蜣螂优化算法下"互联网+营销服务"虚拟机器人应用模型
Application model of"Internet+Marketing service"virtual robot under dung beetle optimizer algorithm
何玮 1周雨湉 2俞阳 1康雨萌 1朱萌 1钱旭盛1
作者信息
- 1. 国网江苏营销服务中心,江苏 南京 210000
- 2. 伦敦大学 国王学院,伦敦 WC2R 2LS
- 折叠
摘要
为了应对新形势下的电力营销服务形势,提升互联网时代的电网优质服务水平,利用蜣螂优化(dung beetle optimizer,DBO)算法,设计了一种"互联网+营销服务"虚拟机器人模型.首先针对电网营销部门可能发生的人机交互情景开展交互分析与关系框架设计,然后基于深度Q网络(deep Q network,DQN)建立虚拟机器人自主学习模型,引入DBO算法完成模型超参数的高效寻优并通过训练完成优化后的模型学习,最终将实际的电力营销数据输入到模型中进行实验测试.在特定的测试环境下综合检测模型的实际应用情况,测试结果表明:该虚拟机器人模型在功能性实验、非功能性实验和安全性实验中模型运转和系统运转正常率达到100%,能够较好地实现人机交互功能,达到全天候客户需求精准响应的战略目标.
Abstract
In order to deal with the situation of power marketing service under the new situation and improve the service of power grid in the internet era,a virtual robot model of"Internet+ Marketing service"was designed by the dung beetle optimizer(DBO)algorithm.Firstly,interac-tion analysis and relationship framework design were carried out for possible human-machine in-teraction scenarios in the power grid marketing department.Then,a virtual robot autonomous learning model was established based on deep Q network(DQN),and the DBO algorithm was in-troduced to efficiently optimize the model hyperparameters.The optimized model learning was completed through training.Finally,actual power marketing data was input into the model for experimental testing.The actual application of the virtual robot model was comprehensively test-ed in a specific testing environment.The test results show that the normal operation rate of the model and system in functional experiments,non functional experiments,and safety experiments reach 100%,which can achieve human-machine interaction function and the strategic goal of ac-curate response to customer needs.
关键词
深度Q网络/虚拟机器人/蜣螂优化(DBO)算法/超参数寻优/电力营销服务Key words
deep Q network/virtual robot/dung beetle optimizer algorithm/hyperparametric op-timization/power marketing service引用本文复制引用
基金项目
国网江苏电力研发实施类信息化项目(SGJSYF00YHX)
出版年
2024
国家科技期刊平台
NSTL
万方数据