仿生咀嚼过程中食品物料力声行为及质地感知评价机制

Food texture is an important index of food quality evaluation. It also is a key factor for customers to choose food. So, methods of food texture evaluation have been paid much attention. But, there are bigger differences between two kinds of inspection results for food texture by a nowaday instrument and by a human chewing, which do not reflect the true relation between food texture and human chewing sense. So, a new type of measurement instrument for food texture is urgently demanded. Measure method of food texture by bionic masticator is a new approach. In the project, bionics and modern detective technology are used to study change mechanisms and control rules of representative food material. And their impact factors are researched. Change principles of physics state and chemistry components of food materials in imitated chewing will be found. CATIA software are used to analyze structures of bionic chewing equipment. A computer simulation model of the bionic chewing equipment is developed. Emulate analysis for the bionic chewing equipment is done in order to acquire optimization structure, movement parameters, materials of key parts such as bionic maxilla and bionic mandible of bionic chewing equipment. Based on tradition Distinct Element Method, a redevelopment on masticate equipment are done. A bionic masticate digit dynamic simulation model of representative food materials are founded in order to gain optimal structure and parameters. Mechanics and acoustics behavior of food materials is investigated. Modern detecting and analysis technologies are used to extract mechanics properties，air acoustic characteristics, solid vibration spectrum characters in chewing of bionic masticate equipment and to explore relationship between the attributes and food texture quality. Disciplinarians of acoustics transmit characteristics in producing, transfer, absord of sound wave of representative food materials are analyzed by modern acoustics spectrum analysis theory. Acoustic transfer mathematic models of representative food materials are made up. Acoustic transfer regulations and their impact factors are researched to discover inherence relation between bionic chewing attributes and food texture quality. And food masticatory acoustic mechanism will be explained. The masticate parameters which are prominence correlative to food texture characters such as hardness, fracturability, springiness, chewiness, cohesiveness are chosen by Principal Component Analysis. Texture pattern of representative food materials are recognized by the optimized intelligent methods. Evaluating mechanism to apperceive texture in bionic chewing is been researched. A new method of solid vibration spectrum instead of air sound wave combining with mechanics characters will be used to measure quantificationally food texture. Appraising accuracy is more than 90%. All the researches are for laying a foundation in order to develop new food texture inspection instrument.

目前食品质地仪器检测与人咀嚼的食品感官真实感觉存在较大差异，不能反映食品质构与人咀嚼特征的真实关系,急需探索食品质地检测新方法。本项目利用仿生和数值分析方法对仿生咀嚼装置结构形式和参数进行分析，构建仿生咀嚼计算机仿真模型，在模拟分析基础上进行实验研究，获得咀嚼装置最优结构形式和最佳运动参数；利用现代检测技术，对食品物料仿生咀嚼机理和操控准则及其影响因素进行研究，揭示食品物料在仿生咀嚼过程中物理状态及化学成分的变化机理；利用现代声波谱分析理论，揭示典型食品物料的声波产生、传递、吸收等声学规律，构建食品物料声学传播数学模型，解释食品咀嚼声学机理，发现仿生咀嚼力声学特征与食品质地指标之间存在的某种内在关系的规律，探求仿生咀嚼食品质地感知评价机制，利用固体振动波代替空气声波与力学特征相结合的优化智能新方法定量判断食品质地品质，检测正确率在90%以上，为开发新型现代食品质地检测仪器奠定基础。

目前食品质地仪器检测与人咀嚼时真实感觉存在较大差异，不能反映食品质构与人咀嚼特征真实关系。本项目利用离散元法优化了舌块表面的倾斜度。构建了仿咀嚼装置仿真模型，优化了咀嚼装置结构形式和运动参数，开发出口腔环境和咀嚼参数可自动控制的仿咀嚼装置。咀嚼效果验证结果表明，仿咀嚼装置可反映食品质构与人咀嚼特征的真实关系。利用有限元法分析了咀嚼时下颌运动规律对食品物料破坏效果的影响，结果表明：研磨模式会产生更大的剪切应力和密赛斯应力，对弹性和韧性食品有良好的破坏效果，咀嚼速率越大，破坏效果越好；切割模式产生的主应力最大，适合切咬脆性食品。利用离散元法建立了咀嚼破碎模型，模拟食品咀嚼的动态破碎过程，结果表明，咀嚼破碎模型分析结果与感官评价结果一致。利用离散元法分析了几何形状对食品质地特性的影响,结果表明：截面积相同，样品厚度越大，硬度、弹性和断裂能量越大；且圆柱体的断裂能量最小，最易断裂。研究了人类咀嚼过程中牛肉的微观结构和物理化学变化，结果表明，咀嚼运动几乎对牛肉微观结构没有影响，牛肉化学性质的变化很小。对颗粒状和黏合状两种食物团的研究揭示了两种食物团在咀嚼中的状态及变化规律；食品质地与肌电信号特征值和苹果微观结构与宏观质构间关系研究揭示了人类咀嚼食物的机制。利用现代声波谱技术对4种典型脆性食品物料声波信号进行了处理和分析，并提取了断裂声音信号的时频特征；利用模态分析技术分析了12种材料（4种金属材料、4种非金属材料和4种食品物料）的振动特性，揭示了食品物料和工程材料的断裂声音及振动波的传播特征和规律。苹果、胡萝卜和休闲食品的振动特性与质构参数相关性的探索，以及利用扫描电镜对细胞壁结构变化规律的观察，揭示了食品脆性、硬度与断裂声音信号特征及微观结构的关系。4种典型物料脆性预测模型的研究表明，基于希尔伯特黄变换的BP神经网络预测食品脆度的正确率可达到90%以上，为开发新型现代食品质地检测仪器奠定了基础。