Research On Quantum Neural Network Model Of Structure And Algorithm

Quantum computing, which is the combination of information science and quantum mechanics, is an interdiscipline full of vitality. Quantum computing represented by quantum neural network owing to its high degree of parallelism, the exponential storage capacity as well as the heuristic acceleration for classical algorithms, has a great advantage and bears a strong vitality, which has become the cutting-edge research field of many international scholars. The introduction of the quantum computing mechanism in the traditional neural network is in order to improve the nonlinear approximation capability of neural networks, convergence, stability and other properties of the algorithm. Therefore, the research on quantum neural network has important theoretical and practical significance.This paper reviews the research status and development trend of the quantum neural networks, summarizes an overview of the basic theoretical knowledge in quantum mechanics, discusses the current design concept, neural structure, network form and learning algorithm of several kinds of quantum neural networks and analyzes the network computational advantages and limitations. In view of the defect that the traditional neural network perceptron can not solve the XOR problem, this paper proposes a quantum perceptron neuron model which can solve this defect. At the same time, a quantum neural network model is established with traditional neural network framework so that this network model is proved theoretically which has a good astringency, and the corresponding learning algorithm used to calculate the network parameters is given out. Through the simulation experiments compared with RBF neural network, the effectiveness and superiority of the network model and the learning algorithm are proved in aspect of the approximation capability for the nonlinear function.This paper analyzes the deficiencies of the quantum neural network model, proposes a kind of quantum neural network models based on CNOT gate, and analyzes the characteristics of the imitation of CNOT gate. Then, a kind of quantum neural network models based on CNOT gate is established with traditional neural network framework so that this network model is proved theoretically which has a good continuity, and the corresponding learning algorithm used to calculate the network parameters is given out. Through the simulation experiments compared with the quantum neural network model, the effectiveness and superiority of the network model and the learning algorithm are proved in aspect of the approximation capability for the nonlinear function.