足球比分直播

基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf

返回
基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf_第1页
第1页 / 共73页
基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf_第2页
第2页 / 共73页
基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf_第3页
第3页 / 共73页
基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf_第4页
第4页 / 共73页
基于机器嗅觉的中华绒螯蟹新鲜度检测技术分析.pdf_第5页
第5页 / 共73页
点击查看更多>>
资源描述:
论文审阅认定书 研究生 在规定的学习年限内,按照研究生培养方案的要求,完成了研究生课程的学习,成绩合格;在我的指导下完成本学位论文 ,经审阅,论文中的观点、数据、表述和结构为我所认同,论文撰写格式符合学校的相关规定,同意将本论文作为学位申请论文送专家评审。 导师签字 年 月 日 万方数据致谢 首先要衷心 感谢 我的导师 徐本连教授 在整个论文过程中的悉心指导 和思想启迪, 徐老师 无论是科学研究还是生活作风都将是我 一生学习的榜样。在每次课题组汇报工作之时, 徐老师 都会就我课题中出现的问题予以详尽的说明和点拨,然后给我开辟新的思想方向,不断鼓励我继续努力、精益求精,这其中都使得我自身获得深入的思考,加快了我完成论文的进程。 在论文完成之际,希望借此机会在这里再次向 徐老师 表示衷心的感谢 此外,还要特别感谢 朱培逸老师 ,可以说这篇论文的顺利完成是与 朱老师 的指导帮助是分不开的,无论在选题、开题的各个阶段, 朱老师 都给我提出了诸多宝贵的意见,并且定期地与我交流论文进展的情况,使我的论文逐步的完善。 朱老师 勤恳负责的态度日趋的感染我奋发 向上 ,同时也要感谢 朱老师 在日常生活中给予的关心和支持,最后真诚地说声谢谢 同时还要感谢诸多一起奋斗的研究生同学,大家互相帮助和支持,还有一群本科学弟学妹们给我的鼓励和欢乐。 最后要感谢的是我的父母, 他们不仅给我物质生活上的保障,也是我学习上前行的精神动力,你们给予的理解和支持使得我顺利的完成硕士论文,在今后的日子里,我会继续努力地学习和工作,不辜负你们的殷切希望 万方数据I 摘 要 水产品新鲜度检测是当前食品质量安全问题的一个重要话题 , 由于中华绒螯蟹 在 濒临 死亡或 死亡的状态下挥发出有毒的物质,而人体一旦食用将会造成严重的威胁, 在当前市场贸易中, 对中华绒螯蟹 新鲜度的 传统 检测为 感官检测,这往往受到主观性的影响,在大量中华绒螯蟹处于捆绑的情况下,最终所得到的结果是有偏的, 因此研究无损、实时、准确的中华绒螯蟹新鲜度检测技术是当前水产品市场发展的迫切需求,特别是基于机器嗅觉的中华绒螯蟹新鲜度检测技术研究具有更重要的现实意义。 理 化指标 ( 挥发性盐基氮 ) 可以实现中华绒螯蟹的 新鲜度检测, 但是需要以蟹肉为实验对象,整个实验包含高温蒸馏等诸多复杂环节,这无法满足市场快速交易的 目的,伴随着传感器技术和 模式识别 技术的 发展,机器嗅觉系统来实现水产品的完整活体检测将具有广泛的应用前景,本课题对基于机器嗅觉的中华绒螯蟹新鲜度检测技术进行了研究,主要成果有 结合完整活体的检测目的搭建了 机器嗅觉系统, 考虑中华绒螯蟹的外形、体积因素,采用静态顶空法设计气味采样平台, 针 对活体状态下所挥发出的特征性气体,研究机器嗅觉传感器阵列的选型,为了后续的 信息采集和 处理, 编写了上位机界面,实现软件系统的各个模块 功能。 鉴于当前尚未有 以中华绒螯蟹 完整活体的状态进行气味信息采集的研究 , 需要 在试验中对诸如 顶空温度、顶空体积、顶空时间、洗气时间、采样时间等参数设置 进行分析研究 , 确定合理的实验方案。 然后对气味信息依次进行 均值滤波、基线处理和异常数据的剔除等操作,选择 稳态信息和瞬时信息两方面有效特征。 由于所使用的传感器为非线性的金属半导体材质,那么所获得的气味信息也必定包括非线性的特性,传统的线性降维算法无法提取其中的 非线性特征,本课题提出基于拉普拉斯特征映射算法用于气味信息的特征提取, 最终的可视化分类结果相对于传统算法有 很大的区分度。 以 挥发性盐基氮作为新鲜度评价的标准 , 用多元线性回归模型建立了气味响应与挥发性盐基氮含量的 相关性 , 从而 反映用气味信息来评价新鲜度是切实可行的 ,本课题 基于 BP 神经网络对 中华绒螯蟹 新鲜度建立预测模型,基本上可以认为该模型对于评价 中华绒螯蟹 新鲜度是有效的。 该论文有图 33 幅,表 8 个,参考文献 73 篇。 关键词 机器嗅觉;中华绒螯蟹; 拉普拉斯特征映射 ;预测模型 万方数据II Abstract Aquatic products freshness detection is an important topic in the current issue of food quality and safety. As Chinese mitten crabs volatilize poisonous substances in the state of dying or death, it will pose a serious threat once human consumption. In the current market trade, traditional freshness detection of crabs is sensory detection which is often affected by the subjectivity. In the case of a large number of Chinese mitten crabs are bundled, the final results obtained are biased. So research on destructive, real-time and accurate Chinese mitten crab freshness detection technology is an urgent demand of the current aquatic products market development. In particular, research on freshness detection approach for Chinese mitten crabs based on machine olfaction has more important practical significance. Physical and chemical indicator TVB-N can detect the freshness of Chinese mitten crabs, but the crab meat to be used as the experiment subject. The entire experiment contains high-temperature distillation and many other complex aspects which can not meet the purpose of rapid market transaction. With the development of sensor technology and pattern recognition technology, the complete living detection of aquatic products achieved by machine olfactory system will have broad application prospects. The subject of Chinese mitten crabs freshness detection techniques based on machine olfaction has been studied, the main results are The machine olfactory system was built combined with the purpose of complete living detection. Considering the shape and volume of Chinese mitten crabs, we designed the odor sampling plat by static headspace .For the characteristics gas were volatiled in living state, we research on the selection of machine olfactory sensor arrays. For subsequent ination collection and processing, the PC interface which can realize the function of each module of the software system were made. In current, there is no research on odor ination collection in state of complete living body. In the experiment, such as headspace temperature, headspace volume, headspace time, scrubbing time, sampling time and other parameters were need to be analysised and studied for determining a reasonable experiment solution. Then, we sequentially pered mean filtering, baseline processing, removing abnormal data and other operations for smell ination. Steady-state ination and transient ination were chosen for effective feature. Since the materials of sensors are non-linear metal-semiconductor, the obtained odor ination must also include non-linear characteristic. Traditional linear dimension 万方数据III reduction algorithm can not extract the nonlinear characteristics, the subject proposed laplacian eigenmaps