【作者】 王新政；

【导师】 韩玉杰；

【作者基本信息】 东北林业大学 ， 机械设计及理论， 2012， 博士

【摘要】 禽舍养殖是我国农业发展战略的重要组成部分,规模化、集约化是我国现代化禽类养殖未来的发展方向。我国是畜禽养殖大国,禽蛋产量位居世界第一,然而我国目前的禽蛋产品在食品安全和卫生质量方面却处于劣势。随着禽流感等疾病的不断增加,尤其是H5N1型高致病性禽流感的出现,让人们清醒地意识到,家禽养殖已不再是单纯地涉及经济利益,而且也关系着人类自身的健康及生命安全。因此,给禽舍环境提出了更高的要求。由于鸡在家禽养殖中所占比重最大,也最具代表性,因此,本文以鸡舍为研究对象,探询禽舍养殖的一般规律。禽舍环境控制是利用一系列的工程设施,来保证家禽生长繁育的环境更合理,是禽业集约化和现代化的重要条件,但由于禽舍环境复杂、系统庞大,存在高度的非线性、时变和多变量耦合等特性,使得许多常规的检测手段和控制方法不能满足要求。因此,研究一套具有智能检测功能的禽舍环境测控系统,以及禽舍环境参数的智能控制方法具有重要的理论意义和应用价值。本文在分析禽舍环境内温湿度和空气质量测控机理的基础上,研究了基于无线传感网络平台的禽舍环境智能测控关键技术。给出了基于ZigBee协议的禽舍环境参数测量的无线传感网络结构及实现方法；提出了禽舍环境参数模糊数据融合检测方法；根据禽类饲养要求和禽类生存特性,分析了禽舍环境内,温湿度和禽类群居特性之间的变化规律,建立了禽舍环境参数变化与测控技术要求之间的关系,构建了一类基于ZigBee协议的禽舍环境测控系统,并对无线传感网络平台的实现进行了详细的分析和介绍,包括硬件选型和电路设计。提出了基于模糊理论的,禽舍环境参数多传感器数据融合的检测方法。针对多传感器数据融合过程中,传感器的可靠度估计值很难计算的问题,以禽舍温度为采集对象,分别提出了基于统计方法和时空融合方法的多传感器模糊贴近度计算方法,进行了实验分析,结果表明基于模糊贴近度统计方法具有较好的检测精度,算法简单,实时性好。针对实际禽舍温湿度系统的非线性、变参数等特性,在神经网络模型研究的基础上,深入研究了变结构温湿度控制系统的设计方法,基于Lyapunov稳定性理论分析了系统的稳定性,并针对不确定性满足匹配条件与非匹配条件,分别设计了两类控制系统,并就其等效系统给出了不确定系统稳定以及滑模的到达条件；同时,为了降低系统设计的保守性,将设计的控制器与神经网络控制技术相结合,给出一类智能变结构的温湿度控制过程系统设计方法。将禽舍温湿度调整过程看作为复杂系统,并针对一类具有系统参数摄动、关联不确定函数和外界干扰的禽舍温湿度调节复杂系统,基于Lyapunov稳定性理论和相应的合理假设,进而提出了一类分散变结构禽舍温湿度控制方案。此外,考虑到实际禽舍温湿度过程中系统对象不能确切已知,引入神经网络在线学习的方法对系统不确定性、扰动及系统关联函数进行在线估计,同时保证了闭环系统的渐近稳定性及各个子系统滑动模态的存在性和可达性。此外,为了消除控制的抖动,给出了基于边界层的改进分散滑模变结构控制的设计方法。为验证本文提出的基于无线传感器网络的禽舍环境测控系统的可行性,搭建了试验平台,通过无线节点对温度、湿度进行测试,并给出风机的继电器输出形式。开发了基于MCGS组态的禽舍环境监控系统,通过实际运行结果显示,监控系统运行正常,数据实时性高,控制曲线符合实际要求,具有广泛的推广应用价值。更多还原

【Abstract】 Poultry house breeding is an important part of our country agricultural development, and its strategy and scale are the future direction of development of modern poultry farming. Meanwhile, with the increasing of the avian flu and other diseases, especially H5N1highly pathogenic avian influenza, people are clearly aware poultry breeding is no longer simply involve economic interests, and also the human relations to their health and life safety. Therefore, it is necessary to put forward a higher demand to the poultry house environment. Poultry house environmental control is a series of projects and facilities, to ensure a reasonable environment for poultry growth and breeding, and also is an important condition for the poultry industry intensification and modernization; however, the poultry house environment is complex, large systems, highly nonlinear, time variable and multivariable coupling such that many of the conventional means of detection and control methods can not meet the requirements. Study the poultry house environment with intelligent detection monitoring system and the poultry house environment parameters of the intelligent control method has important theoretical significance and application value.In this paper, the analysis of the temperature and humidity in the poultry house environment and air quality monitoring and control mechanism on the basis of the poultry house environment based on wireless sensor network platform for intelligent monitoring and control of key technologies. Measurement of poultry house environmental parameters based on the Zigbee protocol wireless sensor network structure and implementation; the poultry house environment parameters of fuzzy data fusion detection; poultry breeding requirements and poultry survival characteristics within the poultry house environment, the variation of temperature and humidity, and poultry between the social characteristics of the relationship between the poultry house environment parameters change with the measurement and control technology requirements, build a the poultry house environment monitoring and control systems based on Zigbee protocol, and wireless sensor network platform, the class carried out a detailed analysis and presentation, including hardware selection and circuit design.The poultry house environment parameters based on fuzzy theory multi-sensor data fusion detection for multi-sensor data fusion, sensor reliability estimated value difficult to calculate, the poultry house temperature for a collection of objects, respectively, based on statistical The methods and spatial and temporal multi-sensor fusion methods fuzzy closeness method of calculation and experimental results verify the validity and accuracy of the proposed method.The controlling course of poultry house environment was a process of controlling the temperature and humidity of the medium in poultry house to make the moisture content to a certain expected value. The objective of poultry house control is to decrease the consumption of energy and to increase the quality of lumber at a minimum time. Conventional controller is widely applied in current poultry house, but it is based on the schedules developed by using empirical testing with the aim of providing reasonable throughout with acceptable levels of defects. This cannot adequately follow the different performance indexes, such as the minimum time and/or the energy of poultry house, etc. According to the characteristic of real poultry house environment, we examined the utilization of variable structure control and intelligent variable structure control in poultry house control system based on Neural Network model and the closed-loop stability was proved by using Lyapunov method. Moreover, for matched and mismatched uncertainties, two different control systems are developed for the poultry house system. And also for the boundedness of the uncertainties, the different kind of stable conditions are discussed.In addition, for the poultry house process, it is considered as a complicated system and the decentralized variable structure control system is presented and the sliding condition is also analyzed such that the stable condition of the closed-loop is given. More specifically, the dynamics characteristics of the poultry house environment were represented by two interconnected subsystems:heat-temperature process subsystem and temperature-moisture process subsystem. And then based on variable structure control theory, a robust poultry house control system is developed for this nonlinear interconnected poultry house environment such that the moisture content of poultry house will reach and be stabilized at the desired set point. Moreover, in order to reduce the control chattering, the modified decentralized variable structure controller is also developed by using a boundary value.To verify the feasibility of the proposed poultry house environment measurement and control system based on wireless sensor networks, set up a test platform through the wireless node to test the temperature, humidity, and gives the fan relay output in the form. Developed MCGS configuration of the poultry house environment monitoring system, through actual operating results to show the monitoring system is running properly, real-time data, the control curve in line with the actual requirements, with a wide range of application value.更多还原