دانلود رایگان مقاله انگلیسی مدل شبکه بیزی برای پیش بینی بار خنک کننده ساختمان های تجاری - اشپرینگر 2018

عنوان فارسی
مدل شبکه بیزی برای پیش بینی بار خنک کننده ساختمان های تجاری
عنوان انگلیسی
A Bayesian Network model for predicting cooling load of commercial buildings
صفحات مقاله فارسی
0
صفحات مقاله انگلیسی
15
سال انتشار
2018
نشریه
اشپرینگر - Springer
فرمت مقاله انگلیسی
PDF
نوع مقاله
ISI
نوع نگارش
مقالات پژوهشی (تحقیقاتی)
رفرنس
دارد
پایگاه
اسکوپوس
کد محصول
E8926
رشته های مرتبط با این مقاله
مهندسی کامپیوتر، فناوری اطلاعات
گرایش های مرتبط با این مقاله
هوش مصنوعی، شبکه های کامپیوتری
مجله
شبیه سازی ساختمان - Building Simulation
دانشگاه
Electricity Infrastructure and Buildings Division - Pacific Northwest National Laboratory - Battelle Boulevard - USA
کلمات کلیدی
مدل شبکه بیزی، پیش بینی بار خنک کننده، مجموعه داده های آموزشی، عدم اطمینان
doi یا شناسه دیجیتال
https://doi.org/10.1007/s12273-017-0382-z
چکیده

Abstract


Cooling load prediction is indispensable to many building energy saving strategies. In this paper, we proposed a new method for predicting the cooling load of commercial buildings. The proposed approach employs a Bayesian Network model to relate the cooling load to outdoor weather conditions and internal building activities. The proposed method is computationally efficient and implementable for use in real buildings, as it does not involve sophisticated mathematical theories. In this paper, we described the proposed method and demonstrated its use via a case study. In this case study, we considered three candidate models for cooling load prediction and they are the proposed Bayesian Network model, a Support Vector Machine model, and an Artificial Neural Network model. We trained the three models with fourteen different training data datasets, each of which had varying amounts and quality of data that were sampled on-site. The prediction results for a testing week shows that the Bayesian Network model achieves similar accuracy as the Support Vector Machine model but better accuracy than the Artificial Neural Network model. Notable in this comparison is that the training process of the Bayesian Network model is fifty-eight times faster than that of the Artificial Neural Network model. The results also suggest that all three models will have much larger prediction deviations if the testing data points are not covered by the training dataset for the studied case (The maximum absolute deviation of the predictions that are not covered by the training dataset can be up to seven times larger than that of the predictions covered by the training dataset). In addition, we also found the uncertainties in the weather forecast significantly affected the accuracy of the cooling load prediction for the studied case and the Support Vector Machine model was more sensitive to those uncertainties than the other two models.

نتیجه گیری

4 Conclusion


This paper proposes a Bayesian Network model for predicting the cooling load of a commercial building. We show that the proposed Bayesian Network model has the potential of achieving similar or better performance than a Support Vector Machine model or an Artificial Neural Network model. In the case study, the Bayesian Network model used the lowest CPU time for training when the amount of the training data is more than ten weeks. The CPU time cost by the Artificial Neural Network model is higher than that of the Bayesian Network model by up to 5700%. Moreover, using the Bayesian Network model does not require background in sophisticated mathematical theories. These benefits suggest that the Bayesian Network model is promising for real-world applications.


In this paper, we also explore the relationship between performance of the candidate prediction models and the amount of data available to train the models. We found all the three models can’t extrapolate the training dataset. For the studied case, the three models tend to have much larger prediction deviation if the testing data point lies far distance from the training dataset. On the other hand, we also noted that increasing the amount of the training data, but not the percentage of the testing data points that are covered by the training dataset, doesn’t benefit the prediction a lot. Based on the above statement, we suggest to increase the percentage of the testing data points that are covered by the training dataset, rather than only the amount of data in the training dataset, if the three models are employed for prediction.


بدون دیدگاه