基于多参数动态时序的技术预测方法与实证研究

Technology Forecasting Method and Empirical Research Based on Multi-parameter Dynamic Time Series

为能够更加精细地识别技术发展变化趋势,借鉴统计学中数据分析指标,在对技术研究有关文献与该技术发展变化关系进行分析的基础上,通过技术主题提取、模型构建与技术预测3个步骤,引入平均值、波动区间、波幅时长3个模型参数,提出利用技术主题词频差值反映一定时间区间内技术发展变化特征的多参数动态时序技术预测模型,并将该模型方法应用于挥发性有机物处理技术领域的技术预测,提取得到8个技术主题,识别出热分解技术是其中的重点技术,结合该领域研究和发展实际,证明该模型方法可行、有效。

In order to be able to identify the trends of technological development and change more finely, drawing on the data analysis indicators in statistics, on the basis of analyzing the relationship between the relevant literature of technical research and the development and change of the technology, through three steps of technology topics extraction, model construction and technology forecasting, three model parameters of average value, fluctuation range and volatility time are introduced, and a multi-parameter dynamic time series technology forecasting model is proposed to reflect the characteristics of technological development and changes within a certain time interval by using the words frequency difference of the technology topics. Furthermore, the model method is applied to the technology forecasting in the field of volatile organic compound treatment technology, eight technology topics are extracted, and the thermal decomposition technology is identified as the key technology, combined with the research and development practice in this field, it is proved that the proposed model method is feasible and effective.