Estimation of Correlation Functions by Stochastic Approximation

Two techniques for estimating the correlation function of a stationary random process using the method of stochastic approximation are developed. Both provide updated estimates as successive T-second observations are processed. Both assume a functional form for the correlation function that depends on a number of parameters that are estimated. One technique is based on the mean-square error; the other maximizes a likelihood function. Examples are included.     

Using high-energy proton fluence to improve risk prediction for consequences of solar particle events

The potential for exposure to large solar particle events (SPEs) with high energy levels is a major concern during interplanetary transfer and extra-vehicular activities (EVAs) on the lunar and Mars surface. Previously, we have used data from the last 5 solar cycles to estimate percentiles of dose to a typical blood-forming organ (BFO) for a hypothetical astronaut in a nominally shielded spacecraft during a 120-d lunar mission. As part of this process, we made use of complete energy spectra for 34 large historical SPEs to calculate what the BFO mGy-Eq dose would have been in the above lunar scenario for each SPE. From these calculated doses, we then developed a prediction model for BFO dose based solely on an assumed value of integrated fluence above 30 MeV (Φ₃₀) for an otherwise unspecified future SPE. In this study, we reasoned that since BFO dose is determined more by protons with higher energies than by those with lower energies, more accurate BFO dose prediction models could be developed using integrated fluence above 60 (Φ₆₀) and above 100 MeV (Φ₁₀₀) as predictors instead of Φ₃₀. However to calculate the unconditional probability of a BFO dose exceeding a pre-specified limit (“BFO dose risk”), one must also take into account the distribution of the predictor (Φ_30,Φ₆₀, or Φ₁₀₀), as estimated from historical SPEs. But Φ₆₀ and Φ₁₀₀ have more variability, and less available historical information on which to estimate their distributions over many SPE occurrences, than does Φ₃₀. Therefore, when estimating BFO dose risk there is a tradeoff between increased BFO dose prediction at a given energy threshold and decreased accuracy of models for describing the distribution of that threshold over future SPEs as the threshold increases. Even when taking the second of these two factors into account, we still arrived at the conclusion that overall prediction improves as the energy level threshold increases from 30 to 60 to 100 MeV. These results can be applied to the development of approaches to improve radiation protection of astronauts and the optimization of mission planning for future space missions.     

Optimum Number of Stages in a Staged Vehicle

The problem of optimum partitioning of a staged vehicle (SV) and determination of the minimal number of stages to provide for a maximum velocity in the vertical SV motion in a homogeneous gravity field is considered at a given value of payload mass.