| Robotic process automation software can help enterprise employees to do repetitive work,which can not only improve business efficiency,but also reduce the possibility of manual operation errors.When the robot process automation software is used to perform the function of automatic login system,it needs to identify the verification code,but the software cannot automatically identify the verification code,which can not achieve the purpose of automatic login system.Therefore,the method of identifying the verification code based on quantum convolutional neural network is studied.This paper studies the construction of quantum convolution neural network model,the quantum convolution layer and the classical full connection layer of hybrid quantum convolution neural network optimized by particle swarm optimization algorithm,and the specific application of this model in verification code recognition.Based on the research of classical convolutional neural network model and quantum computing,this paper designs a hybrid quantum convolutional neural network model.The input of this model needs to convert classical data into quantum state coding,use quantum convolutional layer,input the observed results into the classical pooling layer,and finally use the classical classification method to get the output.The experimental results show that the accuracy of the hybrid quantum convolution neural network model in training set and test set is improved.Combining the particle swarm optimization algorithm with the hybrid quantum convolution neural network model,the angle value of the single bit rotary gate that constitutes the quantum convolution layer and the parameters of the full connection layer are optimized,and a PSO-HQCNN model for verification code recognition is obtained.The verification code recognition model is the key point for RPA software to realize the automatic login system.The verification code data set is obtained by using the RPA software writing process.After the data set is segmented and compressed,it is sent to the hybrid quantum convolution neural network model for training to obtain the verification code recognition model.Use Python related activities in RPA software to embed the verification code identification model into the process of automatic login system.After repeated login tests,the success rate of login system is more than 90%,which can meet the needs of automatic login system. |
