Energy Efficiency Management And Energy Consumption Prediction Of Public Buildings Based On Data Mining

In recent years,with the improvement of the monitoring quality of public buildings energy efficiency monitoring systems and the popularity of building automation systems,the building operation monitoring platform has accumulated massive real-time energy consumption data,these data often contain the building energy-saving breakthrough direction.Moreover,the building energy consumption data has the characteristics of large volume of data and high dimension,which makes conventional analysis method is difficult to discover and summarize the correlation between the energy consumption and the energy consumption data hidden in the energy consumption data.The data mining technology has the ability to deal with massive data and discover potential,novel and useful knowledge.The data mining technology provides a way for people to know the information and knowledge hidden in the data.In this paper,the data mining technology is used to analyze the energy consumption of public buildings in order to improve the energy efficiency of public buildings and the rationality of energy conservation strategies.The main contents and methods are as follows:1.Research on the method of filling the missing values of building energy consumption data.The k-means filling method based on Mahalanobis distance is used to fill the missing values of public buildings energy consumption data.In order to reduce the dimension of energy consumption data and use the intrinsic information contained in the attribute set,correlation analysis is used to find the attributes that are closely related to missing attributes,and uses the information entropy to compute the coefficient matrix of the nearest neighborhood to quantify the influence of non missing data attributes on missing data attributes.The experimental results show that the algorithm can effectively use the intrinsic relationship between energy consumption data attributes and effectively fill the building energy consumption missing data.2.The LODCD algorithm is proposed which can be used to detect abnormal data of building energy consumption.To solve the problem that the traditional outlier detection algorithm has high time complexity and poor parameter robustness,the local outlier optimized detection algorithm(LODCD)based on clustering and density was proposed on the basis of improved COF.The algorithm used clustering method to handled the original data set for selected the candidate outlier datasets and reducing the time complexity of the algorithm.Meanwhile,in order to improve the detection accuracy of the algorithm,the information entropy is introduced to determine the outlier attribute of the object when calculating the distance between data objects.After determining the outlier attribute of the datasets,using the new outlier factor LCOF to measured the outlier degree of the data in the candidate outlier datasets.The algorithm reduces the time complexity and the detection accuracy of the parameter dependence in the premise of ensuring the accuracy.The simulation results show the effectiveness and feasibility of this method.3.The building energy consumption prediction model is constructed based on decision tree algorithm.C4.5 decision tree algorithm is used to deal with the energy consumption data of public buildings,and an easy-to-understand forecast model of public building energy consumption is established,and use C4.5 algorithm to select the main factors that affect the energy consumption of the building for construct the public building Energy consumption regression analysis model.Moreover,in order to achieve the purpose of improving the speed of the algorithm to scan the data set,the principal component analysis is used to deal with the original attributes,and the dimension of building energy consumption is reduced by extracting the principal components of the original attributes.The experimental results show that the model can predict the building energy consumption,and provide a theoretical basis for building energy efficiency.4.The static and dynamic association rules mining is used to analyze the energy consumption data of commercial buildings,and analyze the feasibility of mining association results in public building energy efficiency management through the modeling.The Apriori algorithm is used to analyze the inherent correlation and interlocking relationship between the energy consumption data,meteorological parameters and personnel activities of public buildings from static and dynamic aspects,and use the professional knowledge to interpret the strong association rules to find a key entry point for building energy efficiency.SketchUp and OpenStudio are used to simulate the annual energy consumption of commercial buildings to verify the impact of mining association rules on building energy efficiency.The experimental results show that the method of public building energy consumption data mining processing is proposed in this paper has the characteristics of universality and real-time,which can be effectively applied to the prediction of building energy consumption and guide the improvement of the overall energy efficiency of the building.