情绪是人对外界或自我刺激产生的心理和生理反应。如果人机交互(Human Computer Interaction, HCI)系统可以识别情绪，那么心理疾病的诊断和心理学的研究将会更加客观和有效。本文提出了一种仅基于光电容积脉搏波(photoplethysmography, PPG)的情绪识别方法对情感进行分类。由多波长近红外透射光谱法测量获得的脉搏波，经过特征点检测，获得其信号特征，使用不同的机器学习算法验证由PPG信号识别情绪的性能。结果表明，使用脉搏波进行情绪四分类，其识别正确率为96.2%，且单个样本的测试时间短。这意味着基于PPG信号可以实现快速的多类情绪识别，对无创检测、可穿戴设备和临床实践具有潜在的价值。 Emotion is person’s psychological reaction to external or self-stimulation with the physiological re-action. If Human Computer Interaction (HCI) system can be used to recognize emotion, the diagnosis of mental illness and psychological research will be more objective and effective. In this paper, an emotion recognition method based on photoplethysmography (PPG) is proposed to classify emo-tions. The signals measured by multi-wavelength near-infrared transmission spectroscopy are de-tected by feature points, and the signal characteristics are obtained. Different machine learning al-gorithms are used to verify the performance of emotion recognition by PPG signals. The results show that when PPG is used for emotion classification, the accuracy of recognition is 96.2%, and the testing time of a single sample is short. This means that fast multi-class emotion recognition can be achieved based on PPG signals, which has potential value for non-invasive detection, wearable de-vices and clinical practice.