太阳能—相变蓄热新风控制系统的研究

【作者】 王丽丽；

【导师】 周晓光； 杨军；

【作者基本信息】 北京邮电大学 ， 机械电子工程， 2010， 博士

【摘要】 太阳能-相变蓄热新风系统是把太阳能技术和相变蓄热技术相结合用于房间有组织通风换气的新型辅助供暖系统。国内针对太阳能-相变蓄热新风系统的研究相对较少,对其热力学建模、控制算法及系统辨识的研究也不多见。而太阳能-相变蓄热新风系统的建模以及控制问题对整个系统的热利用效率至关重要,关系到能源消耗问题,将很可能成为这个领域的研究热点。为便于对新风系统理论分析和实验研究,先后搭建了太阳能-相变蓄热新风系统实验平台和模拟系统实验平台。在此基础上,通过对太阳能-相变蓄热新风模拟系统的研究和分析,本文提出了系统的运行模式及其控制方法；深入分析了系统各部件的数学模型；研究了太阳能-相变蓄热新风温度的控制策略,提出把模糊控制算法应用到太阳能-相变蓄热新风系统的控制中,并最终实现了基于CAN总线技术的太阳能-相变蓄热新风模拟系统的模糊控制。此外,本文还利用采集的新风温度和热媒水流量的数据,进行了基于自适应模糊神经网络的非线性系统建模与参数辨识的仿真研究。文中的主要内容和成果如下：1.利用一套太阳能热水—相变蓄热—空气/水新风换热体系,研究了太阳能-相变蓄热新风系统的工作原理,分析了在不同的供暖阶段系统的主要运行模式和各部分的运行规律,最终设计实现了基于CAN总线的太阳能-相变蓄热新风系统的测量与控制。2.设计并搭建了太阳能-相变蓄热新风模拟测控系统,首次提出了系统的三种控制模式及模式转换控制方案。在充分考虑系统非线性、滞后性的基础上,确定了系统总体控制目标,为整个控制系统的设计与性能评价确立了理论依据。3.基于太阳能-相变蓄热新风模拟系统,设计完成了手动控制实验。通过运行测控系统验证了系统设计的正确性,掌握了系统各部件的运行特性,同时获得了大量有效的实验数据,为新风模拟系统模糊控制的实现、自适应神经模糊建模与参数辨识提供了数据支持。4.研究并构建了太阳能-相变蓄热新风模拟系统的风机盘管数学模型、房间温度数学模型。其中房间温度数学模型充分考虑了外界环境温度以及热负载等干扰因素,以更准确地模拟实际系统特性,为验证模糊控制器设计的合理性与有效性夯实了基础.5.作为本文的主要研究内容,提出将Mamdani型的模糊控制算法用于太阳能-相变蓄热新风模拟系统的新风温度控制中,设计实现了系统三种运行模式的模糊控制器,并在建立的风机盘管的模型上进行了MATLAB仿真。6.针对太阳能-相变蓄热新风模拟系统的数学模型难以精确建立的特点,设计了基于自适应神经模糊推理的控制器。应用BP神经元网络的误差反传算法,对输入变量隶属度函数和输出变量进行参数辨识。结果表明：该算法采用神经网络与模糊推理结合的建模和参数辨识方法,可以很大程度上逼近太阳能-相变蓄热新风模拟系统的特性,误差几乎为零。7.设计并进行大量的太阳能-相变蓄热新风系统模糊控制实验,验证本文提出模糊控制算法的合理性。实验结果表明,本文所实现的模糊控制系统具有响应快速、鲁棒性强、稳定性高的特点,满足对新风出口温度的控制要求。本文的工作可以为今后太阳能-相变蓄热新风系统的控制研究提供理论基础和技术支持。更多还原

【Abstract】 In this dissertation, all the research work is mainly about a new type of fresh air system with the help of the solar energy and LHTS (phase change thermal storage) which can achieve organized ventilation.Only few of studies is on this type of fresh air system in China, so are the thermodynamic modeling, control algorithm and system identification. However, the modeling and control of this fresh air system is of great importance to the thermal efficiency of the system and related to energy consumption. So, the study in this dissertation will certainly become a hot topic in this area.To facilitate the theoretical analysis and experimental research of this new type of fresh air system, an experimental platform and a simulating platform were developed in turn. In this dissertation, the advanced monitoring systems were firstly realized with the help of Labwindows/CVI and CAN bus by using these two platforms. Then, taking the simulating fresh air system as the object, a series of studies were done thoroughly on the thermodynamic modeling, control strategies and parameter identification:the fan coil and room air temperature model were established; fuzzy control algorithm was proposed to this fresh air system, and also the simulation was done on the fan coil model; basing on the acquisition of fresh air temperature and heat water flow, the simulation of nonlinear system identification was realized using the adaptive neuro-fuzzy inference method; finally, lots of fuzzy control experiments on the simulating fresh air system were designed and accomplished. The main contents and results of this dissertation are as follows:The novel structure of a fresh air system was firstly introduced. And then the characteristics and the operation modes of the system were analyzed in detail. Measurement and control of the fresh air system with solar collectors and LHTS was finally realized basing on CAN bus.Secondly, an observing and controlling system of the simulating fresh air system was designed and finally established. The operation modes of the system and the way it transfers between modes were presented for the first time. In full consideration of the nonlinerity and hysteresis of the fresh air system, the overall control goal was determined. This work lays theoretical foundation for the establishment and performance evaluation of the whole controlling system.Afterwards, the manual experiments were designed based on the simulating fresh air system. Through experiments, the validity of the system is verified, the operation features of the system components are mastered, and a mass of valid test data are achieved, which provide sufficient data support for the fuzzy control realization of the fresh air system, adaptive neuro-fuzzy modeling and parameter identification.Fourthly, the fan coil and the room temperature model were focused on and established. In the room temperature model, the external environment such as out air temperature and heat load disturbances are all taken into consideration, which makes the control model simulate the actual system characteristics more accurately. This work is the solid basis to verify the reasonability and effectiveness of the design of the fuzzy controller.Fifthly, the Mamdani fuzzy algorithm was proposed for the first time to control the temperature of the simulating fresh air system according to the nonlinearity and hysteresis of the system. The fuzzy controllers of three operation modes of the system are also designed in workspace and simulated numerically with MATLAB.Sixthly, adaptive’neural fuzzy inference controller was designed according to the characteristics of the mathematical model, which combines the neural network theory with the T-S fuzzy model. Regarding the deviation and the rate of the deviation as the input and the controlling variable as output, the proposed controller was realized by continuous BP neural network using the approximation ability of the network.The adaptive neuro-fuzzy controller of the simulation results show a reasonable identification of the parameters, and the simulated output can approach the original output very closely and the error is almost zero, which verifies the proposed adaptive neural modeling.Finally, a number of fuzzy experiments were designed and carried out to verify the correctess and feasibility of the fuzzy arithmetic. The result shows that the fuzzy controller is of good performance of robustness and adaptability, and it meets the control requirement of the fresh air temperature.This study provides theoretical basis and technical support for future application of the fresh air system with solar collectors and LHTS.更多还原