A study on incentive mechanism of the large-scale construction performance audit based on the computational experiment

The characteristics of large-scale constructions, such as multi-agent, multi-stage and multi-target have brought great difficulties to improve the construction´s performance. This paper builds a computational experiment model of the incentive mechanism, and simulates the behavioral strategies of all participants by using computational experiment artificial system. The experiment results show that: the incentive mechanism of performance audit can optimize the project´s performance to a certain extend; in the single-period audit, the subcontractors are more sensitive to the incentive level, and show a strong risk aversion characteristic; while in the multi-period audit, the incentive level lose effectiveness on improving the constructions´ performance, it´s more critical to take the individual´s equity preference into consideration.