Combining these two objective functions using weighting coefficients according to the importance of of each one, we get one objective function:
where and are the weighting coefficients.
By minimizing the new objective function over the link lengths, we can get values for the link lengths. If the generated function is differentiable, then we can get a direct solution to the problem. Usually, there is more than one solution for selecting the link lengths (for more than two link robots). In this case, some other rules can be used to select among these solutions. For example, each link may have bounds on its length. Also, selecting a solution which gives similar length for several links is better from the manufacturing point of view. On the other hand, if the function is not differentiable, then we can use a heuristic optimization technique such as pattern search which does not require gradient information.