学前教育中基于CPG与DDPG结合的机器人运动控制与幼儿互动研究  

Robot motion control and children’s interaction in preschool education based on CPG and DDPG

作　　者：吴文青 WU Wenqing(Xianyang Polytechnic Institute,Xianyang Shaanxi 712000,China)

机构地区：[1]咸阳职业技术学院,陕西咸阳712000

出　　处：《自动化与仪器仪表》2024年第9期289-293,共5页Automation & Instrumentation

基　　金：陕西省职业技术教育学会课题《〈幼儿园管理〉课程思政研究与实践》(SGKCSZ2020-1471);咸阳职业技术学院2024年度教育教学改革研究项目《产教融合背景下高职院校学生通用性职业能力的研究》(2024JYC09)。

摘　　要：学前教育中的机器人教学可以帮助幼儿培养动手能力、逻辑思维能力等。研究设计了一个交互式的智能机器人,通过中枢模式生成器模块生成机器人的运动模式,并通过深度确定性策略梯度算法不断优化机器人的动作策略。研究结果表明,在0~1 000回合内,滑动平均奖励值呈稳定上升趋势,在1 000回合之后,奖励值稳定在750附近。基于中枢模式生成器与深度确定性策略梯度结合的机器人运动控制方法可以有效提高机器人与幼儿的互动体验,有助于促进学前教育中的幼儿运动发展。结合中枢模式生成器和深度确定性策略梯度的机器人能够实现更加自然和有效的运动控制,同时提高机器人与幼儿互动的质量和效果。研究为未来的幼儿教育和机器人研究提供了新的思路和方法。Robot teaching in preschool education can help children develop hands-on ability and logical thinking ability.In this paper,an interactive intelligent robot is designed,which generates the robot’s motion pattern through the central pattern generator module,and continuously optimizes the robot's action strategy through the deep deterministic strategy gradient algorithm.The results show that the sliding average reward value increases steadily during 0-1000 rounds,and stabilizes around 750 after 1000 rounds.The robot motion control method based on the combination of central pattern generator and deep deterministic strategy gradient can effectively improve the interactive experience between robots and children,and help promote the motor development of children in preschool education.Robots that combine a central pattern generator with a depth deterministic strategy gradient can achieve more natural and effective motion control while improving the quality and effectiveness of robot interactions with young children.The research provides new ideas and methods for future early childhood education and robot research.

关 键 词：学前教育 机器人运动控制 中枢模式生成器 深度确定性策略梯度 幼儿互动 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]