update LSTM 2 rtu model

lstm.ipynb

@@ -1135,7 +1135,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 110,
    "metadata": {},
    "outputs": [
     {
@@ -1272,7 +1272,7 @@
        "[2 rows x 65 columns]"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 110,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1294,7 +1294,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 111,
    "metadata": {},
    "outputs": [
     {
@@ -1315,7 +1315,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 179,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[<matplotlib.lines.Line2D at 0x207d4f52890>]"
+      ]
+     },
+     "execution_count": 179,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "plt.plot(df_filtered['rtu_002_rf_vfd_spd_fbk_tn'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 112,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1334,36 +1354,38 @@
     "#  'solar_radiation_set_1']]\n",
     "\n",
     "df_filtered = df_filtered.loc[:,['date','hp_hws_temp',\n",
-    " 'rtu_003_sa_temp',\n",
-    " 'rtu_003_oadmpr_pct',\n",
-    " 'rtu_003_ra_temp',\n",
-    " 'rtu_003_oa_temp',\n",
-    " 'rtu_003_ma_temp',\n",
-    " 'rtu_003_sf_vfd_spd_fbk_tn',\n",
-    " 'rtu_003_rf_vfd_spd_fbk_tn',\n",
-    " 'rtu_004_sa_temp',\n",
-    " 'rtu_004_oadmpr_pct',\n",
-    " 'rtu_004_ra_temp',\n",
-    " 'rtu_004_oa_temp',\n",
-    " 'rtu_004_ma_temp',\n",
-    " 'rtu_004_sf_vfd_spd_fbk_tn',\n",
-    " 'rtu_004_rf_vfd_spd_fbk_tn',\n",
-    "#  'rtu_001_sa_temp',\n",
-    "#  'rtu_001_oadmpr_pct',\n",
-    "#  'rtu_001_ra_temp',\n",
-    "#  'rtu_001_oa_temp',\n",
-    "#  'rtu_001_ma_temp',\n",
-    "#  'rtu_001_sf_vfd_spd_fbk_tn',\n",
-    "#  'rtu_001_rf_vfd_spd_fbk_tn',\n",
-    "#  'rtu_002_sa_temp',\n",
-    "#  'rtu_002_oadmpr_pct',\n",
-    "#  'rtu_002_ra_temp',\n",
-    "#  'rtu_002_oa_temp',\n",
-    "#  'rtu_002_ma_temp',\n",
-    "#  'rtu_002_sf_vfd_spd_fbk_tn',\n",
-    "#  'rtu_002_rf_vfd_spd_fbk_tn',\n",
-    " 'rtu_003_sat_sp_tn',\n",
-    " 'rtu_004_sat_sp_tn',\n",
+    "#  'rtu_003_sa_temp',\n",
+    "#  'rtu_003_oadmpr_pct',\n",
+    "#  'rtu_003_ra_temp',\n",
+    "#  'rtu_003_oa_temp',\n",
+    "#  'rtu_003_ma_temp',\n",
+    "#  'rtu_003_sf_vfd_spd_fbk_tn',\n",
+    "#  'rtu_003_rf_vfd_spd_fbk_tn',\n",
+    "#  'rtu_004_sa_temp',\n",
+    "#  'rtu_004_oadmpr_pct',\n",
+    "#  'rtu_004_ra_temp',\n",
+    "#  'rtu_004_oa_temp',\n",
+    "#  'rtu_004_ma_temp',\n",
+    "#  'rtu_004_sf_vfd_spd_fbk_tn',\n",
+    "#  'rtu_004_rf_vfd_spd_fbk_tn',\n",
+    " 'rtu_001_sa_temp',\n",
+    " 'rtu_001_oadmpr_pct',\n",
+    " 'rtu_001_ra_temp',\n",
+    " 'rtu_001_oa_temp',\n",
+    " 'rtu_001_ma_temp',\n",
+    " 'rtu_001_sf_vfd_spd_fbk_tn',\n",
+    " 'rtu_001_rf_vfd_spd_fbk_tn',\n",
+    " 'rtu_002_sa_temp',\n",
+    " 'rtu_002_oadmpr_pct',\n",
+    " 'rtu_002_ra_temp',\n",
+    " 'rtu_002_oa_temp',\n",
+    " 'rtu_002_ma_temp',\n",
+    " 'rtu_002_sf_vfd_spd_fbk_tn',\n",
+    " 'rtu_002_rf_vfd_spd_fbk_tn',\n",
+    "#  'rtu_003_sat_sp_tn',\n",
+    "#  'rtu_004_sat_sp_tn',\n",
+    " 'rtu_001_sat_sp_tn',\n",
+    " 'rtu_002_sat_sp_tn',\n",
     " 'air_temp_set_1',\n",
     " 'air_temp_set_2',\n",
     " 'dew_point_temperature_set_1d',\n",
@@ -1373,7 +1395,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": 113,
    "metadata": {},
    "outputs": [
     {
@@ -1413,7 +1435,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 24,
+   "execution_count": 114,
    "metadata": {},
    "outputs": [
     {
@@ -1459,7 +1481,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 25,
+   "execution_count": 115,
    "metadata": {},
    "outputs": [
     {
@@ -1475,22 +1497,22 @@
      "output_type": "stream",
      "text": [
       "Epoch 1/2\n",
-      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step - loss: 0.0577\n",
-      "Epoch 1: val_loss improved from inf to 0.08234, saving model to lstm_2rtu_smooth_03.keras\n",
-      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m234s\u001b[0m 29ms/step - loss: 0.0577 - val_loss: 0.0823\n",
+      "\u001b[1m8066/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 0.0548\n",
