摘要: Quantum Neural Network (QNN) is a young and outlying science built upon the combination of classical neural network quantum computing. Making use linear superposition, this paper presents M-P based on analysis conventional network. Moreover, working principle proposed its corresponding weight updating algorithm are expatiated in two cases input state being orthogonal non-orthogonal basic set, respectively. In addition, not only validates that can realize some functions, such as “XOR”, but also verifies feasibility validity learning by simple examples.