├── README.md
└── LSTM_Github.ipynb


/README.md:
--------------------------------------------------------------------------------
1 | # LSTM-based-Capacity-Estimation-of-Lithium-Ion-Battery-for-Electric-Vehicles
2 | 
3 | Lithium Ion batteries have been extensively used for many applications such as laptops, mobile phones and electric vehicles due its long cycle lie, high power and high energy densities. The life of battery is affected by many different factors including cycles, discharge current, charge current, charge voltage, temperature, and state of charge ranges (depth of discharge). 
4 | This project predicts the capacity of the Lithium Ion battery with LSTM based on the Voltage, Current and Temperature of the charging cycles [1]. Keras deep learning library has been utilized to implement LSTM. Lithium ion battery data has been taken from NASA Ames Prognostics Data Repository [2]. 
5 | 
6 | [1] Choi, Yohwan, et al. "Machine Learning-Based Lithium-Ion Battery Capacity Estimation Exploiting Multi-Channel Charging Profiles." IEEE Access 7 (2019): 75143-75152.
7 | 
8 | [2] B. Saha and K. Goebel (2007). "Battery Data Set", NASA Ames Prognostics Data Repository (https://ti.arc.nasa.gov/tech/dash/groups/pcoe/prognostic-data-repository/#battery), NASA Ames Research Center, Moffett Field, CA 
9 | 


--------------------------------------------------------------------------------
/LSTM_Github.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": 1,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "import numpy as np\n",
 10 |     "from mat2json import loadMat\n",
 11 |     "from keras.models import Sequential\n",
 12 |     "from keras.layers import Dense\n",
 13 |     "import keras\n",
 14 |     "from keras.layers import Dense, Dropout, Flatten\n",
 15 |     "from keras.layers import LeakyReLU\n",
 16 |     "from keras.layers import LSTM\n",
 17 |     "from keras.callbacks import EarlyStopping\n",
 18 |     "from sklearn.preprocessing import MinMaxScaler\n",
 19 |     "import matplotlib.pyplot as plt"
 20 |    ]
 21 |   },
 22 |   {
 23 |    "cell_type": "code",
 24 |    "execution_count": 2,
 25 |    "metadata": {},
 26 |    "outputs": [],
 27 |    "source": [
 28 |     "B0005 = loadMat('B0005.mat')"
 29 |    ]
 30 |   },
 31 |   {
 32 |    "cell_type": "code",
 33 |    "execution_count": 3,
 34 |    "metadata": {},
 35 |    "outputs": [],
 36 |    "source": [
 37 |     "def extract_discharge(Battery):\n",
 38 |     "    \n",
 39 |     "    cap = []\n",
 40 |     "    i = 1\n",
 41 |     "    for Bat in Battery:\n",
 42 |     "        if Bat['cycle'] == 'discharge':\n",
 43 |     "            cap.append((Bat['data']['Capacity'][0]))\n",
 44 |     "            i+=1\n",
 45 |     "    return cap"
 46 |    ]
 47 |   },
 48 |   {
 49 |    "cell_type": "code",
 50 |    "execution_count": 6,
 51 |    "metadata": {},
 52 |    "outputs": [],
 53 |    "source": [
 54 |     "A = extract_charge_VIT(B0005)\n",
 55 |     "#print(A)\n",
 56 |     "InitC = 1.85;\n",
 57 |     "cap5 = extract_discharge(B0005);"
 58 |    ]
 59 |   },
 60 |   {
 61 |    "cell_type": "code",
 62 |    "execution_count": 7,
 63 |    "metadata": {},
 64 |    "outputs": [],
 65 |    "source": [
 66 |     "xData_raw = A\n",
 67 |     "comp = len(A) - len(cap5); \n",
 68 |     "yData_raw = np.vstack((InitC*np.ones((comp, 1)), np.reshape(cap5, (len(cap5), 1))))"
 69 |    ]
 70 |   },
 71 |   {
 72 |    "cell_type": "code",
 73 |    "execution_count": 8,
 74 |    "metadata": {},
 75 |    "outputs": [],
 76 |    "source": [
 77 |     "xminmax = MinMaxScaler(feature_range=(0, 1)) # xData_raw feature scaling \n",
 78 |     "xData = xminmax.fit_transform(xData_raw) # feature scaling of xData_raw\n",
 79 |     "#print(xData)\n",
 80 |     "yminmax = MinMaxScaler(feature_range=(0, 1)) # yData_raw feature scaling \n",
 81 |     "yData = yminmax.fit_transform(yData_raw) # feature scaling of yData_raw\n",
