20.500.12556/RUL-186799 Napovedovanje preostanka življenjske dobe baterije z uporabo nevronskih mrež LSTM Prediction of battery remaining useful life using LSTM neural networks Litij-ionske baterije med uporabo postopoma izgubljajo kapaciteto zaradi kompleksnih, nelinearnih degradacijskih procesov, odvisnih od pogojev delovanja. Natančno napovedovanje preostale uporabne življenjske dobe je zato ključno za zasnovo baterije, za zagotavljanje varnosti, zanesljivosti in stroškovne učinkovitosti baterijskih sistemov. Za napovedovanje upadanja kapacitete devetih litij-ionskih celic, staranih pod različnimi pogoji cikliranja, je bila uporabljena nevronska mreža LSTM. Iz meritev toka, napetosti in temperature so bile za vsak cikel izračunane vhodne značilke, saj je bil namen raziskati zmogljivosti računsko učinkovitih metod strojnega učenja. Preizkušena sta bila dva pristopa: sprotno napovedovanje kapacitete na podlagi vseh preteklih obratovalnih podatkov ter napovedovanje nadaljnjega poteka degradacije na podlagi prvih 200 ciklov delovanja. Model je bil ovrednoten z metodo izločitve ene baterije (leave-one-battery-out) z metrikami RMSE, MAE in R2. Model je pri sprotnem napovedovanju dosegel MAE = 1,34 in RMSE = 1,65 ter razmeroma zanesljivo napovedal konec življenjske dobe baterijskih celic. Napovedovanje nadaljnjega poteka degradacije na podlagi prvih 200 ciklov se je izkazalo za manj ustrezno (MAE = 8,56, RMSE = 9,77): napovedana kapaciteta je po začetnem odstopanju ostala približno konstantna in ni sledila dejanskemu upadanju. Lithium-ion batteries gradually lose capacity during operation due to complex, nonlinear degradation processes that depend on operating conditions. Accurate prediction of the remaining useful life is therefore essential for battery design and for ensuring the safety, reliability, and cost-effectiveness of battery systems. An LSTM neural network was used to predict the capacity degradation of nine lithium-ion cells aged under different cycling conditions. Input features were calculated for each cycle from current, voltage, and temperature measurements, with the aim of investigating the performance of computationally efficient machine learning methods. Two approaches were evaluated: continuous capacity prediction based on all previously available operating data, and prediction of the subsequent degradation trajectory based on the first 200 operating cycles. The model was evaluated using a leave-one-battery-out cross-validation approach and the RMSE, MAE, and R² metrics. In continuous capacity prediction, the model achieved an MAE of 1.34 and an RMSE of 1.65. It also predicted the end of life of the battery cells with relatively high reliability. In contrast, predicting the subsequent degradation trajectory based on the first 200 cycles proved unsuitable, with an MAE of 8.56 and an RMSE of 9.77. After an initial deviation, the predicted capacity remained approximately constant and did not follow the actual capacity decline. litij-ionske baterije Degradacija Napovedovanje LSTM lithium-ion battery degradation forecasting LSTM true false false [A. Žabkar] Slovenski jezik Angleški jezik Diplomsko delo/naloga 2026-09-05 08:33:10 2026-09-05 08:33:14 2026-09-08 10:41:44 0000-00-00 00:00:00 2026 0 Ljubljana 0 IX, 33 f. 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 621.352:004.94(043.2) 289389 290339331 49319.pdf 49319.pdf 1 D061290F924ED18A09ABC969F769076C b3f847bcf0148e1842c572ab45507fb69af50d63aee25772a4778aa43d7b8c78 aae16afd-a8f3-11f1-8bc5-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=246170 Fakulteta za strojništvo 0 0 0