Construction Schedule-cost Optimization of High-rise Buildings Based on Reinforcement Learning
In order to solve the problem of intelligent optimization and control of the schedule-cost target of high-rise building construction,this paper proposed a comprehensive optimization algorithm framework method for the schedule-cost of high-rise building construction based on reinforcement learning algorithm and considering the uncertainty of process duration.This method integrated the PERT principle to introduce uncertainty,and developed a schedule-cost optimization model based on the Double DQN algorithm architecture,which has been verified and applied in a class-A public high-rise building commercial complex project.After model optimization,the total construction period and cost optimization of the project have been significantly improved,achieving cost reduction and efficiency improvement of the project,which is conductive to scientific planning,decision-making and risk management of the construction process.The optimization algorithm framework developed by this study based on reinforcement learning overcomes the limitations of traditional methods,enhancing the intelligent level of high-rise building construction control,and empowers scientific construction management.
NETL
NSTL
万方数据