 82 |     "#print(yData)"
 83 |    ]
 84 |   },
 85 |   {
 86 |    "cell_type": "code",
 87 |    "execution_count": 12,
 88 |    "metadata": {},
 89 |    "outputs": [],
 90 |    "source": [
 91 |     "# split the data into train and test\n",
 92 |     "X_train, X_test, y_train, y_test = train_test_split(xData, yData, test_size = 0.20,shuffle = False)# split the data into train and test"
 93 |    ]
 94 |   },
 95 |   {
 96 |    "cell_type": "code",
 97 |    "execution_count": 13,
 98 |    "metadata": {},
 99 |    "outputs": [
100 |     {
101 |      "name": "stdout",
102 |      "output_type": "stream",
103 |      "text": [
104 |       "Epoch 1/200\n",
105 |       "4/4 [==============================] - 0s 79ms/step - loss: 0.3492 - val_loss: 0.2346\n",
106 |       "Epoch 2/200\n",
107 |       "4/4 [==============================] - 0s 10ms/step - loss: 0.2331 - val_loss: 0.1640\n",
108 |       "Epoch 3/200\n",
109 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.1775 - val_loss: 0.1282\n",
110 |       "Epoch 4/200\n",
111 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.1432 - val_loss: 0.1115\n",
112 |       "Epoch 5/200\n",
113 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.1179 - val_loss: 0.1024\n",
114 |       "Epoch 6/200\n",
115 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.1155 - val_loss: 0.0951\n",
116 |       "Epoch 7/200\n",
117 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.1033 - val_loss: 0.0883\n",
118 |       "Epoch 8/200\n",
119 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0989 - val_loss: 0.0822\n",
120 |       "Epoch 9/200\n",
121 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0866 - val_loss: 0.0771\n",
122 |       "Epoch 10/200\n",
123 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0808 - val_loss: 0.0730\n",
124 |       "Epoch 11/200\n",
125 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0827 - val_loss: 0.0692\n",
126 |       "Epoch 12/200\n",
127 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0781 - val_loss: 0.0658\n",
128 |       "Epoch 13/200\n",
129 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0744 - val_loss: 0.0624\n",
130 |       "Epoch 14/200\n",
131 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0698 - val_loss: 0.0585\n",
132 |       "Epoch 15/200\n",
133 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0659 - val_loss: 0.0547\n",
134 |       "Epoch 16/200\n",
135 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0607 - val_loss: 0.0509\n",
136 |       "Epoch 17/200\n",
137 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0576 - val_loss: 0.0475\n",
138 |       "Epoch 18/200\n",
139 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0557 - val_loss: 0.0444\n",
140 |       "Epoch 19/200\n",
141 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0508 - val_loss: 0.0421\n",
142 |       "Epoch 20/200\n",
143 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0459 - val_loss: 0.0401\n",
144 |       "Epoch 21/200\n",
145 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0478 - val_loss: 0.0381\n",
146 |       "Epoch 22/200\n",
147 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0474 - val_loss: 0.0365\n",
148 |       "Epoch 23/200\n",
149 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0457 - val_loss: 0.0351\n",
150 |       "Epoch 24/200\n",
151 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0394 - val_loss: 0.0339\n",
152 |       "Epoch 25/200\n",
153 |       "4/4 [==============================] - 0s 10ms/step - loss: 0.0480 - val_loss: 0.0323\n",
154 |       "Epoch 26/200\n",
155 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0402 - val_loss: 0.0306\n",
156 |       "Epoch 27/200\n",
157 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0395 - val_loss: 0.0294\n",
158 |       "Epoch 28/200\n",
159 |       "4/4 [==============================] - 0s 7ms/step - loss: 0.0404 - val_loss: 0.0282\n",
