Distributed spacecraft missions (DSM) are gaining traction in the space community for the potential to deploy multiple simple and low-cost spacecraft to provide high temporal resolution of observations over regions of interest. Designing a DSM, however, remains challenging due to the many architectural decisions involved at both the spacecraft-level and system-level and due to the presence of multiple conflicting objectives of maximizing performance while minimizing cost and risk. One proven approach to architecting DSMs is to gather a team of design experts from multiple disciplines and leverage their combined experiences to quickly identify a few feasible mission architectures. The rapid design process allows the team to refine the design problem several times to better capture the stakeholders' needs. A typical design team, however, only explores a handful of alternative missions, which limits their understanding of the key design decisions that determine a DSM's metrics. Another popular approach is to convert the design problem into an optimization problem and rely on search algorithms to explore more of the tradespace. In particular, multiobjective evolutionary algorithms (MOEA) have shown promise on DSM design problems, but they are considered computationally inefficient because they generally don't leverage the available domain- or problem-specific knowledge. To identify promising missions, MOEAs require evaluating hundreds or thousands of candidate solutions using computationally expensive simulations to compute their metrics. These computational burdens hinder designers from iterating through multiple problem formulations. This thesis proposes a new tradespace exploration tool that combines the efficiency of expert design heuristics with the explorative power of an MOEA. The tool exploits the available expert knowledge to push the exploration to the most promising regions of the tradespace while searching other regions of the tradespace for novel solutions not captured by the knowledge. First, expert design knowledge is encoded as knowledge-dependent operators so that they can easily be incorporated into MOEAs. Next, an MOEA is augmented with an adaptive operator selection strategy (AOS) that allows an MOEA to efficiently utilize multiple evolutionary operators by constantly monitoring each operator's ability to create high-quality solutions and adapting the search strategy to apply the most effective ones. Given several knowledge-dependent operators along with conventional, knowledge-independent operators, the AOS can explore the tradespace by leveraging both the expert design heuristics and the explorative power of an MOEA. This thesis also develops an MOEA that can extract new knowledge by applying a data mining algorithm to candidate solutions generated during an optimization run. This tool is useful when there are few or no design heuristics available for a given problem. The proposed tool encodes the extracted knowledge as evolutionary operators and uses them with an AOS to guide the remainder of the optimization process. The extracted knowledge is also provided to the user in an easy-to-understand form, with the hope that the information can help decipher the results and elucidate the key design decisions. The efficacy of the proposed tradespace exploration tools are demonstrated on a DSM design problem for climate monitoring. The results shows that combining an MOEA with knowledge from experts or data mining algorithms leads to significant improvements in computational efficiency over a conventional MOEA.