Prediction of Individual Cow Milking Time for Milking Parlor Simulation Models

Abstract

A prediction method for individual cow milking time for use in milking parlor simulation models was formulated as a function of pulsation ratio, vacuum, and milk yield per cow per milking. Milking time was divided into two components: lag time, consisting of time from milking unit attachment to initiation of milk flow, and milk flow time, consisting of time from initiation of milk flow to end of measurable milk flow. Least squares ANOVA indicated that linear effects of pulsation ratio, vacuum, and milk yield and quadratic and cubic effects of milk yield affected milk flow time and lag time. Quadratic effects of pulsation ratio and vacuum and the interaction between pulsation ratio and vacuum only affected milk flow time. Small but significant effects on milk yield per milking were due to pulsation ratio (linear), vacuum (linear and quadratic), and pulsation ratio x vacuum interaction. Based on the analysis, a prediction equation for milking time was formulated, tested, and found to be unsuitable for generation of simulated milking times. Fitting shifted gamma probability distributions to the milk flow time data at various pulsation ratios, vacuums, and milk yields per milking was suitable for generation of simulated milking times. © 1993, American Dairy Science Association. All rights reserved.