刺吸电位技术（Electrical penetration graph，EPG）是一种用来记录植食性刺吸式昆虫在寄主植物上取食行为的电生理技术。通过对EPG波形的分析，可以识别昆虫在植物不同组织中的取食行为。然而，面对EPG波形数据中的噪声、不同类别波形之间的微小差异以及波形发生的确切时间，研究者需要耗费大量时间和精力进行波形判读和标记。随着人工智能技术的发展，尤其是机器学习和深度学习的引入，EPG波形的高效识别与准确分析的自动化和智能化逐步变为现实。人工智能技术能够快速从大量复杂的EPG波形中提取出有用信息并识别出昆虫的具体取食行为，为EPG技术的智能化发展提供了技术支撑。本文详述了到目前为止EPG波形自动识别和参数统计分析的研究进展，展望了未来EPG技术与人工智能结合的应用前景，期望为EPG技术的智能化发展提供一些借鉴。

Electrical penetration graph (EPG) technology is an electrophysiological technique used to record the probing and feeding behaviors of herbivorous, piercing-sucking insects. By analyzing EPG waveforms, it is possible to identify different feeding behaviors of insects within different plant tissues. However, because of noise in the EPG waveform data, subtle differences between different types of waveforms, and in the exact time when the waveforms occur, researchers need to spend a lot of time and effort to interpret waveform data. With the development of artificial intelligence technology, particularly the introduction of machine learning and deep learning, automation of recognition and accurate analysis of EPG waveforms has gradually become a reality. Artificial intelligence technology can quickly extract useful information from many complex EPG waveforms and identify the species-specific feeding behaviors of insects, thereby providing technical support for the intelligent development of EPG technology. This article elaborates on progress in research on the automatic recognition of EPG waveforms and associated statistical analysis, and outlines prospects for combining EPG technology with artificial intelligence and the intelligent development of EPG technology.