摘要
大型复杂机电设备或特种高危贵重物品仓储、运输过程常需要对设备状态及周围环境信息进行实时监控。而在监控过程中经常遇到多终端复杂任务协调难题,由于多终端任务调度是一个NP(non-deterministic polynomial)优化问题,高效的调度策略可极大缩短调度等待时间。基于此,文章提出一种融合多种策略的灰狼调度方法,建立新的编码方式与离散空间映射,应用混沌系统不确定性初始化种群,以自适应惯性权重来平衡调度方法的全局优化性能与收敛速度,引入反向学习思想扩大算法寻优范围。实验结果表明调度方法将系统等待时间有效缩减10.06%,其终端负载及最小等待时间明显优于其他算法,验证了调度策略的有效性及优越性。
Large-scale complex electromechanical equipment or special high-risk valuables warehousing and transportation are often monitored in real time on equipment status and surrounding environmental information.In the monitoring process,multi-terminal complex task coordination problems are often encountered.Because the multi-terminal task scheduling is a NP(non-deterministic polynomial) optimization problem,the efficient scheduling strategy can greatly shorten the scheduling waiting time.Based on this,this article proposes a grey wolf scheduling method that integrates a variety of strategies,establishes a new encoding method and discrete space mapping,and uses the uncertainties of the chaos system to initialize the type.With the speed of convergence,the scope of the extension of reverse learning ideas is introduced.The experimental results show that the scheduling method effectively reduces the system waiting time by 10.06%.Its terminal load and minimum waiting time are significantly better than other algorithms,which verify the effectiveness and superiority of the scheduling strategy.
作者
许富景
杜少成
张燕
刘强
XU Fujing;DU Shaocheng;ZHANG Yan;LIU Qiang(School of Automation and Software Engineering,Shanxi University,Taiyuan 030006,China)
出处
《中国测试》
CAS
北大核心
2024年第10期73-80,共8页
China Measurement & Test
基金
省部共建动态测试技术国家重点实验室基金(2022-SYSJJ-02)。
关键词
数据终端
复杂任务
灰狼算法
调度策略
data terminals
complex tasks
grey wolf algorithm
scheduling strategies