Temperature, Humidity and Energy Consumption Forecasting in the Poultry Hall Using Artificial Neural Networknetwork

ماشین های کشاورزی

Article 19, Volume 7, Issue 2 - Serial Number 14, 2017, Pages 546-557   

Document Type: Short Paper

DOI: 10.22067/jam.v7i2.50799

Authors

N. Gholamrezaei^{* 1}; K. Qaderi²; K. Jafari Naeimi¹

¹Department of Biosystems Engineering, Faculty of Agriculture, Shahid Bahonar University of Kerman, Iran

²Department of Water Science and Engineering, Faculty of Agriculture, Shahid Bahonar University of Kerman, Iran

Abstract

Introduction
Energy consumption management is one of the most important issues in poultry halls management. Considering the situation of poultry as one of the largest and most developed industries, it is needed to control growing condition based on world standards. The neural networks as one of the intelligent methods are applied in a lot of fields such as classification, pattern recognition, prediction and modeling of processes. To detect and classify several agricultural crops, a research was conducted based on texture and color feature. The highest classification accuracy for vegetables, grains and fruits with using artificial neural network were 80%, 86% and 70%. In this research, the ability to Multilayer Perceptron (MLP) Neural Network in predicting energy consumption, temperature and humidity in different coordinate placement of electronic control unit sensors in the poultry house environment was examined.
Materials and Methods
The experiments were conducted in a poultry unit (3000 pieces) that is located in Fars province, Marvdasht city, Ramjerd town, with dimensions of 32 meters long, 7 meters wide and 2.2 meters height. To determine the appropriate placement of the sensor, 60 different points in terms of length, width and height in poultry were selected. Initially, the data was divided into two datasets. 80 percent of total data as a training set and 20 percent of total data as a test set. From180 observations, 144 data were used to train network and 36 data were used to test the process. There are several criteria for evaluating predictive models that they are mainly based according to the difference between the predicted outputs and actual outputs. To evaluate the performance of the model, two statistical indexes, mean squared error (MSE) and the coefficient of determination (R²) were used.
Results and Discussions
In this study, to train artificial neural network for predicting the temperature, humidity and energy consumption, the trainlm algorithm (Levenberg-Marquardt) was used. To simulate temperature, humidity and energy consumption, networks were trained with two and three layers, respectively. Network with two layers with10 neurons in the hidden layer and one neuron in the output layer with (R²) equal to 0.96 and (MSE) equal to 0.00912, was given the best result for predicting temperature. For humidity electronic sensors, results showed that network with three layers with the 10 neurons in the first hidden layer, 20 neurons in the second hidden layer and one neuron in the output layer with (R²) equal to 0.8 and (MSE) equal to 0.00783 was the best for predicting humidity. Finally, network with two layers with 10 neurons in the first hidden layer, 10 neurons in the second hidden layer and one neuron in the output layer was selected as the optimal structure for predicting energy consumption. For this topology, (R²) and MSE were determined to 0.98 and 0.00114, respectively. Linear and multivariate regression for the parameters affecting temperature, humidity and energy consumption of electronic sensors was determined by the STATGR software. Correlation coefficients indicated that parameters such as length, height and width of the electronic control sensors placed in the poultry hall justified 82% of the temperature changes, 61% of the humidity changes and 92% of the energy consumption changes. Therefore, comparing with correlation coefficients obtained from the neural network models, the highest correlation coefficient was related to energy parameter and the lowest correlation was linked to humidity parameter.
Conclusion
The results of the study indicated the high performance for predicting temperature, humidity and energy consumption. The networks hadthree inputs including length, width and height of electronic sensor positions and an output for temperature, humidity and energy consumption. For training networks the multiple layer perceptron (MLP) with error back propagation learning algorithm (BP) was used. Functions activity for all networks in hidden layers were tangentsigmoid and in the output layer, linear (purelin). Comparing the results of artificial neural network and logistic regression model showed that artificial neural network model with correlation coefficients of 0.98 (energy), 0.96 (temperature) and 0.8 (humidity) provided closer data to the actual data compared with regression models with correlation coefficients of 0.92, 0.82 and 0.61 for the energy, temperature and humidity respectively.

