中文摘要：

根据肠道蠕动机制提出一种应用于反馈式人工肛门括约肌的直肠感知功能预测模型。该模型通过小波包分析，将结肠收缩压力信号的能量分层作为特征向量，采用支持向量机进行模式识别。仿真实验中，首先直接将收缩压力信号幅值作为特征向量，分别采用BP网络、支持向量机进行模式识别，随后将所提出的预测模型与两种方法进行比较。仿真结果说明，基于小波包分析与支持向量机相结合的预测模型具有更快的训练速度和更好的识别效果，具有良好的应用前景。

英文摘要：

According to mechanism of colon wri proposed, which is applied in feedback artificial ggling, a prediction model of rectum＇ s perceptive function is anal sphincter. Layered energy of colon systolic pressure sigrials is obtained as feature vectors through wavelet packet analysis. Support vector machine （SVM） is adopted for pattern recognition. During emulation test, amplitude of systolic pressure signals is adopted as feature vectors, pattern recognition of BP neural network and SVM are studied and compared with the prediction model proposed in the paper. Simulation results indicate that the prediction model based on wavelet packet analysis combined with SVM has faster training speed and better effective recognition.