多源信息融合方法的精密鱼类喂养

发布时间：2018-10-16 12:51

【摘要】：水产养殖是提高鱼类食物或者渔业资源的有效途径。近些年,差不多50%的海鲜源自水产养殖,并以每年大约9%的增长率增长。然而在水产养殖过程中,由于技术和资源利用不当,造成差不多有60%的饲料被浪费。合适的水产养殖管理,例如:水质管理、鱼类喂食管理、鱼病诊断管理和技术资源的有效利用等能有效解决以上问题。针对水产养殖中的鱼类生长所采用的传感技术、无线传感网络、监控系统、机器视觉以及建模策略已不再新颖。大多数与建模相关的文章多与多因子缺省数学模型相关,技术相关的文章缺少整合系统。因此,本文的目的是研究与讨论影响采食量建模的因子,并以信息融合的方式整合在水产养殖中应用的技术。此外,基于对声波压力数据的时间序列分析,构建依据鱼类生物量探索水情变化的软件系统。本文简要地确定并描述了当前的采食量建模发展现状,并概述传感器技术、机器视觉技术在水产养殖方面的应用,例如水质、鱼类行为、鱼病、废弃物管理。作者提出环境因子、营养因素、饲养因素、生理学因素是影响采食量所需考虑的重要因素,并以此为依据对饲养策略进行建模,以更有助于鱼类的生长。作者同时提出了基于传感器与机器视觉,对多方面的水产养殖因素例如水质、鱼类投喂、鱼类行为、鱼病诊断与废弃物管理的进行融合。此外,在实验室环境下设计的监控系统已被用于研究鱼类生物量对声波压力频率、振幅、周期的影响。我们采用声波压力传感器、模拟接收器(示波器)与软件程序flukview来检测在水产养殖用箱体中的波动现象。另外,文章还对在有鱼、无鱼状态下的时间函数的参数进行了对比与数理统计。同样地,以统计学的方式解释了基于不同音程的离差测量(MOD)以及中心趋势测量(MCT)结果。本研究证明了声频测量是一种简单而有效的集约化水产养殖环境分析工具。相对于在空气中传播,声波信号在水中传播迅速,因此可以通过声波压力频率测量出来。本研究成果是对基于声波压力测量来控制饲养废弃物、高效饲养来提高水产养殖鱼类生长的创新。本研究在结论部分对使用声波压力测量的需求、以及集成声波技术与机器视觉技术的混合系统进行了展望。
[Abstract]:Aquaculture is an effective way to improve fish food or fishery resources. In recent years, almost 50% of seafood has come from aquaculture, growing at an annual rate of about 9%. However, in aquaculture, almost 60% of feed is wasted due to improper use of technology and resources. Appropriate aquaculture management, such as water quality management, fish feeding management, fish disease diagnosis management and effective use of technical resources, can effectively solve the above problems. The sensing techniques, wireless sensor networks, monitoring systems, machine vision and modeling strategies for fish growth in aquaculture are no longer novel. Most of the articles related to modeling are related to the multi-factor default mathematical model, and the technical related articles lack of integrated system. Therefore, the purpose of this paper is to study and discuss the factors that affect the modeling of feed intake, and integrate the techniques used in aquaculture by information fusion. In addition, based on the time series analysis of acoustic pressure data, a software system is constructed to explore water regime changes based on fish biomass. This paper briefly defines and describes the current situation of food intake modeling, and summarizes the applications of sensor technology and machine vision technology in aquaculture, such as water quality, fish behavior, fish disease, waste management. The author puts forward that environmental factors, nutrition factors, feeding factors and physiological factors are the important factors that need to be considered to influence the feed intake, and based on these factors, the feeding strategy is modeled, so as to be more helpful to the growth of fish. At the same time, based on sensor and machine vision, the authors proposed the fusion of various aquaculture factors such as water quality, fish feeding, fish behavior, fish disease diagnosis and waste management. In addition, the monitoring system designed in laboratory environment has been used to study the effects of fish biomass on acoustic pressure frequency, amplitude and period. Acoustic pressure sensors, analog receivers (oscilloscopes) and software program flukview are used to detect fluctuations in aquiculture tanks. In addition, the parameters of time function under the condition of fish and fish are compared with mathematical statistics. Similarly, the results of deviation measurement (MOD) based on different intervals and central trend measurement (MCT) are statistically explained. This study proved that audio measurement is a simple and effective tool for intensive aquaculture environment analysis. Sound signals travel rapidly in water relative to air, and can therefore be measured by acoustic pressure frequencies. This research is based on acoustic pressure measurement to control the feeding waste, high-efficiency feeding to improve the growth of aquaculture fish innovation. In the conclusion of this study, the requirement of acoustic pressure measurement and the hybrid system of acoustic wave technology and machine vision technology are prospected.
【学位授予单位】：中国农业大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：S951.2

【参考文献】