160 |       "Epoch 29/200\n",
161 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0380 - val_loss: 0.0271\n",
162 |       "Epoch 30/200\n",
163 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0350 - val_loss: 0.0261\n",
164 |       "Epoch 31/200\n",
165 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0351 - val_loss: 0.0251\n",
166 |       "Epoch 32/200\n",
167 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0371 - val_loss: 0.0244\n",
168 |       "Epoch 33/200\n",
169 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0334 - val_loss: 0.0234\n",
170 |       "Epoch 34/200\n",
171 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0303 - val_loss: 0.0221\n",
172 |       "Epoch 35/200\n",
173 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0347 - val_loss: 0.0208\n",
174 |       "Epoch 36/200\n",
175 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0290 - val_loss: 0.0193\n",
176 |       "Epoch 37/200\n",
177 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0291 - val_loss: 0.0181\n",
178 |       "Epoch 38/200\n",
179 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0271 - val_loss: 0.0171\n",
180 |       "Epoch 39/200\n",
181 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0287 - val_loss: 0.0164\n",
182 |       "Epoch 40/200\n",
183 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0247 - val_loss: 0.0159\n",
184 |       "Epoch 41/200\n",
185 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0257 - val_loss: 0.0152\n",
186 |       "Epoch 42/200\n",
187 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0245 - val_loss: 0.0142\n",
188 |       "Epoch 43/200\n",
189 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0238 - val_loss: 0.0132\n",
190 |       "Epoch 44/200\n",
191 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0201 - val_loss: 0.0125\n",
192 |       "Epoch 45/200\n",
193 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0242 - val_loss: 0.0117\n",
194 |       "Epoch 46/200\n",
195 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0198 - val_loss: 0.0111\n",
196 |       "Epoch 47/200\n",
197 |       "4/4 [==============================] - 0s 7ms/step - loss: 0.0182 - val_loss: 0.0108\n",
198 |       "Epoch 48/200\n",
199 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0186 - val_loss: 0.0104\n",
200 |       "Epoch 49/200\n",
201 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0164 - val_loss: 0.0100\n",
202 |       "Epoch 50/200\n",
203 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0189 - val_loss: 0.0096\n",
204 |       "Epoch 51/200\n",
205 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0220 - val_loss: 0.0090\n",
206 |       "Epoch 52/200\n",
207 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0145 - val_loss: 0.0086\n",
208 |       "Epoch 53/200\n",
209 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0171 - val_loss: 0.0085\n",
210 |       "Epoch 54/200\n",
211 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0153 - val_loss: 0.0083\n",
212 |       "Epoch 55/200\n",
213 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0152 - val_loss: 0.0078\n",
214 |       "Epoch 56/200\n",
215 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0146 - val_loss: 0.0074\n",
216 |       "Epoch 57/200\n",
217 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0162 - val_loss: 0.0071\n",
218 |       "Epoch 58/200\n",
219 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0179 - val_loss: 0.0070\n",
220 |       "Epoch 59/200\n",
221 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0142 - val_loss: 0.0071\n",
222 |       "Epoch 60/200\n",
223 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0170 - val_loss: 0.0070\n",
224 |       "Epoch 61/200\n",
225 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0151 - val_loss: 0.0068\n",
226 |       "Epoch 62/200\n",
227 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0152 - val_loss: 0.0066\n",