+      "Epoch 1: val_loss improved from inf to 0.34488, saving model to lstm_2rtu_smooth_04.keras\n",
+      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m165s\u001b[0m 20ms/step - loss: 0.0548 - val_loss: 0.3449\n",
       "Epoch 2/2\n",
-      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step - loss: 8.6307e-04\n",
-      "Epoch 2: val_loss improved from 0.08234 to 0.07465, saving model to lstm_2rtu_smooth_03.keras\n",
-      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m232s\u001b[0m 29ms/step - loss: 8.6305e-04 - val_loss: 0.0747\n"
+      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 0.0014\n",
+      "Epoch 2: val_loss improved from 0.34488 to 0.27523, saving model to lstm_2rtu_smooth_04.keras\n",
+      "\u001b[1m8067/8067\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m160s\u001b[0m 20ms/step - loss: 0.0014 - val_loss: 0.2752\n"
      ]
     },
     {
      "data": {
       "text/plain": [
-       "<keras.src.callbacks.history.History at 0x207d62cebd0>"
+       "<keras.src.callbacks.history.History at 0x207bef58850>"
       ]
      },
-     "execution_count": 25,
+     "execution_count": 115,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1529,7 +1551,7 @@
     "\n",
     "model.compile(optimizer='adam', loss='mean_squared_error')\n",
     "\n",
-    "checkpoint_path = \"lstm_2rtu_smooth_03.keras\"\n",
+    "checkpoint_path = \"lstm_2rtu_smooth_04.keras\" #\"lstm_2rtu_smooth_03.keras\"--> 3,4rtu\n",
     "checkpoint_callback = ModelCheckpoint(filepath=checkpoint_path, monitor='val_loss', verbose=1, save_best_only=True, mode='min')\n",
     "model.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=2, batch_size=64, verbose=1, callbacks=[checkpoint_callback])"
    ]
@@ -1675,12 +1697,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 284,
+   "execution_count": 89,
    "metadata": {},
    "outputs": [],
    "source": [
     "from tensorflow.keras.models import load_model\n",
-    "checkpoint_path = \"lstm_4rtu_smooth_02.keras\"\n",
+    "checkpoint_path = \"lstm_2rtu_smooth_03.keras\"\n",
     "model = load_model(checkpoint_path)"
    ]
   },
@@ -1693,15 +1715,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 26,
+   "execution_count": 116,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "\u001b[1m19190/19190\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m95s\u001b[0m 5ms/step\n",
-      "\u001b[1m16134/16134\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m83s\u001b[0m 5ms/step\n"
+      "\u001b[1m19190/19190\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m78s\u001b[0m 4ms/step\n",
+      "\u001b[1m16134/16134\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m63s\u001b[0m 4ms/step\n"
      ]
     }
    ],
@@ -1712,14 +1734,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 35,
+   "execution_count": 148,
    "metadata": {},
    "outputs": [],
    "source": [
     "%matplotlib qt\n",
-    "var = 7\n",
-    "plt.plot(y_test[:,var], label='Original Testing Data', color='blue')\n",
-    "plt.plot(test_predict1[:,var], label='Predicted Testing Data', color='red',alpha=0.8)\n",
+    "var = 5\n",
+    "plt.plot(testdataset_df['date'][31:], y_test[:,var], label='Original Testing Data')\n",
+    "plt.plot(testdataset_df['date'][31:] ,test_predict1[:,var], label='Predicted Testing Data')\n",
+    "\n",
     "# anomalies = np.where(abs(test_predict1[:,var] - y_test[:,var]) > 0.38)\n",
     "# plt.scatter(anomalies,test_predict1[anomalies,var], color='black',marker =\"o\",s=100 )\n",
     "\n",
@@ -1786,15 +1809,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 43,
+   "execution_count": 150,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[[ 0.00183319  0.01064136  0.01099219  0.01235903  0.01590783  0.0086837\n",
-      "  -0.00136298]]\n"
+      "[[ 0.00056891 -0.00128509  0.00987136 -0.00519864  0.01173089  0.00975552\n",
+      "   0.01170705]]\n"
      ]
     }
    ],
@@ -1892,20 +1915,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 41,
+   "execution_count": 185,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x207c1f8ce90>]"
