摘要
针对当前早教机器人的数字图像识别分类方法精度较低,导致出现早教机器人的机械臂数字书写的准确率不够理想的缺陷,研究提出一种智能模型对其进行改进。首先,基于一种鲸鱼优化算法(Whale optimization algorithm,WOA)改进的脉冲耦合神经网络(Pulse coupled neural network,PCNN)实现数字图像预处理;然后结合WOA和海鸥优化算法(Seagull Optimization Algorithm,SOA),以对SOA的性能进行改进;最后,利用改进的SOA对BP神经网络(BP neural network,BPNN)进行改进。结合上述内容,构建基于改进BPNN的数字图像识别分类模型,提升早教机器人的机械臂书写数字的准确率。结果显示,在应用该模型后,早教机器人的数字书写准确率为98.61%,比其他模型高7.53%~10.18%。因此,研究提出的方法能够有效提升早教机器人的机械臂书写数字的准确率,为儿童早教提供新的路径。
In view of the low accuracy of the current digital image recognition and classification methods for early education robots,which leads to the unsatisfactory accuracy of the robot arm digital writing of early education robots,an intelligent model is proposed to improve it.Firstly,digital image preprocessing is implemented based on a Whale optimization algorithm(WOA)improved Pulse Coupled Neural Network(PCNN);Then combine WOA and Seagull Optimization Algorithm(SOA)to improve the performance of SOA;Finally,the improved SOA is used to improve the BP neural network(BPNN).Combining the above content,a digital image recognition and classification model based on improved BPNN is constructed to improve the accuracy of the robot's robotic arm in writing digits.The results show that after applying this model,the digital writing accuracy of the early education robot is 98.61%,which is 7.53%~10.18%higher than other models.Therefore,the method proposed in the study can effectively improve the accuracy of the robotic arm for early childhood education to write numbers,providing a new path for children’s early childhood education.
作者
党晓梅
DANG Xiaomei(Xianyang Polytechnic Institute,Xianyang Shaanxi,712000,China)
出处
《自动化与仪器仪表》
2024年第8期261-265,共5页
Automation & Instrumentation
基金
陕西省职业技术教育学会课题《〈早期教育概论〉课程思政研究与实践》(SGKCSZ2020-1497)。
关键词
早教机器人
机械臂
BPNN
数字识别
海鸥优化算法
early education robot
mechanical arm
BPNN
digital identification
seagull optimization algorithm
