This way we can be sure that we are always using the same version and also it avoids setting it up in the configuration

See Table 1 of https://arxiv.org/pdf/2210.08060.pdf

Create function to compute $\frac{\partial V}{\partial \dot{q}}$ from the optimal input $u^*$

To sort it out we will need to complete the following steps:

• Finish polishing the paper according to the reviews
• Add Anastasija as co-author of the paper
• Add Remark about mechanical systems. If there is sufficient space, add example in paper, otherwise we will only add a link pointing to the GitHub repository
• Create a public repository and add link to the paper to disseminate result

Implement modified DMP:

and compare if it performs better than classic DMP.

If we do this, it will be easier to review the code (we will be able to quickly see the visual diff in GitHub 😀)

Will need to use a self-hosted MATLAB or make the repository public (will not do it for now)

Hi @demerdjieva, I suspect that we missed an minus sign here. Please also double-check this equation by checking (11) and (12) of our IFAC paper and also the computations that you wrote for the double pendulum.

Suggestion: create manually a trajectory where the second link do a circular motion while the first link is stationary

Paper: https://link.springer.com/article/10.1007/s10514-022-10067-4
Code: https://github.com/Slifer64/novel-DMP-constraints/blob/main/simulations/fig1a_offline_opt_pos_vs_vel.m

Search for realistic values for the mass and length parameters of 2DoF robot