【作者】 史云；

【导师】 钱东平；

【作者基本信息】 河北农业大学 ， 农业机械化工程， 2009， 博士

【摘要】 水资源问题是二十一世纪我国乃至世界面临的主要问题。随着世界人口的增长及工农业生产的发展,需水量在日益增长,水已经变得比以往任何时候都要珍贵。与此同时,由于人类的生产和生活,导致水体的污染,水质恶化,使得有限的水资源更加紧张。长期以来,石油类物质一直是水和土壤中的重要污染源。它不仅对人的身体健康带来极大危害,而且使水质恶化,严重破坏水体生态平衡。因此各国都加强了油类物质对水体和土壤的污染的治理。目前,我国对于水中油含量的检测手段落后,与国际先进水平存在差距,传统的检测方法操作过程繁琐,容易引入误差,且不能现场作业。难以满足当今技术水平的要求。为了取得具有代表性的正确数据,使分析数据具有与现代测试技术水平相应的准确性和先进性,不断提高分析成果的可比性和应用效果,检测的方法和仪器是非常重要的。只有保证了这两方面才能保证快速和准确地测量出水中油类污染物含量,以达到保护和治理水污染的目的。因此,论文的目的就是开展水中油污染检测方法、技术和检测设备的研究,通过本课题的研究,探索出一套适合我国国情的水质污染现场检测技术和检测设备,以实现油类污染的现场检测,为地下水资源调查与评价、水资源保护以及农业发展提供科学依据。论文在简要分析地下水油类污染现场检测技术的基本理论-朗伯比尔定律的基础上,对实现石油类污染现场检测的方法原理进行研究,提出了系统测量的基本模型,设计了系统实现的总体方案及工作流程。针对基本模型在实现油污染现场检测中存在的问题,本文又提出了将径向基函数神经网络的软测量方法应用于油含量现场检测中的设计思路,利用神经网络很强的非线性逼近能力和学习能力,建立了目标系统的径向基函数神经网络模型,取得了很好的效果。论文运用了一些关键技术和思想,进行了系统的软硬件设计,在硬件设计中,采用气体滤波相关轮和窄带红外滤光片等技术,完成了光学系统的硬件设计,解决了红外探测器的信号衰减、气体浓度信号失真等关键问题。采用单片机智能控制方法和技术,实现了数据的采集、存储、传输、控制、运算等功能;在软件设计中,采用了嵌入式多任务内核(RTX51Tiny)作为其开发平台,扩展系统硬件功能,完善了抗干扰措施。最后,通过系统测试,验证了系统的准确性和可靠性。通过本文的研究,得到以下结论:论文通过对矿物油含量测量机理、光学系统、智能化检测技术以及基于神经网络的软测量方法等方面的研究,对油类污染现场检测的技术关键逐一进行了解决,最终探索出一套适合现场使用的检测方法,完成了检测系统的设计及一系列的试验测试。通过一系列试验测试表明,该系统符合设计要求。论文集先进的传感器技术、计算机技术、单片机技术及光学检测技术于一体,以此来实现地下水的油类污染的现场检测。研究中充分利用单片机技术,实现了检测系统的智能控制;此外,本文采用基于神经网络的软测量方法,对矿物油的吸光度与油的浓度所确立的目标系统进行建模,并利用神经网络很强的非线性逼近能力和学习能力,以实现本文的最终目标,提高检测系统的准确度和测量范围,简化了测量过程。因此,论文成果充分体现了创新性,为地下水石油类污染检测提供了新技术和新设备,开拓了新的思路。本项成果的推广应用,将会使我国地下水污染检测技术达到一个新的水平,为我国地下水污染检测整体水平的提高起到重要的促进作用。更多还原

【Abstract】 The water resources question is the main matter in our country and even the world in the 21st century. Along with world population’s growth and the industry and agriculture development, the requirement of water is growing day by day and the water has got more precious than the past. Meanwhile, as a result of humanity’s production and the life, it causes the pollution of the water body and the water quality to worsen. It makes the limited water resources be less. Petroleum-series of Contaminants has been the important source of pollution in water and soil for a long time. It not only brings the enormous harm to person’s health, but also destroys the water body ecological equilibrium seriously. Therefore many countries strengthen to fight against the pollution of Petroleum-series of Contaminants in the Environmental Water and soil. At present, the method of detecting Petroleum-series Contaminants in our country is backward. Between our country and the developed country has the gap. The operating process of the traditional examination method is tedious, easy to cause the error, and cannot in-site detect. It is difficult to satisfy the technical level’s wanting. In order to obtain the representative correct data to improve the accuracy of application, the examination method and the instrument is very important. Only we guaranteed the advance of the method and the precision of the instrument, we could guarantee detecting the contents of Petroleum-series Contaminants fleetly and truly. We can gain our ends for solving the problem of water pollution and conservation. Therefore the purpose of the essay is to study the method of detecting the Petroleum-series Contaminants in water and develop the device. Though the study of the essay, we will find out the detecting method, technique and device adopting to our country situation. Then we can realize the in-situ detecting the Petroleum-series Contaminants and provide the scientific basis for the research, evaluation and conservation of water resource, and development of agriculture.The essay studies the theory of in-situ detecting the Petroleum-series Contaminants and proposes the system survey’s fundamental model, designs the overall solutions and work flow of system realization on the basis of simply analyzing the elementary theory - Beer-Lambert law. In view of fundamental model have some faults on in-situ detecting the Petroleum-series Contaminants, this article also proposed a new designing thought. The thought is to apply the soft sensing method based on radial basis function (RBF) neural network to detect the Petroleum-series Contaminants in water. Applying the strong nonlinear approximate and studied ability of neural network, we have established the neural network model of system based RBF and had made the very good progress. The essay has utilized some key technologies and the thought to finish the system’s software and hardware design. In the hardware design of optical system, we have applied some techniques including the gas filter theory of theory and the infrared narrow band filter to solve the key question involving infrared acquisition aid’s signal attenuation, the gas strength signal distortion and so on. Using the Single chip computer (SCM) intelligent control method and technique, the system has realized functions including data gathering, memory, transmission, control, operation and so on. In the software design, the system has used the embedded multitasking kernel (RTX51Tiny) to become the development platform. It expands the system’s hardware function and consummates the ant jamming measure. Finally we have confirmed system’s accuracy and the reliability through the system testing.We obtain the following conclusion through this article researching:This essay has resolved the key problem of in-situ detecting the Petroleum-series Contaminants through the study of theory of Contaminant content , the optical system, intellectualized examination technology as well as soft sense technique based on neural network. We have explored a set of methods fitting in-situ detecting the Petroleum-series Contaminants. Finally we have completed the detect system’s design and a series of experiment. Through a series of indoor simulation and the field test ,it indicated that this set of methods and the detect system can meet the requirement of field observation long time.The essay integrated the advanced sensor technology, the computer technology, the SCM technology and optical detect technology and realized in-situ detect of the Petroleum-series Contaminants in water. During the study of this essay, we realized intelligent control of the detect system through making the best of the SCM technique. In addition, this essay sets up objective system modeling based on the neural network soft sense technique aiming at the gas’s absorbency and the density. The essay’s purpose is to improve the system’s accuracy and the measuring range and simplify the measuring process. We realized the purpose through using nonlinear approximate ability and learning ability of neural network .Therefore, the paper’s achievement has manifested the innovation fully and provided the new technology and the new equipment for the detecting Petroleum-series contaminants in underground water.The popularization and application of this achievement will enable our country underground water condiment examination technology to achieve a new level. It will promote the whole detecting level of underground water condiment improvement in our country.更多还原