Abstract: The tugboat scheduling is a typical multiprocessor tasks scheduling problem. A kind of port tugboat scheduling problem with multi-anchorage bases and different operation modes is presented, and the tugboat scheduling optimization model which objective is to minimize both the maximum completion time and total operation fuel wastage of all tugboats is described. The evolutionary strategy algorithm is applied to optimize the tugboat operation scheduling, and the multi-dimension real encoding and decoding based on the roulette probability assignment is introduced. Finally, the calculation results prove that the evolutionary strategy algorithm can effectively optimize the port tugboat scheduling problem, and the optimization results of the maximum completion time are improved and reduced about 16% than the simulation results on condition of different tugboat operation modes. Moreover, the calculation results show that different operation modes have much effect on the tugboat scheduling.

Key words: tugboat scheduling, multi-anchorage bases, multiprocessor tasks, evolutionary strategy algorithm, multi-objectives optimization

摘要： 拖轮调度是典型的多处理器任务调度问题，针对多停泊基地和不同作业模式下的拖轮调度，以最大完工时间和总作业油耗最小化为目标，建立了拖轮调度多目标优化模型。采用演化策略算法对多停泊基地拖轮调度优化问题进行计算，提出一种基于轮盘赌概率分配的编码和解码方法。计算结果表明了演化策略算法的有效性和可行性，优化后的最大完工时间最小值相比仿真运行结果取得较大的改善，不同作业模式下均减少了约16%；计算结果还表明不同作业模式对拖轮调度结果会产生较大影响。

关键词: 拖轮调度, 多停泊基地, 多处理器任务, 演化策略算法, 多目标优化