228 |       "Epoch 63/200\n",
229 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0153 - val_loss: 0.0066\n",
230 |       "Epoch 64/200\n",
231 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0163 - val_loss: 0.0065\n",
232 |       "Epoch 65/200\n",
233 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0146 - val_loss: 0.0063\n",
234 |       "Epoch 66/200\n",
235 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0174 - val_loss: 0.0061\n",
236 |       "Epoch 67/200\n",
237 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0135 - val_loss: 0.0062\n",
238 |       "Epoch 68/200\n",
239 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0130 - val_loss: 0.0066\n",
240 |       "Epoch 69/200\n",
241 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0158 - val_loss: 0.0065\n",
242 |       "Epoch 70/200\n",
243 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0145 - val_loss: 0.0062\n",
244 |       "Epoch 71/200\n",
245 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0131 - val_loss: 0.0060\n",
246 |       "Epoch 72/200\n",
247 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0131 - val_loss: 0.0060\n",
248 |       "Epoch 73/200\n",
249 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0109 - val_loss: 0.0062\n",
250 |       "Epoch 74/200\n",
251 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0162 - val_loss: 0.0066\n",
252 |       "Epoch 75/200\n",
253 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0111 - val_loss: 0.0067\n",
254 |       "Epoch 76/200\n",
255 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0136 - val_loss: 0.0060\n",
256 |       "Epoch 77/200\n",
257 |       "4/4 [==============================] - 0s 16ms/step - loss: 0.0129 - val_loss: 0.0058\n",
258 |       "Epoch 78/200\n",
259 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0148 - val_loss: 0.0058\n",
260 |       "Epoch 79/200\n",
261 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0124 - val_loss: 0.0063\n",
262 |       "Epoch 80/200\n",
263 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0129 - val_loss: 0.0066\n",
264 |       "Epoch 81/200\n",
265 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0112 - val_loss: 0.0060\n",
266 |       "Epoch 82/200\n",
267 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0124 - val_loss: 0.0059\n",
268 |       "Epoch 83/200\n",
269 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0134 - val_loss: 0.0059\n",
270 |       "Epoch 84/200\n",
271 |       "4/4 [==============================] - 0s 13ms/step - loss: 0.0127 - val_loss: 0.0061\n",
272 |       "Epoch 85/200\n",
273 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0123 - val_loss: 0.0059\n",
274 |       "Epoch 86/200\n",
275 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0124 - val_loss: 0.0056\n",
276 |       "Epoch 87/200\n",
277 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0127 - val_loss: 0.0055\n",
278 |       "Epoch 88/200\n",
279 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0123 - val_loss: 0.0056\n",
280 |       "Epoch 89/200\n",
281 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0116 - val_loss: 0.0057\n",
282 |       "Epoch 90/200\n",
283 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0144 - val_loss: 0.0058\n",
284 |       "Epoch 91/200\n",
285 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0105 - val_loss: 0.0056\n",
286 |       "Epoch 92/200\n",
287 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0114 - val_loss: 0.0056\n",
288 |       "Epoch 93/200\n",
289 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0124 - val_loss: 0.0056\n",
290 |       "Epoch 94/200\n",
291 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0111 - val_loss: 0.0057\n",
292 |       "Epoch 95/200\n",
293 |       "4/4 [==============================] - 0s 12ms/step - loss: 0.0128 - val_loss: 0.0057\n",
294 |       "Epoch 96/200\n",
