一种用于工作流引擎的任务预测与分配算法  

AN ALGORITHM FOR TASK PREDICTION AND ASSIGNMENT IN WORKFLOW ENGINE

作　　者：钱鹰[1] 王寸涛 韦庆杰[1] 

机构地区：[1]重庆邮电大学图形图像与多媒体实验室,重庆400065

出　　处：《计算机应用与软件》2014年第8期75-78,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(61171060);教育部留学回国人才科研启动基金项目(教外司留(2009)1509号)

摘　　要：为了检测工作流系统在中短期内是否会出现任务超载的情况,首先利用灰色预测算法来预测不同类型任务在预测窗口内的变化趋势,并根据预测结果建立一种负载均衡偏差可控的整数线性规划模型,以求解最优的任务分配策略。当工作流系统在预测窗口内出现任务超载时,该方法可计算出在保证不同工作人员之间任务负载均衡和任务分配合理性的情况下所需增加的最少工作人员数量,并且在一定程度上提高了任务处理的时效性。In order to detect whether the tasks in the workflow system will overload in short-medium term,we first make use of the grey prediction algorithm to predict the variation trend of different types of tasks within the prediction window,and establish an integer linear programming model supporting the controllable load-balancing deviation according to the prediction results so as to solve the problem of optimal task allocation strategy. When the tasks become overloaded for the staff in the prediction window,this method can calculate the least increased number of staff in the workflow system. At the same time,it can ensure the load balance among different staff and the rationality of task allocation. And to a certain extent,it can also improve the timeliness of task processing.

关 键 词：工作流 任务分配 负载均衡 负载预测 

分 类 号：TP317.1[自动化与计算机技术—计算机软件与理论]