Investment model cost parameters for VSC HVDC transmission infrastructure continue to be associated with high uncertainty and their validity remains a crucial challenge. Thus, it is the key objective of this analysis to identify a new cost parameter set providing better investment cost estimates than currently available cost parameter sets. This parameter estimation is based on a previously conducted review of investment model cost parameters including its collection of existing cost parameter sets and project cost reference data. By using a particle swarm optimisation, the overall error function of the review's evaluation methodology is minimised to obtain an optimal parameter set. The results show, however, that the optimised parameter sets are far from being realistic and useful, which is why an improved overall error function is developed. Effectively penalising negative and near-zero cost parameter coefficients, this new overall error function delivers a realistic and well-performing cost parameter set when being minimised. In fact, the new parameter set produces better cost estimates for back-to-back, interconnector, and offshore wind connection projects than any of the existing cost parameter sets. Therefore, it is a valuable contribution and shall be considered in future grid investment analyses involving VSC HVDC technology.