Industrial parks with manufacturing and service enterprises and integrated parks refer to large-scale complex buildings with a certain type of industry as the main body, with apartments, hotels, offices, production and manufacturing as supporting elements. The integrated park will provide separate power supply and heating for its buildings. By analyzing the heat and electricity characteristics of the park users, the energy characteristics of the corresponding users and the entire park will be obtained, which is of great significance to the economic operation of the park and energy conservation and emission reduction. This paper mainly uses self-encoding technology to analyze the heat and power load of the park from three dimensions: user, category (cluster analysis result) and park. Firstly, the characteristics of the individual users' electricity consumption and heat load are analyzed by self-encoding technology. The user analysis results are clustered and analyzed by the cluster analysis method to obtain several user categories. Finally, by analyzing the proportion of each category and the energy consumption characteristics, the relationship between the heat and power load of the entire integrated park and its energy consumption characteristics are obtained.