宇宙飞船常处在温度变化剧烈的太空环境中,所以控制宇航舱的温度恒定适宜成为一个关乎宇航员生存的重要问题.简明介绍了软测量技术和RBF神经网络的原理,根据以上两个原理结合控制模拟太空舱温度的主要两个方面因素,放热源开度和控制的放热温度,对单位时间产生的热量进行软测量研究,在MATLAB平台上运用神经网络工具箱函数建立软测量模型,并进行仿真,仿真的结果得出了RBF神经网络相比BP神经网络其放热量软测量模型具有更高的准确度;因此,在模拟太空舱调控舱内温度的过程中,可以利用软测量的方法对模拟太空舱内的放热量进行动态测量,能够在温度传感器出现故障的情况下对放热参数进行估计.%The spacecraft in the space environment suffers from dramatic temperature changes frequently,so controling the temperature becomes a problem about the astronauts to survive.At first,this paper introduces the soft measurement technology and the principle of RBF neural network concisely.Then according to the two theories and combining the two aspects of temperature control modul:heat source opening and the exothermal temperature,the heat generated by the unit of time is researched by soft measurement technique.Finally,by using neural network toolbox the author establishes the soft measurement model in the MATLAB platform,and does simulation.From the simulation results we can conclude that the RBF neural network has higher accuracy than BP network in the soft measurement model.As a result,in the process of controling temperature of capsule,this paper takes the advantage of the soft measurement method to carry on the dynamic measurement for the capsule temperature.Under the condition of the temperature sensor breaking down,it can also estimate parameters of heat.
展开▼
