Neural Network Adaptive Contribution Evaluation and Prediction Model of Energy Intensity

Development of low carbon environmental protection energy becomes requirements, peak in carbon, carbon neutral requirement is raised, extensive international track evaluation index, to evaluate a region cut carbon emission reductions, energy intensity in the whole society as a comprehensive key indicators, both associated with control of the total energy consumption, also related to the benefit of the unit of energy output. This paper proposes a adaptive neural network computing, from the perspective of the connotation of decomposition by multiple elements of social energy intensity contribution model, can be decomposed contribution of various elements on the current energy consumption and to quantify the energy consumption intensity influence key factor prediction, calculation under a given energy intensity target, social economy, efficiency, structure and matching index of growth. Solve the blindness of energy intensity prediction and the matching of policy objectives.