Research on Dynamic Decision-making of Organizational Culture Fusion after Merger and Acquisition Based on Deep Reinforcement Learning and Agent Simulation

Abstract

Organizational culture integration is a critical link in post-merger integration and directly affects the realization of merger and acquisition synergies. However, culture integration is dynamic, complex, and uncertain, making traditional static decision-making methods insufficient for handling multi-agent interaction and sudden conflicts. This study constructs a dynamic decision-making model for post-merger organizational culture fusion by integrating deep reinforcement learning and agent simulation. First, a multi-agent simulation system is developed to represent management agents, core employee agents, and ordinary employee agents, and to model cultural cognition, behavioral preferences, and interaction rules. Second, a deep Q-network is embedded to optimize integration decisions under state variables such as cultural difference, conflict level, and employee identification. Third, 30 merger and acquisition cases from manufacturing, finance, and technology industries are used for scenario simulation and validation. Experimental results show that the proposed model increases decision response speed by 32.6%, reduces cultural conflict occurrence by 28.9%, and improves integration success rate by 25.3%, demonstrating its effectiveness for dynamic cultural integration decision-making.