295 |       "4/4 [==============================] - 0s 9ms/step - loss: 0.0093 - val_loss: 0.0059\n",
296 |       "Epoch 97/200\n",
297 |       "4/4 [==============================] - 0s 8ms/step - loss: 0.0104 - val_loss: 0.0058\n",
298 |       "Model: \"sequential\"\n",
299 |       "_________________________________________________________________\n",
300 |       "Layer (type)                 Output Shape              Param #   \n",
301 |       "=================================================================\n",
302 |       "lstm (LSTM)                  (None, 50)                16200     \n",
303 |       "_________________________________________________________________\n",
304 |       "dropout (Dropout)            (None, 50)                0         \n",
305 |       "_________________________________________________________________\n",
306 |       "dense (Dense)                (None, 1)                 51        \n",
307 |       "=================================================================\n",
308 |       "Total params: 16,251\n",
309 |       "Trainable params: 16,251\n",
310 |       "Non-trainable params: 0\n",
311 |       "_________________________________________________________________\n"
312 |      ]
313 |     }
314 |    ],
315 |    "source": [
316 |     "model = Sequential()\n",
317 |     "model.add(LSTM(50, input_shape=(X_train.shape[1], X_train.shape[2])))\n",
318 |     "model.add(Dropout(0.2))\n",
319 |     "model.add(Dense(1))\n",
320 |     "model.compile(loss='mean_squared_error', optimizer='adam')\n",
321 |     "\n",
322 |     "history = model.fit(X_train, y_train, epochs=200, batch_size=32, validation_data=(X_test, y_test), \n",
323 |     "                    callbacks=[EarlyStopping(monitor='val_loss', patience=10)], verbose=1, shuffle=False)\n",
324 |     "\n",
325 |     "# Training Phase\n",
326 |     "model.summary()"
327 |    ]
328 |   },
329 |   {
330 |    "cell_type": "code",
331 |    "execution_count": 14,
332 |    "metadata": {},
333 |    "outputs": [],
334 |    "source": [
335 |     "#Predicting network output\n",
336 |     "y_predict = model.predict(X_test) # predictions on test data"
337 |    ]
338 |   },
339 |   {
340 |    "cell_type": "code",
341 |    "execution_count": 15,
342 |    "metadata": {},
343 |    "outputs": [
344 |     {
345 |      "name": "stdout",
346 |      "output_type": "stream",
347 |      "text": [
348 |       "0.041771157767402624\n"
349 |      ]
350 |     }
351 |    ],
352 |    "source": [
353 |     "# calculate the prediction error\n",
354 |     "mape = np.sum(abs(y_test[:,0]- y_predict))/np.size(y_test)\n",
355 |     "print(mape)"
356 |    ]
357 |   },
358 |   {
359 |    "cell_type": "code",
360 |    "execution_count": 16,
361 |    "metadata": {},
362 |    "outputs": [],
363 |    "source": [
364 |     "y_predict_actual = yminmax.inverse_transform(y_predict.reshape(-1,1)) # get the original a"
365 |    ]
366 |   },
367 |   {
368 |    "cell_type": "code",
369 |    "execution_count": 17,
370 |    "metadata": {},
371 |    "outputs": [],
372 |    "source": [
373 |     "y_test_actual = yminmax.inverse_transform(y_test[:,0].reshape(-1,1)) # get the original "
374 |    ]
375 |   },
376 |   {
377 |    "cell_type": "code",
378 |    "execution_count": 18,
379 |    "metadata": {},
380 |    "outputs": [
381 |     {
382 |      "data": {
383 |       "text/plain": [
384 |        "[<matplotlib.lines.Line2D at 0x2dbc1c9f908>]"
385 |       ]
386 |      },
387 |      "execution_count": 18,
388 |      "metadata": {},
389 |      "output_type": "execute_result"
390 |     },
391 |     {
392 |      "data": {
393 |       "image/png": "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\n",
394 |       "text/plain": [
395 |        "<Figure size 432x288 with 1 Axes>"
396 |       ]
397 |      },
398 |      "metadata": {
399 |       "needs_background": "light"
400 |      },
401 |      "output_type": "display_data"