algorithms for extracting odor ination characteristics, the final visualization classification results have great distinction compared to the traditional algorithms. TVB-N was used as freshness uation criteria. The relevance between odor responses and TVB-N were established by multiple linear regression model, it reflected that smell ination were used to uate the freshness was feasible. The subject established prediction model for Chinese mitten crabs freshness based on BP neural network. Basically, we can consider that the model for uation of the freshness of Chinese mitten crabs is valid. This thesis has 33 figures, 8 tables and 73 references. Keywords machine olfaction; Chinese mitten crab; laplacian eigenmaps; prediction model 万方数据IV 目 录 摘 要 .................................................................. I 目 录 ................................................................. IV 图清单 ............................................................... VIII 表清单 .................................................................. X 变量注释表 ............................................................. XI 1 绪论 .................................................................. 1 1.1 研究背景和意义 ....................................................... 1 1.2 国内外相关技术的研究现状 ............................................. 1 1.3 课题的主要研究内容 ................................................... 5 2 机器嗅觉系统的设计 .................................................... 7 2.1 机器嗅觉系统的原理和整体方案设计 ..................................... 7 2.2 机器嗅觉系统的硬件设计 ............................................... 8 2.3 机器嗅觉系统的软件设计 .............................................. 15 2.4 机器嗅觉系统的调试 .................................................. 21 2.5 本章小结 ............................................................ 22 3 基于机器嗅觉系统的气味信息采集与处理 ................................. 23 3.1 气味信息的采集 ...................................................... 23 3.2 气味信息的预处理 .................................................... 26 3.3 特征选择 ............................................................ 29 3.4 特征提取 ............................................................ 30 3.5 本章小结 ............................................................ 38 4 中华绒螯蟹新鲜度评价和预测模型研究 ................................... 39 4.1 新鲜度评价标准 ...................................................... 39 4.2 气味响应曲线与新鲜度标准的相关性 .................................... 40 4.3 BP 神经网络建立新鲜度预测模型 ....................................... 44 4.4 本章小结 ............................................................ 50 5 总结与展望 ........................................................... 51 5.1 总结 ................................................................ 51 5.2 展望 ................................................................ 52 万方数据V 参考文献 ................................................................................................................................ 53 附录 1 ..................................................................................................................................... 58 附录 2 ..................................................................................................................................... 59 附录 3 ..................................................................................................................................... 60 作者简历 ................................................................................................................................ 62 学位论文原创性声明 ............................................................................................................ 63 学位论文数据集 .................................................................................................................... 64 万方数据VI Contents Abstract ................................................................................................................................... II Contents ................................................................................................................................. VI List of Figures ..................................................................................................................... VIII List of Tables ........................................................................................................................... X List of Variables..................................................................................................................... XI 1 Introduction .......................................................................................................................... 1 1.1 Background and Significance of the Subject ...................................................................... 1 1.2 Research Status of Related Technologies ........................................................................... 