【作者】 史兵；

【导师】 赵德安；

【作者基本信息】 江苏大学 ， 农业电气化与自动化， 2013， 博士

【摘要】 随着人们生活水平的不断提高,对水产品的需求也在持续增长,我国已成为水产品生产和消费的大国,2012年全国水产品总产量达到5906万吨,取得了巨大的经济效益。河蟹是我国主要的水产品种之一,味道鲜美、营养丰富,除了食用之外,河蟹的甲壳等部分还可以使用在工业生产、医学治疗等诸多方面,已取得了良好的社会效益。河蟹在我国分布广泛,养殖方式有池塘养殖、湖泊养殖、河段围栏养殖、稻田养殖等多种形式,受地理环境特点的影响,池塘养殖是养殖户主要采取的养殖方式之一。池塘养殖虽然实施起来较为便利,投入较少,但也存在着养殖分散、多数养殖户不具备相应专业知识,缺乏规范化、标准化的指导,从而导致养殖过程自动化程度低、产品规格差异大、病害严重、产品安全无保障等诸多问题。为了解决池塘养殖存在的这些问题,在江苏省“物联网示范工程”项目的支撑下,开展了河蟹池塘养殖智能支持系统一些关键技术的研究。本文主要围绕以下工作展开：首先,针对大范围、多数据的参数获取技术展开研究。根据河蟹池塘养殖基地通常地势平坦、范围广等实际情况,引入了无线传感器网络技术,在养殖基地建立了一套无线传感器网络系统。通过对GAF拓扑控制算法的深入研究,提出了GAF-G改进拓扑算法,实现了固定分簇、固定簇头的拓扑结构,并实现了簇内节点与簇头通信,簇间簇头之间通信的路由关系。通过对池塘水体断面数据的分析,确定了池塘水体各主要指标的分布情况,从而为传感器节点的布置提供了依据,以实现优化配置。其次,针对池塘各水质因子之间的相互关系展开研究,确定各因子之间的内在联系,并针对主要因子提出不同的调节措施。在深入分析各因子内在关系的基础上,利用基于模糊PID控制算法,对典型河蟹养殖池中溶氧量进行控制,并通过仿真与实际测试确定了控制效果。接着,针对池塘水质评价工作展开研究,明确了水质评价的重要意义和方法。通过专家咨询的方式,确定了针对河蟹池塘养殖的最主要水质指标：溶解氧、酸碱度(pH值)、温度和盐度。并对这四项指标进行量化分级,分为Ⅰ-V级,共五级水质。详细介绍了目前较为常用的几种水质评价算法,并通过几种算法的应用实例,比较几种算法的优缺点及适用场合。然后,针对河蟹养殖专家支持系统展开研究。介绍了专家系统在农业方面的应用及发展情况,和专家系统推理机的工作原理,建立了河蟹养殖病害知识库。通过深入研究Rete匹配算法的工作方式,提出了基于索引列表的改进策略。通过分析冲突集的工作方式,提出了基于指针链表的改进策略,并通过仿真确认了改进效果。最后,针对智能系统的实现展开工作,利用现有的网络技术、数据库技术和通信技术实现了系统集成,完成了智能管理系统的建设。更多还原

【Abstract】 With the improvement of people’s living standard and the increasing demand for aquatic products, China has become one of the largest aquaculture and consumer countries. In2012, the national aquatic production reached59.06million tons. The huge economic benefits were brought. Crab is one of the major aquaculture species, and it is one kind of delicious and nutritious food. In addition to food, some parts of the crab can also be used in industrial production, medical treatment and other aspects. The social benefits are achieved significantly. Crab breeding widely distributes in China, and there are several breeding methods, such as pond breeding, lake breeding, and fence breeding. Due to the characteristics of the geographical environment, pond breeding is one of the mainly methods. Although pond breeding is more convenient and less investment, there are also some problems, such as scattered farming, lack of standardization guidance, and so on. All these disadvantages result in a low degree of automation of the breeding process, large differences in product specifications, serious diseases, lack of product safety and many other issues.In order to solve these problems in crab breeding, some of the key technologies of intelligent support system for crab breeding were researched, which was based on the demonstration projects of internet of things in Jiangsu province. The work mainly focuses on the following aspects.Firstly, research on technology of collecting large-scale, multi-data was carried out. According to the actual situation of crab breeding in ponds, the technology of wireless sensor network was introduced, and a breeding center based on the wireless sensor network system was established. Through in-depth study on topology algorithm of GAF, one kind of improved topology algorithm of GAF-G was proposed. The improved algorithm implemented the topology of fixed clustering, the fixed cluster head, and implemented communications in clusters. Based on study on distribution of the main water factors in crab ponds, the deployment of sensor nodes was determined and the optimal allocation was achieved. Secondly, research on the intrinsic link between each water factor in crab ponds was carried out. Based on analysis of the intrinsic relationship of each factor, one kind of fuzzy PID algorithm was proposed to control dissolved oxygen in crab ponds. The results of simulation and actual testing confirmed the control effects.Thirdly, research on the evaluation system for pond water quality, and evaluation methods were carried out. According to experts’advice, the most important indicators for water quality of crab ponds were determined:dissolved oxygen, pH, temperature and salinity. In addition, quantitative classification of four indicators was divided into Ⅰ～Ⅴ level. Several evaluation algorithms’for water quality were introduced in details in the evaluation system, and through several application examples, the advantages and disadvantages of different algorithms were described.Fourthly, research on crab breeding expert support system was carried out. The development and applications of expert system in agriculture were described, and the reasoning mechanism of expert system was discussed in details. One kind of knowledge base for crab breeding was established. Through in-depth research on Rete algorithm, one kind of improved matching algorithm of Rete was proposed based on method of index. By analyzing the working process of conflict set, one method based on chain was introduced. The improved effectiveness of new method was confirmed by simulation.Finally, the system integration of intelligent system was implemented, the technology of network, database and communications were adopted to accomplish the intelligent system.更多还原