基于混合神经网络的多维视觉传感信号模式分类

Multidimensional Visual Sensing Signal Pattern Classification Based on Hybrid Neural Networks

传感器采集的数字信号分类精度差,导致关键信息的丢失。为了提高传感数据的可靠性和有效性,提出基于混合神经网络的多维视觉传感信号模式分类方法。结合卷积神经网络(CNN)、循环神经网络(RNN)构建混合神经网络,以更有效地表示多维视觉数据中的特征;其中,卷积神经网络负责对多维的空间信号进行去噪处理并提取特征;循环神经网络负责对时域和频域信号进行特征提取;混合神经网络通过联合训练CNN和RNN各自的参数,以调整其权重,并且结合两者从不同层级提取的特征来实现多维视觉传感信号模式的分类。仿真结果表明,使用所提方法进行分类时,信号光滑度保持在0.9以上,传感信号分类结果与实际结果拟合度较高,有效实现多维视觉传感信号模式分类。

The poor classification accuracy of digital signals collected by sensors leads to the loss of key information.In order to improve the Reliability and effectiveness of sensing data,a multi-dimensional visual sensing signal pattern classification method based on hybrid neural networks is proposed.Combining convolutional neural network(CNN)and recurrent neural network(RNN),a hybrid neural net-work is constructed to represent the features in multidimensional visual data more effectively.Among them,convolutional neural network is responsible for denoising multi-dimensional spatial signals and extracting features,and recurrent neural network is responsible for fea-ture extraction of time-domain and frequency-domain signals.The hybrid neural network adjusts the weights of CNN and RNN by jointly training their respective parameters,and combines the features extracted from different levels to achieve multi-dimensional visual sensing signal pattern classification.The simulation results show that when using the proposed method for classification,the signal smoothness remains above 0.9,and the sensor signal classification results have a high fit with the actual results,effectively achieving multi-dimen-sional visual sensing signal pattern classification.