Using Artificial Intelligence Methods to Predict Doses from Large Solar Particle Events in Space Department of Nuclear Engineering, The University of Tennessee

Format:

Conference/Event

Author/Creator:

Nichols, Theodore F., author.

Conference Name:

International Conference On Environmental Systems (2004-07-19 : Colorado Springs, Colorado, United States)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2004

Summary:

When planning space missions, radiation effects due to large solar particle events (SPEs) can become a major concern since doses can become mission threatening to both the crew and the spacecraft electronic components. As mission duration increases, the possibility that a significant dose is delivered also increases, especially during the more active parts of the solar cycle. Therefore, a method of predicting when certain limiting doses will be reached following the onset of a large SPE needs to be available. Typical dose versus time profiles of a SPE can be represented by a Weibull functional form, which is comprised of three unknown parameters. Since these dose-time profiles are nonlinear functions, the use of artificial neural networks as the forecasting mechanism is ideal. In this work we report on the status of development of a "nowcast" methodology that utilizes a set of artificial neural networks that can forecast profiles of dose versus time since event onset using dose and dose rate information obtained early on as the event begins

Notes:

Vendor supplied data

Publisher Number:

2004-01-2324

Access Restriction:

Restricted for use by site license