VEhGLSf3vu1H833E+MD8i7sz3ryAlj6LjKjkYuCciFuX7Rcd1IPCPiFgSEW8CV5I+c73h83V+ROwaEfuSxk3+TvHvV7lKsSwg7QWVrPX719DJQpKA84FZEXF22aLpwOQ8P5k0ltGTcQ2TNCTPDySNo8wiJY2jioorIr4aEaMjYiyp++LGiJhUdFySNpL0jtI8qR/+IQr+O0bEM8A8Se/JTQcAjxQdV5mJrOqCguLjegrYQ9Kg/L9Zer8K/XwBSNo8324FfBT4PcW/X+UqxTIdOC4fFbUHsKysu6prenKgqLdNwN6k3bUHSLuW95H6IzclDeLOBq4HNunhuHYE7s1xPQR8M7e/C7gLmEPqOhhQ4Hu3H3B1b4grb//+PD0MnJzbC/075hh2Alrz3/KPwNBeEtdGwHPA4LK23hDXacCj+XP/W2BA0Z+vHNetpMR1P3BAke8XKcEvBN4k7b0eXykW0gEoPyWNeT5IOtJsrbbrch9mZlZVQ3dDmZlZ5zhZmJlZVU4WZmZWlZOFmZlV5WRhZmZVOVlYr6NUCXhCm7bPSapYBC1Xnd2s9tG9vb1f5aKTHa3zaUnHdbB8pKQrqjzH2PLqom2WjdCqyr8fl/STzsReZXsjcoXSe0rnrpQtu1arKiH/XNJ6uf37kvZf121b7+ZkYb3RxaST/sqV1zAqXER8MiIeqbLOzyPiog6WPx0RR1Va3glfAH65Do9fTU4OfwS+AlwIXJErHJQcE6kS8g7AMFKBPYD/oZiKvtaDnCysN7oCOLR03YJc5HEkcKukiUrXrXhI0nfaPrDtL3FJX5J0ap6/WdIPJbUqXVvi/ZKuzNcAOLPsMf+qdD2R+yT9ovQLus12bpbUlOdflnRW/tV9h6Thuf1USV/K89tIuj6vc4+kd5fHmudvzcvukfSBTrxP/0yqVVQyMv/6ny3pu2Wxvizpe3mP4HpJu+X4n5B0eF5nfVIy/k5E/CEifkw6+/ftZBSr6qb1BzYg1x+KiCeBTSVt0YmYrY9ysrBeJyKWks7YPTg3tQCXkUovf4dUk2on4P2Sjuzi078R6XoXPyeVRDiB9Ev540pVRf8PcCywV6RCjm8Bk6o850bAHflX9y2kax+0NRX4aV7nA6QzcMstBsZHKoZ4LHBORxuUtDXwfES8Xta8U37se4FjJZVqAm1EKs2yPfAScCaphMxHgNMBIuLNiDgsVtVHIyJ+GhEfb7Pd63KsL5GSesk9pDpOVqecLKy3Ku+KKnVBvR+4OVKhuRWkL+B9Kzy+kun59kHg4YhYmL9wnyAVXDuAdEGgu5VKxB9AKjfRkTdI116AVFJ7bPnC3L0zKiKuAoiI1yJieZvnWB/4paQHSSUtOhwPISXOJW3aboiIZRHxGqk0xZiy+Ep7IA8Cf4lUqO/BtrFWExET8rYHkJJ2yWLS3p/Vqf7VVzErxDTgh0qXgRwUETMlje7E41aw+o+gDdssL/0SX1k2X7rfn1RL58KI+GoXYn0zVtXNeYu1+7/6PLCIdCW2fsBrVdZ/lcqvrW0c5fG9/bojYmVZNddOi4jXJE0jlROfkZs3zDFZnfKehfVKEfEyqdroBawa2L4L+KCkzfI4wkTgL20eugjYPHcpDQAO6+KmbwCOKqsyuomkMVUe06GIeAmYX+oykzRA6SI65QYDCyNiJfAx0iVFO/J3urhXsC4kbaxVJbD7A4eSCv6V/BOp+J/VKScL680uJv3SvhggUmnlk0hJ5H5gZkSsVhY6d6+cTkosM1j9C62qfITT10lX3XsgP8eIdXsZQEoA/5mf83ag7WDwz4DJku4nXejqlSpxvgI8rnQt6J6wETA9x1+6guPP4e3B8W1I1XWtTrnqrFkfJekjwK4R8fVeEMcuEfGNIuOw2vKYhVkfFRFXSdq06DhI3yM/KDoIqy3vWZiZWVUeszAzs6qcLMzMrConCzMzq8rJwszMqnKyMDOzqv4/rEJyjarpKpwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}