Parameter estimation for a linear differential equation from multiple measured time series data

Hi Everyone,

I have followed the example showed in the following link.

https://www.mathworks.com/help/optim/ug/fit-ode-problem-based-least-squares.html

It was a very great way to help met get started with estimating 3 parameters for a differential equation from measured data. However, I have about 50 time series data with 50 different initial conditions. I wonder how the example in the link above can be modified for me to use all time series data, instead of only one, to determine the best fit for my differential equation.

I feel that obtaining 50 different set of parameters by fiting 50 data set independently then averaging the parameters is not a great way to estimate the parameters.

Thank you and I am open to suggestions.