Increased precision in sampling using regression modeling, with an application to electric load research

Thesis/Dissertation · Fri Jan 01 04:00:00 EST 1988

OSTI ID:5223878

Oberg, K M

A model is given for situations in survey sampling in which the characteristic of interest is an expected value of the dependent variable in a regression. For each sample unit, a regression can be used to estimate the expected value of the characteristic of interest for a given set of values of the explanatory variables. The model can be used to calculate the expected value and variance of an estimator of the population total of the expected value of the characteristic of interest, for a given set of values of the explanatory variables. The application involves the estimation of a class-load curve on the system peak day of an electric utility. The conventional method uses, for each customer in the sample, the customer's actual demand on the system peak day to estimate the customer's expected demand under the conditions of the peak day. The proposed method uses, for each customer in the sample, a model to estimate the customer's expected demand under the conditions of the peak day. The conditions are variables such as the time-of-day and weather. The variance of an estimator of a class expected load curve under the conditions of the peak day may be reduced by using the proposed method instead of the conventional method.

• Research Organization: Michigan Univ., Ann Arbor, MI (USA)
• OSTI ID: 5223878
• Country of Publication: United States
• Language: English

Related Subjects

29 ENERGY PLANNING, POLICY, AND ECONOMY
296000* -- Energy Planning & Policy-- Electric Power
ELECTRIC POWER
LOAD ANALYSIS
MATHEMATICAL MODELS
MATHEMATICS
PEAK LOAD
POWER
PRICES
REGRESSION ANALYSIS
STATISTICS
TIME-OF-USE PRICING
WEATHER