Abstract: Earthwork allocation is crucial to hydropower project construction, posing a direct effect on its cost and efficiency. To optimize allocation strategies and machinery utilization efficiency, this study develops a new optimization method for improving earthwork allocation and road transportation systems based on the coupling of the primal-dual interior point method and AnyLogic. First, a dynamic earthwork allocation model is formulated to minimize allocation costs while considering constraints such as excavation and filling schedules and site planning. Then, we develop an AnyLogic-based road transportation system simulation model using queuing theory, focusing on the complicated road transportation conditions in hydropower projects. And the interior point method is adopted to solve the dynamic earthwork allocation model for an optimal allocation scheme. Finally, we couple these two models and apply a multi-agent simulation technique to simulate the dynamic interaction of construction machinery in transit, so that we can achieve visualization of the real-time data and feedback. This facilitates dynamic adjustment for machinery allocation, thereby improving machinery utilization efficiency. Application in a practical engineering case shows this method is effective in lowering earthwork allocation cost, achieving a project cost reduction by 7.3%. It also increases machinery coordination efficiency, achieving an increase of 84.0% and 86.7% in the comprehensive embankment placement rates of the supply and receiving zones respectively. By adopting the optimal allocation scheme, the machinery utilization rate exceeds 60.0% of most of the backhoe excavators, and 95.0% of the dump trucks. These studies are useful for decision-making in earthwork allocation planning.

Key words: earthwork allocation, primal-dual interior point method, AnyLogic, road transportation system, simulation model