train_weibull_model: Training function for Weibull model

In CNAIM: Common Network Asset Indices Methodology (CNAIM)

Training function for Weibull model

Description

This function uses transformer fault statistics data to train a Weibull model: Based on the environmental factors determining a transformer's expected lifetime, the set of all data points is first partitioned into five parts. Then a multilinear estimate for the expected lifetime of a transformer is trained for each part separately, and the corresponding Weibull shape and scale parameters for the five parts are estimated. The function returns the shape and scale parameters needed for the function predict_weibull_model().

Usage

train_weibull_model(transformer_faults_data)

Arguments

transformer_faults_data

Data frame. Contains past data on transformer faults, together with environmental factors. Must contain the following fields: utilisation_pct: Numeric or "Default", placement: "Indoor", "Outdoor" or "Default", altitude_m: Numeric or "Default", distance_from_coast_km: Numeric or "Default", corrosion_category_index: Numeric or "Default", partial_discharge: "Low", "Medium", "High (Not Confirmed)", "High (Confirmed)" or "Default", oil_acidity: Numeric or "Default", temperature_reading: "Normal", "Moderately High", "Very High" or "Default", observed_condition: "No deterioration", "Superficial/minor deterioration", "Slight Deterioration", "Some deterioration", "Substantial deterioration" or "Default" age: Numeric

Value

Data frame. All shape and scale parameters needed for the function predict_weibull_model().

Source

https://www.cnaim.io/docs/fault-analysis/

Examples

train_weibull_model(transformer_faults_data = transformer_11kv_faults)

CNAIM documentation built on Aug. 31, 2022, 9:13 a.m.