Data for the Discrete-Time State-of-Charge Estimator for Latent Heat Thermal Energy Storage Units Based on a Recurrent Neural Network

https://doi.org/10.17035/d.2024.0322710137
Open external identifier URL

dataset

posted on 2024-10-30, 07:34 authored by Jose Bastida HernandezJose Bastida Hernandez, Ivan De la Cruz Loredo, Pranaynil SaikiaPranaynil Saikia, Carlos Ugalde LooCarlos Ugalde Loo

The datasets describe the simulation results of the mathematical model, discrete non-linear state observer and estimators based on long short-term memory recurrent neural networks (LSTM-RNN) of a latent heat-based thermal energy storage for cooling applications (ice tank). Data for the training of the recurrent neural networks is also included. The data is provided in txt files as timeseries. The txt files are named according to the number of the figure and plots. Dataset 1 (Fig 10)Charging and discharging data of state of charge under conditions established in Table 2.The processes are labelled with letter s and number, thus, for charging there are 1681 profiles (s1 to s1681 with time t) in the file: Fig10_charging.txt. For the discharging process there are 451 profiles labelled with letter s and number, as well, in file: Fig10_discharging.txt (s1-s451 with time t).Dataset 2 (Fig 21)State-of-charge (SoC) estimated profiles using the mathematical model of the ice tank and eight different RNN’s architecture. For charging: Fig21a_socs.txt (label soc with time t) and Fig21a_socxy.txt (labels s1 to s8I with time t). For discharging: Fig21b_socs.txt (label soc time t) and Fig21b_socxy.txt (labels s1 to s8 time t).Dataset 3 (Fig 24)Comparison of the SoC employing matrix operations in the RNNs estimator output. The SoC calculated directly by the mathematical model is in file: Fig24_socs.txt (label soc with time t), the output of the RNN1,10 and RNN2,15 are included in Fig24_RNNs.txl (label sE2a for RNN1,10 with time tE2a and sC2a for RNN2,15 with time tC2a).Dataset4 (Fig 26)Comparison of the SoC calculated by the mathematical model of the ice tank, the RNNs estimators and the non-liner observer for five discharging-charging cycles. The SoC of the ice tank model is in file: Fig26_socs.txt (label soc with time t), the SoC estimated by the nonlinear observer is in file: Fig26_soco.txt (label soc with time t), the SoC estimated by the RNN1,10 is in file: Fig26_RNN110 (label soc with time t) and the SoC estimated by RNN2,15 is in file: Fig26_RNN2,15 (label soc with time t).Dataset 5 (Fig 27)Comparison of the SoC calculated by the mathematical model of the ice tank, the nonlinear observer, the RNNs estimators for 25 discharging-charging cycles. The SoC of the ice tank modes in in file: Fig27_socs.txt (label soc with time t), the SoC of the nonlinear observer is in file: Fig27_soco.txt (label soc with time t), the SoC estimated by the RNN1,10 is in file: Fig27_socRNN110.txt (label soc with time t), and the SoC estimated by the RNN2,15 is in file : Fig27_socRNN210.txt (label soc with time t).Dataset 7 (Fig 28)Comparison of the SoC calculated by the mathematical model of the ice tank, the nonlinear observer, and RNNs with different sampling time (ts) (subscript A for ts = 120s and subscript B for ts = 600s) for two charging-discharging cycles. The SoC given by the mathematical model of the ice tank is in: Fig28_socs.txt (label soc with time t). The SoC given by the nonlinear observer is in file : Fig28_soco.txt (label soc with time t). The SoC estimated by the RNN1,10 with a sampling time of 120 s is in file: Fig28_RNN110A.txt (label soc with time t) and with sampling time of 600 s is in file: Fig28_RNN110B.txt (label soc with time t). Finally, the SoC estimated by the RNN2,15 with a sampling time of 120 s is in file: Fig28_RNN215A.txt (label soc with time t) and with sampling time of 600s is in file: Fig28_RNN215B.txt (label soc with time t).Dataset 8 (Fig 30)The estimation of the SoC using the mathematical model and the RNN-LSTM under different initial conditions for charging the files are listed in Table 1.Table 1. Listed of the files of Figure 30 for comparison of the SoC with different initial conditions.<table><tr><td>Charging</td></tr><tr><td>For SoC with mathematical models</td><td>For SoC with RNN estimator</td><td>Labels</td></tr><tr><td>Fig30_socs_blue</td><td>Fig30_RNN_blue</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_red</td><td>Fig30_RNN_red</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_gray</td><td>Fig30_RNN_gray</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_orange</td><td>Fig30_RNN_orage</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_magenta</td><td>Fig30_RNN_magenta</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_green</td><td>Fig30_RNN_green</td><td>label soc with time t</td></tr><tr><td>Fig30_socs_cyan</td><td>Fig30_RNN_cyan</td><td>label soc with time t</td></tr></table> For discharging process the files of comparisons are listed in Table 2.<table><tr><td>Charging</td></tr><tr><td>For SoC with mathematical models</td><td>For SoC with RNN estimator</td><td>Labels</td></tr><tr><td>Fig30B_socs_blue</td><td>Fig30B_RNN_blue</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_red</td><td>Fig30B_RNN_red</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_gray</td><td>Fig30B_RNN_gray</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_orange</td><td>Fig30B_RNN_orage</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_magenta</td><td>Fig30B_RNN_magenta</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_green</td><td>Fig30B_RNN_green</td><td>label soc with time t</td></tr><tr><td>Fig30B_socs_cyan</td><td>Fig30B_RNN_cyan</td><td>label soc with time t</td></tr></table>Dataset 9 (Fig 31)Comparison of the SoC calculated with the mathematical model of the ice tank and the RNN estimator using different initial conditions during training of the RNN. The SoC of the mathematical model is in file: Fig31_socs.txt (label soc with time t) and the SoC estimated by the RNN is in file: Fig31_RNN.txt (label soc with time t).Reserch results based upon these data are published at https://doi.org/10.1016/j.apenergy.2024.123526