System identification dataset

The dataset contains two sub-folders:<br>1. Load sharing 2. function approximation<br>1. Load sharing ("data.zip"):<br>Each folder corresponds with a subject.<br>the monopolar, single differential and double differential data issaved in the corresponding sub-folders 'mono', 'sd' and 'dd' respectively.<br>In each subfolder, the data is saved as '30.mat','50mat',or '70.mat' corresponding with 30%,50% or 70% MVC isometric flexion-extension.<br>The recording protocol can be found the word file 'report.doc' in this folder inThe subsection: experimental recording.<br>Structure of the '.mat' files :<br>They all have the same structure:<br>Raw_Torque : The measured Torque in ADC numbersstructure 'TAB_ARV' , the EMG envelopes for 'BB', 'BR', 'TM', 'TL' (Read report for the methods and acronyms).<br>2. function approximation ("fun_approx.zip")<br><br>Multiple benchmark examples including a piecewise single variable function, five nonlinear dynamic plants with various nonlinear structures, the chaotic Mackey Glass time series (with different signal to noise ratio (SNR) and various chaotic degree) and the real-world Box-Jenkins gas furnace system are considered to verify the effectiveness of the proposed FJWNN model. <br>The description ("info.pdf") and the entire simulated data as well as the results of our method on the training and test sets (in excel files) were provided.