TFR-HSS-Benchmark/samples

TFRD

This is samples from TFRD dataset for temperature field reconstruction of heat source systems task (TFR-HSS).

download site: https://pan.baidu.com/s/14BipTer1fkilbRjrQNbKiQ, password: tfrd

author: Zhiqiang Gong

contact: gongzhiqiang13@nudt.edu.cn

time: 2021-08-24

Structures

TFRD contains three sub-data: namely the HSink, the ADlet, and the DSine corresponding to the three typical sub-task.

• TFRD // The general list of TFRD

  • HSink // Data for HSink sub-task

    • train // Data for training

      The training process read the training samples from train_val.txt and find the corresponding sample in train/

      • train // training samples
      • train_val.txt // sample list for training

    • test

      Each test consists six test sets:

      test 0 represents the general test samples

      test 1 describes the test set where all the components share the same power density

      test 2 describes the test set where 1/4 of the heat sources are with zero-power intensity and the remainder are with random selected power intensity.

      test 3 describes the test set where half of the heat sources are with zero-power intensity and half with random selected power intensity.

      test 4 describes the test set where 3/4 of the heat sources are with zero-power intensity and the remainder are with random selected power intensity.

      test 5 describes the test set where only one heat source is with random selected intensity and the remainder are with zero-power intensity.

      • test_0 // test 0
      • test_0.txt // sample list for test 0
      • test_1 // special test 1
      • test_1.txt // sample list for special test 1
      • test_2 // special test samples 2
      • test_2.txt // sample list for special test 2
      • test_3 // special test samples 3
      • test_3.txt // sample list for special test 3
      • test_4 // special test samples 4
      • test_4.txt // sample list for special test 4
      • test_5 // special test samples 5
      • test_5.txt // sample list for special test 5

  • ADlet // Data for ADlet sub-task

    • train
      • train
      • train_val.txt
    • test
      • test_0
      • test_0.txt
      • test_1
      • test_1.txt
      • test_2
      • test_2.txt
      • test_3
      • test_3.txt
      • test_4
      • test_4.txt
      • test_5
      • test_5.txt

  • DSine // Data for DSine sub-task

    • train
      • train
      • train_val.txt
    • test
      • test_0
      • test_0.txt
      • test_1
      • test_1.txt
      • test_2
      • test_2.txt
      • test_3
      • test_3.txt
      • test_4
      • test_4.txt
      • test_5
      • test_5.txt

Variables

• F: The component information, including the power density, position and shape
• u: Real temperature field of the specific F in 200*200 discretized matrix
• u_obs: Temperature value of the monitoring points
• u_pos: Positions of monitoring points in 200*200 discretized matrix (1 describes area monitoring points and 0 without monitoring points)
• xs, ys, zs: Corresponding coordinates of the 200*200 discretized matrix

Examples

• General examples

HSink	ADlet	DSine

• Special examples of HSink

Test 0	Test 1	Test 2
Test 3	Test 4	Test 5

Citing this work

If this dataset is helpful for your research, please consider citing:

@article{gong2021,
    Author = {Xiaoqian Chen and Zhiqiang Gong and Xiaoyu Zhao and Weien Zhou and Wen Yao},
    Title = {A Machine Learning Modelling Benchmark for Temperature Field Reconstruction of Heat-Source Systems},
    Journal = {arXiv preprint arXiv:2108.08298},
    Year = {2021}
}