Description


GenSSI is a toolbox that requires MATLAB and Symbolic Math Toolbox. It offers a technique for studying structural identifiability using iterative Lie derivatives and identifiability tableaus. The computational steps are: Generate Lie derivatives. It is an iterative process and computes the Lie derivatives of the analytic output with respect to the uncontrolled analytic function, respectively to the input function of the model. The number of iterations is the number of derivatives given by the user. The Lie derivatives evaluated at the initial state form the generating series coefficients. Complete identifiability tableau is a pseudocolor plot representation of the Jacobian of the nonzero generating series coefficients. Reduced identifiability tableaus is made of independent rows from the complete tableau such that the rank remains unchanged. If in the identifiability tableau there are rows with just one element, then the corresponding parameters are globally identifiable, and their rows and columns are eliminated, generating the 2nd reduced identifiability tableau. The procedure is repeated for the parameters that require 2/more equations to be identified until the remaining tableau can't be reduced anymore. Solution of the parameters. If a unique solution is found, the model is structurally globally identifiable. If it has more than one solution, the model is structurally locally identifiable. Otherwise, if no solution could be found, the parameters are structurally nonidentifiable. 