-      ]
-     },
-     "execution_count": 41,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "%matplotlib qt\n",
     "\n",
@@ -1915,17 +1927,22 @@
     "# distance4 = np.linalg.norm((test_predict1[:,22:29]-y_test[:,22:29])-kmeans4.cluster_centers_[0], ord=2, axis = 1)\n",
     "# distance_a = np.linalg.norm(test_predict1[:,8:]-a, ord=2, axis = 1)\n",
     "# plt.plot(y_test[:,23],alpha=0.6)\n",
-    "# plt.plot(y_test[:,3],alpha=0.6)\n",
+    "# plt.plot(y_test[:,7],alpha=0.6)\n",
     "# plt.plot(y_test[:,6],alpha=0.6)\n",
-    "plt.plot(y_test[:,7],alpha=0.6)\n",
+    "# plt.plot(testdataset_df['date'][31:] ,y_test[:,14],alpha=0.6,label='Return air fan speed')\n",
+    "# plt.plot(testdataset_df['date'][31:] ,y_test[:,9],alpha=0.6,label='Damper position')\n",
     "# plt.plot(test_predict1[:,3],alpha=0.6)\n",
-    "plt.plot(abs(distance1)>2.5)\n",
+    "plt.plot(testdataset_df['date'][31:] ,abs(distance2)>3.5,linewidth=2.5,label='Faults')\n",
+    "plt.plot(testdataset_df['date'][31:] ,pd.Series((distance2)>3.5).rolling(window=60,min_periods=1).mean()==1,linewidth=2.5,label='Faults')\n",
     "# plt.plot(abs(distance2)>3.5)\n",
     "# plt.plot(abs(distance3)>5)\n",
     "# plt.plot(abs(distance4)>5)\n",
     "# plt.plot(distance_a>8,c='g')\n",
-    "\n",
-    "    "
+    "plt.title('RTU-2 Faults')\n",
+    "plt.xlabel('Time')\n",
+    "plt.ylabel('Value')\n",
+    "plt.legend()\n",
+    "plt.show()\n"
    ]
   },
   {
@@ -1981,31 +1998,25 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 77,
+   "execution_count": 166,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "array([[ 5.8607887e-02, -2.4713947e-01,  2.4978706e-01, -7.8289807e-01,\n",
-       "        -2.0218764e-01, -2.8860569e-01,  2.7817219e-01,  2.4209845e-01],\n",
-       "       [-2.6845999e-02,  1.2596852e-01,  9.6294099e-01,  2.0099232e-01,\n",
-       "         3.3391420e-02,  7.7613303e-04, -7.1204931e-02, -9.7836025e-02]],\n",
-       "      dtype=float32)"
-      ]
-     },
-     "execution_count": 77,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
-    "pca.components_"
+    "distance2 = np.linalg.norm((test_predict1[:,8:15]-y_test[:,8:15])-kmeans2.cluster_centers_[0], ord=2, axis = 1)\n",
+    "X = pca.transform((test_predict1[:,8:15]-y_test[:,8:15])-kmeans2.cluster_centers_[0])\n",
+    "centroids = pca.transform(kmeans2.cluster_centers_)\n",
+    "\n",
+    "plt.scatter(X[:, 0], X[:, 1], c = abs(distance2)>10 ,cmap='viridis', alpha=0.5)\n",
+    "plt.scatter(centroids[:, 0], centroids[:, 1], marker='x', c='red', s=200, linewidths=2)\n",
+    "plt.title('KMeans Clustering')\n",
+    "plt.xlabel('Feature 1')\n",
+    "plt.ylabel('Feature 2')\n",
+    "plt.show()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 84,
+   "execution_count": 155,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2026,7 +2037,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 38,
+   "execution_count": 154,
    "metadata": {},
    "outputs": [
     {
@@ -2036,7 +2047,7 @@
      "traceback": [
       "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
       "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
-      "Cell \u001b[1;32mIn[38], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m X \u001b[38;5;241m=\u001b[39m \u001b[43mprocessed_data\u001b[49m\n\u001b[0;32m      4\u001b[0m k \u001b[38;5;241m=\u001b[39m \u001b[38;5;241m1\u001b[39m\n\u001b[0;32m      6\u001b[0m kmeans \u001b[38;5;241m=\u001b[39m KMeans(n_clusters\u001b[38;5;241m=\u001b[39mk,random_state\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m4\u001b[39m)\n",
+      "Cell \u001b[1;32mIn[154], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m X \u001b[38;5;241m=\u001b[39m \u001b[43mprocessed_data\u001b[49m\n\u001b[0;32m      4\u001b[0m k \u001b[38;5;241m=\u001b[39m \u001b[38;5;241m2\u001b[39m\n\u001b[0;32m      6\u001b[0m kmeans \u001b[38;5;241m=\u001b[39m KMeans(n_clusters\u001b[38;5;241m=\u001b[39mk,random_state\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m4\u001b[39m)\n",
       "\u001b[1;31mNameError\u001b[0m: name 'processed_data' is not defined"
      ]
     }