English：

Large island infrastructure construction often involves complex stonework excavation,transportation and backfilling operations,where dispatching efficiency directly impacts project costs andschedules. Focusing on the stone supply and demand dispatching for a port project, this study addressesa problem characterized by multiple excavation zones, multiple backfill zones, diverse material types,multimodal transport, and multiple storage nodes, alongside long-cycle, dynamic monthly demands. Toaddress this stonework dispatching problem, a multi-objective optimization model is established, aimingto minimize total land transportation distance, monthly excavation volume fluctuations, and stockpileinventory levels. Given the NP-hard complexity of the problem, a customized genetic algorithm isdesigned and implemented. The core of the algorithm lies in an innovative indirect chromosome encodingscheme that encodes monthly relative excavation intensities rather than specific transportation volumes,thereby enhancing search efficiency. A fitness function evaluates the performance of each strategy througha full project cycle simulator, utilizing weight coefficients to manage complex constraints. The engineeringpractice shows that the algorithm converges effectively, generating a stonework dispatching scheme thatbalances economic efficiency, operational stability and timeliness.