Routing and scheduling field service operation by P-graph
针对现场服务中多任务、多目标、多约束的复杂问题，提出两阶段优化模型：上层按能力分配任务到时间区间，下层用时间约束过程网络合成精确规划路径与调度，迭代优化直至方案无法改进。
Field service operation is a theoretically complex problem. Numerous tasks at various locations need to be assigned to service staff. Skills, capabilities, and the level of local knowledge have to be considered to estimate the time required to complete a task by a group of selected technicians. Multiple objectives involve cost, task priorities, and consumer satisfaction. The cost includes the salary of service staff for a whole route from departing to returning to the base location. The travel cost depends on the proper routing of the service staff and the effective collection of closely located tasks into a single route. Priorities of tasks are defined by a function of their deadlines. Consumer satisfaction is a function of the number of missed appointments. A two-level optimization model is introduced to solve large-scale field service operation problems. At the higher level, tasks are assigned to time intervals based on the capacity and availability of the service groups. Then at the lower level, assigned tasks are exactly scheduled and service routes planned by time-constrained process network synthesis (TCPNS). The execution of the different levels is performed iteratively until the overall work plan cannot be further improved.