【作者】 朱岩；

【导师】 杨历；

【作者基本信息】 河北工业大学 ， 热能工程， 2005， 硕士

【摘要】 设施农业是资源高效利用、技术和知识密集、增值效益显著、工业化程度高的农业新兴产业。现代自动控制温室是设施农业最基本的技术实现形式之一。它要在多变的自然气候条件下,为动植物的生命活动创造一个优化的生长环境,以实现高生产率、高产出、高品质、低成本、高效益、环境友好等可持续发展目标,因而越来越引起人们的重视,并成为一个具有重要意义的研究方向。 温室环境是多变量耦合、时变、非线性的复杂系统。温室环境调控是现代自控温室中最重要的关键技术之一,温室气候的建模是实现温室控制的关键。本文首先系统地分析了温室环境中的主要热力学过程,在参照国内外研究成果的基础上,针对我国现有温室的调控特点,提出了含有控制量的温室温湿度动态模型;接下来详细分析了自控温室系统中的气候控制问题,包括影响作物生长的各个主要环境因子以及对象的控制特点。在分析温室被控对象复杂性的基础上及功能分解的前提下,对温室环境控制系统进行总体设计,提出冬夏季室内控制方案。总的来说,这一部分关于控制方法的研究仅仅作为一种新的尝试,还有待于进一步的深入。 最后本文将模糊理论的知识表达与神经网络的自学习能力有机地结合起来,采用快速学习算法,提出了一种模糊神经网络控制方案,并针对冬季温室内加热系统的控制,在MATLAB语言的SIMULINK平台下分别对模糊神经网络控制、常规PID控制和模糊控制进行仿真比较。从仿真结果可以得知:模糊神经网络控制具有更小的超调,无振荡,平稳性好,达到稳定状态的时间短,稳态误差小,因此其动态特性和静态特性均最为优越。由此验证了模糊神经网络这种控制方法应用于温室自动控制系统的可行性。更多还原

【Abstract】 Equipment agriculture is a burgeoning industry, which can make full use of resource and have high-level of automation. It not only is technique and intelligence concentrated but also will bring you satisfy appreciation profit. Modern automatic control greenhouse is one of the realization forms of facility agriculture, which can make a man-made climate of the greenhouse or eliminate the disadvantages for crop. In such a greenhouse, the produce period of crop can be shortened. Meanwhile the aim of high-quality, high-output, high-benefit can be reached.Greenhouse is a multi-variable coupled time-varying and non-linear system. Its environment control is not only one of the key techniques in modern automatic control but also difficult to control the greenhouse climate. The modeling of greenhouse climate is the key to realizing greenhouse control. This paper has analyzed the main thermodynamics processes in the climate of the greenhouse systematically at first, on the basis of consulting foreign research results, direct against the regulation and control characteristic of the existing greenhouse of our country, have proposed containing the dynamic model of temperature and humidity of greenhouse which controls quantity; And then analyzed in detail the climate in the greenhouse system of the automatic control controls the question, including influence each main environmental factor and the target’s control characteristic that the crop grows. On the basis of analyzing that is accused of the target’s complexity in the greenhouse and under the prerequisite that the function resolve, design overall the environmental control system of the greenhouse, propose controlling the scheme in the room in summer and winter. Generally speaking, this chapter is about controlling the research of the method only as a kind of new try, wait further deepening.This text finally presents a kind of fuzzy neural network method which combines fuzzy knowledge representation with neural network self-learning ability and adopts a fast study algorithm. For the greenhouse heating systematic control in winter, I have been going on the simulation compare with the FNNC、 routine PID control and fuzzy separately by the SIMULINK platform of MATLAB. In the end we can learn from the artificial result that FNNC has lighter exceeding suiting well to control, no shaking, fine stationarity, short time to reach the stable state, little error of the stable state. So its dynamic characteristic and static characteristic is the most superior. Therefore it has verified that the feasibility of using FNNC.更多还原