Linking Strategic and Operational Decision-Making in System of Systems Using Game Theory - the Case of Circular Economy for Photovoltaics

In a System of Systems (SoS), constituent systems (CS) interact. Each CS may pursue individual goals while contributing to shared goals. Often, bi-level decision problems occur on the strategic (SoS design) and operational (SoS operation) planning levels. We describe the location and allocation planning for photovoltaic (PV) recyclers in a circular economy (CE) as an instantiation of this problem. The complex interactions and interdependencies between managerial independent manufacturers, collectors, and recyclers can be understood as SoS with circularity as a shared goal. Each CS must act based on incomplete information, considering systemic factors and information about other CSs. Suitable approaches to support the decision-making for this problem type are needed. Generally, CS's decision-making can be modeled through Game Theory (GT). We conduct a systematic literature review on the use of GT in SoS Engineering (SoSE) to identify how existing approaches address the described bi-level problem. While other bi-level problems and problems with individual and shared goals are discussed, no approach studies the linkage between strategic and operational planning. Therefore, we illustrate the linkage of strategic and operational planning levels for the location and allocation planning of PV recyclers as a non-cooperative game. A market maker approach is proposed to facilitate external coordination and support the achievement of circularity. Approaches from distributed model predictive control (DMPC) are discussed regarding their applicability to bi-level decision problems in SoS. We propose the Distributed Command Governor Approach to solve the distributed non-cooperative game. The proposed model can be transferred to SoS with collaborative management and collaborative or acknowledged operation.