402 |     }
403 |    ],
404 |    "source": [
405 |     "plt.plot(y_predict_actual)"
406 |    ]
407 |   },
408 |   {
409 |    "cell_type": "code",
410 |    "execution_count": 19,
411 |    "metadata": {},
412 |    "outputs": [
413 |     {
414 |      "data": {
415 |       "text/plain": [
416 |        "[<matplotlib.lines.Line2D at 0x2dbc1d256c8>]"
417 |       ]
418 |      },
419 |      "execution_count": 19,
420 |      "metadata": {},
421 |      "output_type": "execute_result"
422 |     },
423 |     {
424 |      "data": {
425 |       "image/png": "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\n",
426 |       "text/plain": [
427 |        "<Figure size 432x288 with 1 Axes>"
428 |       ]
429 |      },
430 |      "metadata": {
431 |       "needs_background": "light"
432 |      },
433 |      "output_type": "display_data"
434 |     }
435 |    ],
436 |    "source": [
437 |     "plt.plot(y_test_actual)"
438 |    ]
439 |   },
440 |   {
441 |    "cell_type": "code",
442 |    "execution_count": 20,
443 |    "metadata": {},
444 |    "outputs": [
445 |     {
446 |      "data": {
447 |       "text/plain": [
448 |        "[Text(0.5, 1.0, 'LSTM')]"
449 |       ]
450 |      },
451 |      "execution_count": 20,
452 |      "metadata": {},
453 |      "output_type": "execute_result"
454 |     },
455 |     {
456 |      "data": {
457 |       "image/png": "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\n",
458 |       "text/plain": [
459 |        "<Figure size 864x576 with 1 Axes>"
460 |       ]
461 |      },
462 |      "metadata": {
463 |       "needs_background": "light"
464 |      },
465 |      "output_type": "display_data"
466 |     }
467 |    ],
468 |    "source": [
469 |     "fig, ax = plt.subplots(1, figsize=(12, 8))\n",
470 |     "ax.plot(np.arange(57), 1.4*np.ones((57, 1)),'k--',linewidth = 2)\n",
471 |     "ax.plot(y_predict_actual, color='black',label='Predicted Capacity')\n",
472 |     "ax.plot(y_test_actual, color='red',label='Actual Capacity')\n",
473 |     "ax.set(xlabel='Discharge Cycles', ylabel='Capacity(Ah)')\n",
474 |     "ax.set_xlim([0,57])\n",
475 |     "#ax.set_ylim([0,2])\n",
476 |     "ax.legend()\n",
477 |     "ax.set(title = 'LSTM')"
478 |    ]
479 |   },
480 |   {
481 |    "cell_type": "code",
482 |    "execution_count": 21,
483 |    "metadata": {},
484 |    "outputs": [
485 |     {
486 |      "data": {
487 |       "image/png": "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\n",
488 |       "text/plain": [
489 |        "<Figure size 576x288 with 1 Axes>"
490 |       ]
491 |      },
492 |      "metadata": {
493 |       "needs_background": "light"
494 |      },
495 |      "output_type": "display_data"
496 |     }
497 |    ],
498 |    "source": [
499 |     "plt.figure(figsize=(8,4))\n",
500 |     "plt.plot(history.history['loss'], label='Train Loss')\n",
501 |     "plt.plot(history.history['val_loss'], label='Test Loss')\n",
502 |     "plt.title('model loss')\n",
503 |     "plt.ylabel('loss')\n",
504 |     "plt.xlabel('epochs')\n",
505 |     "plt.legend(loc='upper right')\n",
506 |     "plt.show();"
507 |    ]
508 |   },
509 |   {
510 |    "cell_type": "code",
511 |    "execution_count": null,
512 |    "metadata": {},
513 |    "outputs": [],
514 |    "source": []
515 |   }
516 |  ],
517 |  "metadata": {
518 |   "kernelspec": {
519 |    "display_name": "Python 3",
520 |    "language": "python",
521 |    "name": "python3"
522 |   },
523 |   "language_info": {
524 |    "codemirror_mode": {
525 |     "name": "ipython",
526 |     "version": 3
527 |    },
528 |    "file_extension": ".py",
529 |    "mimetype": "text/x-python",
530 |    "name": "python",
531 |    "nbconvert_exporter": "python",
532 |    "pygments_lexer": "ipython3",
533 |    "version": "3.7.4"
534 |   }
535 |  },
536 |  "nbformat": 4,
537 |  "nbformat_minor": 4
538 | }
539 | 


--------------------------------------------------------------------------------