پريناز اسماعيلي

In the usual repetitive dynamic games in the literature, after a few rounds, it finally reaches an evolutionary stable strategy, but in some models, during a repetition due to a sudden, discrete jump, the game may experience a catastrophe that shifts the equilibrium to a new level. As two clear examples of this type of events, it can be referred to labor strikes and people's behavior in pandemics. In this thesis, we have tried to observe these discrete changes in order to prevent sudden jumps if it they are undesirable by predicting the time and place of the catastrophe, and if it is desirable, to use it to find out the least necessary effort towards reaching the goals. In order to eva‎luate the effectiveness of catastrophe theory in dynamic evolutionary models, several scenarios of these games are modeled in this thesis. For investigating the worker-employer relationship in the strike threat as an example of an evolutionary game with the possibility of a catastrophe occurrence, four scenarios of classic game are considered. The relationship between the players gradually becomes more complicated. In model 1, information is not exchanged between players during decision making (prisoner's dilemma), in model 2 (chicken game) the possibility of observing decisions is added. In this case, the strike is done and the players decide to compromise or not. In model 3 (ultimatum game) it is possible to bargain, and reaching to the middle solution will be possible. In model 4 (coalition game) in addition to the above items, we will have three categories of players that can form an alliance between each other and each side of the game will tend to cooperate with neutral members to increase their bargaining power. The solution approach of models 1-4 is the replicator dynamics that have many limiting assumptions. A network-based approach is used in models 5 and 6 to release these constraints and by using the mechanisms such as punishment and reward, an attempt will be made to eliminate free-riding and increase cooperation in the pandemics as a clear example of social dilemmas. In each model, with the help of the evolutionary game structure, the evolutionary stable strategy is determined and the catastrophic theory predicts the time and the location of the catastrophe occurrence and its borders. This gives decision makers the managerial insights that the minimum required effort to achieve the goals and prevent the system collapse.

نظريه‌ي فاجعه , بازي هاي تكاملي , بازي معماي زنداني , بازي جوجه , شبيه سازي چندعامله , اعتصاب , پاندمي

Catastrophe theory , evolutionary games , Prisoner dilemma game , Chicken game , Multi-Agent Simulation , strike , pandemy

Parinaz Esmaeili

Seyed Mohammad SeyedHosseini, Ahmad Makui