Simulation-based interpretation of spacecraft particle sensor measurements

粒子传感器，如朗缪尔探针或粒子成像仪，用于许多卫星和实验室等离子体实验，以推断等离子体参数，如密度和温度。在其最简单的形式中，朗缪尔探针由一个电极组成，该电极可以被偏置到可变电压，并且可以从该电极测量电流。收集的电流作为偏置电压的函数，即所谓的探针特性，然后可以基于理论或计算模型来解释，以产生局部等离子体参数。近世纪来，人们对这一问题做了大量的研究。许多理论，最近，几个计算模型已被开发来描述朗缪尔探针在等离子体中的响应。然而，在实践中，探针测量几乎普遍地被解释为能够在操作模式中产生真实的时间答案的简化分析模型。不幸的是，这些解释是出了名的不确定，误差线可以达到100%。这些大的不确定性的原因来自于使用理想化的模型，其中只有一小部分相关的物理过程被考虑在内。例如，一些模型只考虑静止未磁化等离子体中的探针，其他模型考虑等离子体流但忽略环境磁场，还有一些模型考虑磁场但忽略等离子体流或其他效应，如光电子或二次电子发射。在基本上所有的情况下，等离子体被假定为空间均匀的，附近的卫星或实验对象的存在和它们的几何形状被忽略。一个有前途的解决方案，这一困境是解释粒子传感器测量的基础上，详细的计算机模拟能够占多物理等离子体材料相互作用的特点，同时也占的几何形状中进行测量。不幸的是，由于进行这种模拟所需的时间和计算资源，这在真实的时间操作模式中是不可能的。在这项拟议的研究中将探索和开发的一个实际解决办法是为拟支持的每项实验或空间使命建立一个图书馆或数据库，利用适当的回归技术可从中推断出等离子体参数。与目前的做法相比，新的方法将导致显着改善推断等离子体参数的准确性。拟议的研究将集中于为选定的航天器和实验制作解决方案库，开发和评估不同的回归技术。这项研究的预期成果是改变范式的传感器测量的解释，这将依赖于详细的动力学模拟结果，而不是理想化的分析模型。这项建议是根据多年来在卫星环境建模方面的经验提出的。

Particle sensors such as Langmuir probes or particle imagers, are used in many satellites and laboratory plasma experiments to infer plasma parameters such as the density and temperature. In its simplest form a Langmuir probe consists of an electrode which can be biased to variable voltages, and from which the electric current can be measured. The collected current as a function of bias voltage, the so-called probe characteristic, can then be interpreted on the basis of theoretical or computational models, to yield local plasma parameters. Much work has been done on this topic over nearly a century. Many theoretical, and more recently, several computational models have been developed to describe the response of Langmuir probes in a plasma. In practice however, probe measurements are almost universally interpreted in terms of simplified analytic models capable of producing real time answers in operation mode. Unfortunately these interpretations are notoriously uncertain, with error bars that can be of order 100%. The reason for these large uncertainties comes from the use of idealised models in which only a fraction of the relevant physical processes are taken into account. For example, some models only consider a probe in a stationary unmagnetised plasma, others account for plasma flow but neglect ambient magnetic fields, yet others account for a magnetic field, but ignore plasma flow or other effects such as photoelectron or secondary electron emission. In essentially all cases, plasma is assumed to be spatially uniform, and the presence of nearby satellite or experimental objects and their geometry is ignored. A promising solution to this predicament is to interpret particle sensor measurements on the basis of detailed computer simulations capable of accounting for the multiphysics which characterises plasma-material interaction, while also accounting for the geometry in which measurements are made. This is unfortunately not possible in real time operation mode owing to the time and computational resources required to do such simulations. One practical solution to be explored and developed in this proposed research, is to construct a library or data base for each experiment or space mission to be supported, from which plasma parameters could be inferred using a suitable regression technique. Compared to current practice, the new approach would lead to significant improvements in the accuracy of inferred plasma parameters. The proposed research would concentrate on producing solution libraries for selected spacecraft and experiments, developing, and assessing different regression techniques. An expected outcome of this research is a change of paradigm in the interpretation of sensor measurements, which would then rely on detailed kinetic simulation results rather than on idealised analytic models. This proposal is based on many years of experience in satellite-environment modelling.