ARIMA Modeling of Animal Zone Temperature in Weaned Piglet Buildings: Design of the Model

Abstract

Abstract. Predictive models provide an efficient tool for improving environmental control in livestock buildings. In this article, a robust and accurate ARIMA model for forecasting temperature inside a building for weaned piglets in the range 6 to 20 kg live weight was built. The candidate models presented in this article predict 10 min values during a complete production cycle, which makes them suitable as predictive models for improving control strategies. The accuracy of the base model, which used outdoor temperature as a predictor variable, can be improved by appropriately replacing the outliers in the time series. Because accuracy increases with the increase in the number of predictor variables, the model that used four variables (temperature at the air outlet, area of the air outlet through the fan, volume of air extracted, and animal live weight) provided the best results, with a maximum absolute error of 0.840°C, a root mean square error of 0.204°C, and random residuals according to the Ljung-Box statistic. This model used only the values of the last 20 min for the forecast, which suggests low thermal inertia in the animal zone. In addition, the model includes predictor variables that are representative of outdoor conditions, operation of the systems, and animal health status. Keywords: ARIMA, Forecast, Model, Piglet, Temperature.

Validation of an AutoRegressive Integrated Moving Average model for the prediction of animal zone temperature in a weaned piglet building

“…Integration and Moving Average. The ARIMA model investigated in this paper was designed by Besteiro et al (2017). The model allowed for the prediction of 10-min values of animal zone temperature for a complete production cycle of approximately 6 weeks.…”

“…The most usual models of pig growth show an underlying exponential fit (Wellock et al, 2003;Vincek et al, 2012). Exponential fit was used for the three weighings performed at three different times: when the animals entered the room, when Besteiro et al (2017), Model 4.4 was chosen because it showed a level of significance of 0.055, which allowed for the assumption of the randomness of its residuals at a 95% confidence level. This model was the only model that described the evolution of the variable, insofar as it produced a series of residuals with a structure that was similar to that of Gaussian white noise.…”

“…Temperatures were measured using an S-THB-M002 sensor (Onset Computer Corporation©, Bourne, MA, USA). Speed of the air extracted through the ventilation system was measured using a HD2903TTC310 (Delta Ohm, Caselle, PD, Italy) active air speed transmitter (Besteiro et al, 2017). Volume of extracted air was calculated from air velocity and fan section according to the method proposed by Hinz and Linke (1998).…”

“…Overall, the literature reveals a lack of validated models for the prediction of environmental variables in livestock buildings. Many models have been developed but not validated at all (Diesch & Froehlich, 1988;Axaopoulos et al, 1992;Lambert et al, 2001;Besteiro et al, 2017). Other models have been validated over short periods, which does not ensure the robustness of the model or the incorporation of the effects of animal growth.…”

“…Average (ARIMA) model for the prediction of indoor temperature in the animal occupied zone of a conventional pig farm. The model, designed by Besteiro et al (2017), predicts temperature inside a building for weaned piglets in the range 6 to 20 kg live body mass.…”