【作者】 齐湘洪；

【导师】 朱明；

【作者基本信息】 华中农业大学 ， 机械硕士（专业学位）， 2022， 硕士

【摘要】 中国水产养殖产量位居世界首位,其中传统池塘养殖占很大比重。然而,近年来,传统池塘养殖逐渐暴露出生产力低下、环境污染严重、过分依赖人为经验等诸多问题,且随着长江禁捕等政策的出台,池塘传统散养模式面临更大危机,迫切需要转型升级。池塘圈养是一种新型的水产养殖模式,其对池塘地区环境要求低,适应性广,并能显著节约水资源、降低人力成本,增产效果显著,且能耗下降空间较大,具有广阔的发展空间,但是圈养模式目前的自动化、机械化程度仍然较低,从一定程度上限制了圈养模式的大批量推广应用。本文以圈养池塘为研究对象,针对水产养殖中的关键环节及瓶颈,设计了一套自动管控系统,实现对水质数据实时采集与水质状态管控、水产品饲喂、供氧及清污等。论文主要内容为下:(1)完成了自动管控系统总体架构及方案设计。在开展水产养殖管控平台需求分析的基础上,设计了圈养自动管控系统总体架构,包括主控模块、水质监测模块、饲喂模块、供氧模块、清污模块、视频监控及通信模块等。该系统具备多种信息传递方式,且能够通过触摸屏、手机APP等多种手段实现水产养殖管控。(2)完成水产养殖水质数据预处理、关联度分析及水质预测研究。针对水质传感器的数据异常、噪声等问题,选择小波滤波算法对水质数据进行滤波处理;基于实测的多源传感器数据,开展关联度分析,确定各种水质参数间的关联程度;为降低溶解氧传感器出现故障对养殖的影响,针对溶解氧与温度、p H、浊度之间的强关联度,采用人工神经网络方法建立溶解氧水质预测模型。(3)水产养殖管控软件的设计与实现。面向水质监测及系统管控问题,开展管控软件设计工作,包括圈养池塘下位机软件、基于Android Studio IDE开发的手机APP、触摸屏软件等。基于上述软件,实现了圈养池塘的信息实时采集、定时/条件触发式养殖管控,且可在本地及远程的多台设备上同步圈养池塘系统信息,远程/就地发布指令。实验表明,该系统运行效果较好,可以实现高效的自动化水产养殖。更多还原

【Abstract】 China’s aquaculture production ranks first in the world,of which traditional pond culture accounts for a large proportion.However,in recent years,traditional pond aquaculture has gradually exposed many problems such as low productivity,serious environmental pollution,and excessive reliance on the human experience.With the promulgation of policies such as the banning on fishing in the Yangtze River,the traditional pond aquaculture industry is facing with greater crisis,thus leading to more urgent needs for transformation and upgrading.Captive breeding is a new type of aquaculture mode,which has a variety of advantages including low requirements for regional environment,wide adaptability,and significantly saving water resources,reducing labor costs,increasing production notably,large energy consumption reduction space,and broad development space.However,the current degree of automation and mechanization of the captive breeding model is still low,which limits the mass promotion and application to a certain extent.In this work,the captive ponds were taken as the research object,and an automatic control system was designed for the key links and bottlenecks in aquaculture to achieve real-time collection of water quality data,water quality status control,product feeding,oxygen supply,and pollution removal.The main contents of the paper are listed as follows:(1)The overall structure and scheme design of the automatic management and control system were finished.Based on the demand analysis of the aquaculture management and control platform,the overall structure of the captive automatic management and control system was designed,including the main control module,water quality monitoring module,feeding module,oxygen supply module,cleaning module,video monitoring,and communication modules,etc.The system had a variety of information transmission methods,and could realize aquaculture management and control through a touch screen,mobile phone APP,and other means.(2)The investigation of aquaculture water quality data preprocessing,correlation analysis,and water quality prediction was finished.Aiming at problems such as abnormal data and noise of water quality sensors,the wavelet filtering algorithm was selected to filter water quality data.Based on the measured multi-source sensor data,correlation analysis was carried out to determine the degree of correlation between various water quality parameters.In order to reduce the impact of dissolved oxygen sensor failure on aquaculture,an artificial neural network method was used to establish a dissolved oxygen water quality prediction model,based on the strong correlation between dissolved oxygen and temperature,p H,and turbidity.(3)The design and implementation of aquaculture management and control software.The management and control software such as captive pond lower computer software,an Android studio IDE based mobile APP and touch screen software were designed to monitor water quality and conduct system control.On the basis of the above software,real-time collection of information in captive ponds,timing/conditiontriggered aquaculture management and control can be realized,and captive pond system information can be synchronized on multiple local and remote devices,meanwhile the instructions can be issued remotely or locally.At last,the experiments showed that the system was suitable and efficient for captive pond.更多还原