Keywords

Artificial neural network; Electronic control; energy; Modeling; Temperature

References

1. Adamchuk, V. I., A. V. Skotnikov, J. D. Speichinger, and M. F. Kocher. 2004. Development of an instrumented deep-tillage implement for sensing of soil mechanical resistance. Trans. ASAE. 47: 1913-191.

2. Anami, B. S., J. D. Pujari, and R. Yakkhundimath. 2011. Identification and classification of normal and affected agriculture/horticulture produce based on combined color and texture feature extraction. International Journal of Computer Applications in Engineering Sciences 1: 356-360.

3. Arjona, R., P. Ollero, and F. Vidal. 2005. Automation of an olive waste industrial rotary dryer. Journal of Food Engineering 68: 239-242.

4. Azadeh, A., S. F. Ghaderi, and S. Sohrabkhani. 2006. Forecasting electrical consumption by integration of Neural Network, time series and ANOVA.

5. Ishikawa, S., and V. Gulick. 2013. An automated mineral classifier using Raman spectra. Computers and Geosciences 54: 259-268.

6. Jacovides, C. P. 1997. Reply to comment on Statistical procedures for the evaluation of evapotranspiration models. Agricultural Water Management 3: 95-97.

7. Kaul, M., R. L. Hill, and C. Walthall. 2005. Artificial neural networks for corn and soybean yield prediction. Agriculture system 85: 1-18.

8. Keshavarz, J. 2014. Design, fabrication and evaluation of the electronic control system of temperature and humidity, in the industrial poultry farm. M. Sc. Thesis, Faculty of Agricultural, Bahonar University. (In Farsi).

9. Moallem, P., and A. Monajemi. 2007. A heuristic criterion for goodness of multi layer perceptron as a classifier. First Data Mining Conference. Amir Kabir University, Tehran, Iran. (In Farsi).

10. Rashed Mohassel, M. H., H. Najafi, and M. D. Akbarzadeh. 2001. Weed Biology and Control. Ferdowsi University Press, 404p. (In Farsi).

11. Rizzoni, G. 2000. Principels and Applications of Electrical Engineering, 3td ed., McGraw-Hill, USA.

12. Sigari, M. H., H. Sigari, and N. Mozayani. 2012. Estimated time of drying food using computer vision and artificial neural network. Fifth National Conference on Agricultural Machinery Engineering and Mechanisation. Ferdowsi University, Mashhad, Iran. (In Farsi).

13. Simmons, J. D., and B. D. Loit. 1993. Automatic Fan Control, Agricultural Research Service, South Central Poultry Research Laboratory, Mississipi Stat. pp. 323-2230.

14. Storey, N., 1998. Electronics: A system approach, Prentice Hall, Harlow.

15. Torrecilla, J. S., L. Otero, and P. D. Sanz. 2004. A neural network approach for thermal/pressure food processing. Food Engineering 62: 89-95.

16. Vakil-Baghmisheh, M. T. 2002. Farsi Character Recognition Using Artificial Neural Networks. PhD Thesis, Faculty of Electrical Engineering, University of Ljubljana.

17. Visen, N. S., D. S. Jayas, J. Paliwal, and N. D. G. White. 2004. Comparison of two neural network architectures for classification of singulated cereal grains. Journal of Canadian Biosystem Engineering 46: 7-14.

18. Zhang, Q., S. Y. Yang, G. S. Mittal, and S. Yi. 2002. Prediction of performance indices and optimal parameters of rough rice drying using neural network. Biosystems Engineering 83 (3): 281-290.

19. Zhang, Y. F., and J. Y. H. Fun. 1998. A neural network approach for early cost estimation of packaging products. Computers & Industrial Engineering 34: 433-50.