دانلود رایگان مقاله مکانیسم تقسیم کار از دیدگاه بهینه سازی فردی

Abstract

Although the tools of complexity research have been applied to the phenomenon of labor division, its underlying mechanisms are still unclear. Researchers have used evolutionary models to study labor division in terms of global optimization, but focusing on individual optimization is a more realistic, real-world approach. We do this by first developing a multi-agent model that takes into account information-sharing and learning-by-doing and by using simulations to demonstrate the emergence of labor division. We then use a master equation method and find that the computational results are consistent with the results of the simulation. Finally we find that the core underlying mechanisms that cause labor division are learning-by-doing, information cost, and random fluctuation.

4. Conclusion

We have modeled the division of labor as a self-organization process. Driven by individual income optimization, the system moves from disorder to order and a division of labor emerges. We do not set system-wide goals in the simulation process but apply a set of rules governing how agents interact with each other and with their environment. Agents seeking to increase their income is the driving factor in this self-organized evolution.

The rational for labor division is as follows.

(1) Information sharing: Resource information is shared in a closed environment. This is a good approximation of the real-world information exchange process and makes the transaction rules between agents rational and realistic.

(2) Learn-by-doing: This nonlinear mechanism enables agents to further improve what they are good at and—through the effect of positive feedback— become more strongly attached to doing it. Overall efficiency increases, which is an important factor in the emergence of labor division.

(3) Random fluctuation: Stochastic fluctuation is a necessary condition for self-organized evolution and determines the state the agent arrives at in the final stage. In the process of evolution, a small change in the initial state of the system can strongly affect the final state. In our study the emergence of labor division does not depend on stochastic fluctuations, but the final outcome of labor division does depend on them.