Abstract：Compared to partial life cycle assessment，the full life cycle assessment is more complex and

uncertain，with the difficulty of comprehensively considering carbon emissions and their multiple influencing

factors at each stage，resulting in fuzzy assessment boundaries and dynamic repetition of the process. To this end，it is necessary to set clear system boundaries and dynamically adjust them according to reality. Therefore，an

improved algorithm for assessing carbon emissions throughout the full life cycle of wind storage power systems

was proposed. On the basis of determining the system boundary for the full life cycle carbon emission assessment

of wind storage power systems，carbon emission factors were considered for carbon emission calculation，and

carbon emission data from various stages were comprehensively considered to establish an evaluation function.

This function can be used for backpropagation of BP evaluation algorithm，which automatically adjusts the network

structure and weights to clarify the evaluation boundary and improve the accuracy of prediction results. The

experimental results show that the proposed improved algorithm can control the error of carbon emission

assessment results within 3 kg，and can accurately conduct the full life cycle carbon emission assessment of wind

storage power systems.

      Key words：including wind storage power system；full life cycle；carbon emission assessment；carbon emission factor；BP neural network