鸡蛋破损自动检测分级系统的研究-农业机械化工程专业论文.docxVIP

摘 摘 要 鸡蛋是人们日常生活中一种非常有营养的副食品。在鸡蛋加工过程中，剔除破 损蛋是保证加工质量及节约成本的一道重要工序。目前常用的方法是人工检测法， 检测工作劳动强度大，检测效率和精度都很低。因此，探寻科学适用的鸡蛋破损检 测技术是一项非常有意义的工作。 本文采用声学特性检测技术对鸡蛋的破损情况进行检测。为了使鸡蛋蛋壳发出 能够反应其破损状况的声音信号，设计了以单片机为核心的自动敲击装置，该装置 通过光电传感器感知是否有鸡蛋进入敲击装置，并由单片机控制步进电机带动敲击 棒对鸡蛋的大头，中间及小头等三个不同部位进行敲击而使其发声。为保证对此声 音信号进行实时处理，设计了基于DSP的检测分级装置，鸡蛋受敲击后产生的声音 信号经过放大、滤波及AD转换后进入到DSP内进行处理。为及时剔除破损蛋，还 设计了由DSP控制的自动分级执行机构。 对采集到的声音信号，主要是对其进行频谱分析，通过对鸡蛋大头、中间及小 头等三个不同部位声音信号的频谱分析，找出区分好壳蛋和破损蛋的显著特征参数， 在充分考虑各个特征参数的相互之间的独立性的基础上，选出最优的特征参数组合， 并利用Bayes判别原理建立了区分好壳蛋和破损蛋的判别模型。 在软件方面，自动敲击装置的单片机系统软件和检测分级装置的DSP系统软件 均采用汇编语言编写，提高了系统软件执行的效率，而且减小了对系统程序存储器 的扩展要求，降低了系统硬件的复杂性。 由本文设计的自动敲击装置和检测分级装置所共同组成的鸡蛋破损自动检测分 级系统，其工作参数能够满足课题组前期研究成果“禽蛋品质无损自动检测分级设 备”的需求；系统正常工作时，其检测能力可达每小时3600枚蛋；对好壳蛋检测准 确率为93．22％；破损蛋检测准确率为85．61％：系统的总体检测准确率为89．2％。 关ftt讵q：破损检测，数字信号处理器，单片机，步进电机，频谱，Bayes判别 Egg Egg is one of nourishing foods in our daily lifc．Getting rid of cracked eggs is all important procedure to guarantee good quality of processing eggs and decrease its cost． Detecting cracked eggs by hands is very difficul哆for its bad working conditions and laboring intension,as well as the low efficiency and precision．Consequentlyit’S very urgent and valuable to find a scientific，applicable and fast detecting method for removing cracked eggs． In this task,the acoustic characteristics were used for detecting the cracked eggs．An automatic knocking device based OB MCU Was designed to make the eggshell emit characteristic sound．If there was an egg in the device，a photoelectric sensor would send a signal to the MCU．Then the MCU drove three step motors which were connected with knocking sticks to knock different parts of the eggs,the biggest part，the middle part and the smallest part．In order to process the real·time sound signal steadily,a testing and grading device based Oil DSP Was designed．The sound signal from the eggshell Was sent to the DSP through the amplifying circuit，the filter and the AD convertor．And there Was an machine controlled by the DSP,which could get rid of th