A Function Allocation Strategy for Human–Machine Systems in Armored Vehicles Based on Evolutionary Game Theory and System Dynamics

Abstract

The complex battlefield conditions can cause unreasonable function allocations of the human–machine system in armored vehicles, decreasing the combat safety and efficiency. Aiming to optimize the function allocation in typical combat tasks, this study proposes a cooperation strategy by integrating evolutionary game theory with system dynamics. Taking the crew and the automated system as different players in the evolutionary game model, the payment matrix is established. The evolutionary stable strategies of the replicator dynamic system are discussed under different evolution routes, revealing the time-variant dynamic features of the human–machine evolutionary game. Moreover, the system dynamics model is built to explain the internal interaction behavior and mechanism of the human–machine system. The simulation results indicate that the game with different initial system states can converge to different equilibrium points. The analysis of evolutionary processes with different model parameters demonstrates that the game strategies are more sensitive to the cost of an increase in mental workload and the payoff of an increase in trust and decision accuracy. With the adoption of the proposed function allocation strategy, the mental workload coefficient decreases by 36.09%, while the trust level and the decision accuracy increase by 33.59% and 38.83%, respectively. The proposed strategy highlights the significant impact of mental workload, trust, and decision accuracy on game approaches, and explains the internal interaction behavior and mechanism between evolutionary game strategies and the dynamics of the human–machine system. This study can provide a theoretical reference and modeling approach for human–machine cooperation in armored vehicles.