Crane scheduling tasks in steelworks are a matter of uncertainty scheduling with certain probability distribution pattern. To better schedule tasks with this feature, this paper proposes a Dynamic Area Allocation (DAA)-based crane scheduling model according to the following steps. First, Bayesian network, according to the time sequence of crane transportation tasks, is constructed. Then, conditional probability for each network node on the basis of actual crane operating data is calculated for getting the corresponding time–space probability distribution, and then obtaining the spatial distribution by superposing of all crane transportation tasks in the space domain at certain time. At last, tasks are assigned to cranes based on their spatial distribution and the equal probability partition. Simulation testing on the scheduling model is carried out using practical crane transportation tasks in steelworks. Results show that the model based on the dynamic area allocation, with its scheduling period of 15 min, can greatly shorten transportation time and reduce times of collision resulted from crane interference, after compared with the current widely used crane scheduling programme based on the fixed area allocation (FAA).

钢厂起重机调度任务是一个具有一定概率分布模式的不确定性调度问题。为了更好地调度具有此功能的任务，本文根据以下步骤提出了一种基于动态区域分配（DAA）的起重机调度模型。首先，根据起重机运输任务的时间顺序，构建贝叶斯网络。然后，根据起重机实际运行数据，计算每个网络节点的条件概率，得到相应的时空概率分布，然后在空间域内叠加所有起重机运输任务，得到空间分布。最后，根据起重机的空间分布和等概率划分，将任务分配给起重机。以钢厂实际起重机运输任务为例，对调度模型进行仿真测试。结果表明，与目前广泛使用的基于固定区域的起重机调度程序相比，基于动态区域分配的模型调度周期为15 min，可以大大缩短运输时间，减少起重机干扰造成的碰撞次数。分配（美国联邦航空局）。