目的：基于中医“治未病”理论，整理有关咳嗽的中医药膳，在客观数据的基础上对咳嗽药膳用药规律进行较为深入的研究探讨，从实际应用的角度出发，建立一个咳嗽中医药膳数据库，实现药膳疗法的智能筛选，为咳嗽的临床辅助治疗提供参考。方法：采用文献学和统计学的研究方法，以1992年大连出版社出版的王者悦版本《中国药膳大辞典（修订版）》、1997年人民卫生出版社出版的项平版本《中医食疗方全录》及1992年至今网上数字资源库（中国知网、万方）中关于咳嗽的中医药膳文献为研究对象。以用药安全、有效、记录详实为基本原则制定文献的纳入标准和排除标准，对符合要求的数据手工录入至Excel文档。整理相关信息，对收集到的药膳文献进行二次筛选，然后依据药膳名称、组成、剂型等要素对文献内容进行拆分，用Microsoft Excel软件建立数据表。登录中医传承计算平台中的方剂管理系统，将整理好的数据录入该系统。运用数据管理系统和数据分析系统，对咳嗽药膳的药物用药频次、四气、五味、归经；食物的频次、归经等进行统计；然后利用关联规则算法和复杂系统熵聚类算法对数据进行分析。结果：本次研究共筛选出245个有效药膳，药物种类达147种，累计使用370次；药食同源种类有70种，累计使用1997次；食物种类58种，累计使用271次。支持度为12，置信度设为0.7时有1条常用药物组合，即“半夏和茯苓”。药物聚类核心组合为苦杏仁、百合、枇杷叶、茯苓、生姜、川贝母、白果、麦冬、陈皮、甘草等药物之间的配伍。结论：中医药膳治疗咳嗽的药物药膳组成主要以止咳平喘、润肺化痰类药物为主，聚类的核心药物组合中加入了理气健脾和清热泻火类药物，且多为甘温之品，归经为肺、胃、脾三经偏多，药膳剂型以汤剂为主。

Abstract: objective: based on the theory of traditional Chinese medicine "cure not ill", organize the cough of traditional Chinese medicine medicinal food, on the basis of objective data to cough more in-depth research on herbal medicine, from the perspective of practical application, to create a cough medicine medicinal food database, realize intelligent screening of medicinal food therapy, the clinical adjuvant therapy to provide the reference for the cough. Methods: Adopts the research methods of literature study and statistics, in 1992, dalian publishing the king of yue version "big Chinese medicated diet dictionary (revised edition)", people's medical publishing house, 1997 flat version in the diet of TCM on the whole record and since 1992 online digital repository (China hownet, ten thousand) in the TCM herbal literature as the research object of cough. The inclusion criteria and exclusion criteria were established based on the safe, effective and detailed records of drug use, and the data meeting the requirements were manually input into Excel documents. The relevant information was sorted out, the literature of medicinal diet was screened twice, and then the literature content was divided according to the name, composition, dosage form and other elements of medicinal diet, and the data table was established with Microsoft Excel software. Log in the prescription management system of the traditional Chinese Medicine inheritance computing platform and input the sorted data into the system. Using data management system and data analysis system, the frequency of drug use, four qi, five flavor and meridian of cough medicated diet were analyzed. The frequency and meridian of food were counted; Then, association rule algorithm and entropy clustering algorithm of complex system are used to analyze the data. Results: A total of 245 effective medicinal diets were screened out in this study, 147 kinds of medicines were used for 370 times. There were 70 kinds of homology of medicine and food, which were used for 1997 times. 58 kinds of food, 271 times in total. When the support degree was 12 and the confidence degree was 0.7, there was a commonly used drug combination, namely pinellia ternata and Poria cocos. The core combination of drug clustering was the compatibility of bitter almond, lily, loquat leaf, poria cocus, ginger, fritillaria fritillaria, ginkgo, Ophiopogon, tangerine peel and licorice. Conclusion: The composition of TCM medicated diet for treating cough mainly consists of relieving cough, relieving asthma, moistening lung and reducing phlegm. In the cluster of core drugs, drugs for regulating qi and strengthening spleen and clearing heat and reducing fire are added, and most of them are sweet and warm. The meridian of lung, stomach and spleen is more than three meridian, and the dosage form of medicated diet is mainly decoction.