基于Delphi法和BP神经网络的技术预见模型研究

构建了基于Delphi法和BP神经网络的技术预见模型。基于专利分析提取高价值专利以期提供科学、客观的参考点；处理专家评价数据时，为充分考虑专家熟悉程度的影响引入性能指数概念，并运用熵权TOPSIS法评估专利技术综合得分；结合BP神经网络算法计算各技术评估得分，弱化权重计算中人为因素的影响，对比计算结果验证其可行性。以智能制造领域为例开展实证研究，对该领域重点专利技术进行评估，验证了模型的科学性和可行性。

Research on Technology Foresight Model Based on Delphi Method and BP Neural Network

The paper construct the foresight model based on Delphi method and BP neural network. Based on patent analysis, high value patents are extracted so as to provide scientific and objective reference points; When the expert evaluation data is processed, the performance index is introduced so that the influence of expert familiarity is taken into full consideration, and the entropy weighted -TOPSIS method is used to evaluate the comprehensive score of patent; BP neural network algorithm is used to evaluate the score, and the influence of human in the weight calculation is weakened, and the feasibility is verified by comparison. Taking intelligent manufacturing field as an example, empirical studies are conducted and the model is validated.