1 1.3 Main Research Contents of the Subject .............................................................................. 5 2 Design of the Machine Olfactory System ........................................................................... 7 2.1 Principle and the Overall Design of the Machine Olfactory System .................................. 7 2.2 The Hardware Design of the Machine Olfactory System ................................................... 8 2.3 The Software Design of the Machine Olfactory System .................................................. 15 2.4 Debugging of the Machine Olfactory System .................................................................. 21 2.5 Chapter Summary ............................................................................................................. 22 3 Smell Ination Acqusition and Processing Based on the Machine Olfactory System................................................................................................................................................. 23 3.1 Smell Ination Acqusition ........................................................................................... 23 3.2 Smell Ination Preprocessing ...................................................................................... 27 3.3 Feature Selection ............................................................................................................... 29 3.4 Feature Extraction ............................................................................................................. 30 3.5 Chapter Summary ............................................................................................................. 38 4 The uation of Chinese Mitten Crabs Freshness and Prediction Model Research 39 4.1 Freshness uation Standard ......................................................................................... 39 4.2 The Correlation Between Odor Response Curve and Freshness Standard ....................... 40 4.3 The Establishment of Freshness Prediction Model Based on BP Neural Network .......... 44 4.4 Chapter Summary ............................................................................................................. 50 5 Summary and Prospect ..................................................................................................... 51 5.1 Summary ........................................................................................................................... 51 万方数据VII 5.2 Prospect ............................................................................................................................. 52 References .............................................................................................................................. 53 Appendix1 .............................................................................................................................. 58 Appendix2 .............................................................................................................................. 59 Appendix3 .............................................................................................................................. 60 Author’s Resume ................................................................................................................... 62 Declaration of Thesis Originality ........................................................................................ 63 Thesis Data Collection .......................................................................................................... 64 万方数据VIII 图清单 List of Figures 图序号 图名称 页码 图 1-1 一些国外的电子鼻系统 4 Figure 1-1 Some foreign electronic nose systems 4 图 2-1 仿生嗅觉系统与人体嗅觉系统的对应结构图 7 Figure 2-1 The corresponding structure of bionic olfactory system and human olfactory system 7 图 2-2 系统整体设计结构图 8 Figure 2-2 The structure diagram of overall design 8 图 2-3 密封气室结构图 9 Figure 2-3 The structure of sealed chamber 9 图 2-4 电磁阀结构图 10 Figure 2-4 The structure of solenoid valve 10 图 2-5 气动元件控制电路图 10 Figure 2-5 The diagram of pneumatic components control circuit 10 图 2-6 传感器空间排布结构图 13 Figure 2-6 The diagram of sensors spatial arrangement 13 图 2-7 传感器的结构及信号拾取电路图 14 Figure 2-7 The diagram of sensor structure and signal pickup circuit 14 图 2-8 信号放大滤波电路图 15 Figure 2-8 The diagram of signal amplifying and filtering circuit 15 图 2-9 软件系统框架图 15 Figure 2-9 Framework map of software system 15 图 2-10 上位机登录界面图 16 Figure 2-10 PC login interface 16 图 2-11 板卡设置前后对照图 17 Figure 2-11 Contrast diagram before and after the board set 17 图 2-12 数据采集图 18 Figure 2-12 Data collection figure 18 图 2-13 数据存储图 18 Figure 2-13 Data storage figure 18 图 2-14 雷达图 18 Figure 2-14 Radar figure 18 图 2-15 上位机操作界面图 19 Figure 2-15 PC operation interface 19 图 2-16 主程序流程图 20 Figure 2-16 Flow chart of the main program 20 万方数据
展开阅读全文
收藏
下载资源

加入会员免费下载





足球比分直播