Views
3 years ago

System identification, modeling, and prediction for ... - IEEE Xplore

System identification, modeling, and prediction for ... - IEEE Xplore

System identification, modeling, and prediction for ... - IEEE

1944 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000System Identification, Modeling, and Prediction forSpace Weather EnvironmentsDimitris VassiliadisAbstract—By now nonlinear dynamical models and neural networkshave been used to predict and model a wide variety of spaceweather environments. This review starts with the physical basisfor and a brief description of the system approach. Following that,several examples illustrate practical issues in temporal and spatiotemporalprediction and modeling. The concluding remarks discussthe future developments in this research direction.Index Terms—Extraterrestrial exploration, Extraterrestrialphenomena, identification, modeling, neural network applications,nonlinear filters, space vehicle reliability.I. INTRODUCTIONTHE EFFECTS of the space environment on humans andmachines have been studied from the inception of spaceexploration. As human presence in space is in an explosivephase, it is expected that the impact of these effects will bequite significant in this and the next few solar cycles. At thesame time, the modeling effort, size of space databases, andavailability of solar and interplanetary monitors have reachedthe necessary level to make accurate forecasting possible. Thesedevelopments have prompted the establishment of nationalspace weather programs in the U.S. [21], [22] and other countries.This paper discusses the capabilities of system analysisin space weather modeling and forecasting by reviewing pastresearch and suggesting points of development for the future.Space weather disturbances are divided into two maincategories according to their physical causes: first there aredisturbances related to the high-latitude electrodynamic circuitwhich affect technological systems on the earth’s surface.Among these current systems, the most relevant ones to spaceweather are the auroral electrojets flowing in the ionosphere;other current systems, however, that complete the electriccircuit, are also important, e.g., the field-aligned and polar currents.The variation of the net magnetic field, , producedby these sources, draws secondary “geomagnetically-induced”currents in large-scale conductors such as power grid elementsand pipelines [Kappenman, this issue]. The second categoryof disturbances are caused by energetic electrons which areinjected into the inner magnetosphere, especially during magneticstorms, and subsequently trapped in the radiation beltsat altitudes of 3–5 R [2]. The intense electron flux producesmalfunctions and failures in satellites. System analysis methodshave been applied to both categories of space weather effects.Manuscript received January 19, 2000; revised May 8, 2000. This work wassupported by NASA and NSF grants.The author is with Universities Space Research Association, NASA/GoddardSpace Flight Center, Greenbelt, MD 20771 USA.Publisher Item Identifier S 0093-3813(00)11639-X.Before discussing the various methods it is useful to start withsome definitions. The majority of space environment models areempirical since they are derived from sets of observational data.Some of these will eventually be phased out and replaced byphysical models which are based on first principles; the complexityof the space environment, however, essentially guaranteesthe predominance of the empirical component. In practicespace weather applications combine the best available componentsof either approach. A prediction is the output of such aspace environment model for a time interval of activity. Morespecifically, a forecast is a prediction for a time in the future(which has not occurred yet at the time the prediction is made).As a limiting case, a model can be used to nowcast, or describethe present, rather than the future, state of the space environment.In addition, a significant part of model development andtesting is done through retrospective analyses. A forecast’s mostimportant properties are its timeliness and accuracy. A minimumlead time is crucial for operators to respond to spaceweather emergencies. For a typical prediction based on solarwind measurements made at the libration point upstream ofthe earth, the lead time is basically the solar wind propagationtime to reach earth, or 0.5–1 h depending on wind speed. Thelead time can be extended either by predicting the solar windvariations (the increase is up to 1–2 h), still largely an unexploredarea; or, in some cases, by using information directlyfrom the solar surface (several days). Once issued, the forecast isverified against observations, and it is evaluated through a comparisonwith a reference model [7]. During model development,it is standard practice to divide the observations into a data setfor training,orsystem estimation, and one or more test sets usedto measure prediction accuracy.A. Physical Basis for the Systems Approach to Space WeatherMany empirical models for space disturbances follow a systemsapproach: they represent the observed activity by a numberof variables (usually derived from the observational data) whosecoupling is described by (nonlinear) ordinary differential equations.It is important to address the question why such models,including the systems approach, are successful. First we discussthe physical basis for the systems approach, and then give itsbasic features.1) It is an observational fact for many large-scale spaceplasma environments that their activity can be reproduced bya small number of variables. (Here, by large scale, we meanscales which are much larger than particle scales such as theDebye length or the gyroradius, and significantly larger thanmost fluid scales.) Most of the available degrees of freedom inthose environments synchronize so that only a small number0093–3813/00$10.00 © 2000 IEEE

Modeling GaN: Powerful but Challenging - IEEE Xplore
China's Social Networking Problem - IEEE Xplore
Cancer Models for Understanding, Prediction, and Control
Aerodynamic Modeling and System Identification from Flight Data ...
Identification, Modeling, and Correction of Image Artifacts - MREL ...
+ System Support - Modeling, Analysis, and Prediction Program
A one-stop solution in robotic control system design ... - IEEE Xplore
Model Predictive Control System Design and Implementation Using ...
Social Computing, Behavioral-Cultural Modeling and Prediction ...
Individual Risk Predictive Modeling Divide and Conquer - Wisconsin ...
Quality of Experience of VoIP Service: A Survey of ... - IEEE Xplore
Climate Models for Decadal Prediction
utility of animal models for predicting human allergenicity
Regularised Inversion Model Predictive Uncertainty Analysis
Development of multi-metric regression models for predicting ...
Improving Mechanistic-Empirical Models for Predicting HMA Rutting
Mathematical Models in Predicting Natural Gas Consumption ...
Episode Grouping and Predictive Modeling with Health Care claims
Predictive Modeling & Credit-Based Insurance Scores
Prediction and Optimization Models for Online Display Advertising
Comparing Predictions from the CAL3QHCR and AERMOD Models ...
Improving Short-term Predictions and the Identification of ... - NASA
Measuring the Accuracy of Prediction Models - Cleveland Clinic ...
Applications to Hybrid System Identification - Johns Hopkins University
Community Modeling Initiative - Modeling, Analysis, and Prediction ...
Assessment and Future Directions of Nonlinear Model Predictive ...
risk assessment within an integrated multi model oil spill prediction ...
Building energy system identification and control
design, modeling, identification and control of mechatronic ... - Tesis
Descriptive and Predictive models of dengue epidemiology: an ...