Fire Warning for Prefabricated Battery Compartments in Energy Storage Power Stations Based on Stacking Multi-Model Fusion
Due to the independent composition of multiple sensing devices in the temperature monitoring of prefabricated battery compartments in energy storage power stations,the sensitivity of fire warning for high temperature anomalies is low.Therefore,a research on fire warning method for prefabricated battery compartments in energy storage power stations based on Stacking multi-model fusion is proposed.Introduced Stacking ensemble learning method to fuse temperature data of prefabricated battery compartments in multi-dimensional energy storage power stations,and used regularization coefficients to correct the deviation of the fused data.In the fire warning stage,personalized warning temperatures are set based on the development trend of temperature parameters and the time scale requirements of fire warning.Based on the relationship between the fusion result and the warning temperature,it is determined whether to issue a warning.In the test results,the design method achieved effective early warning for different degrees of abnormal temperature and had good sensitivity.
Stacking multi-model fusionfire warningregularization coefficientdeviation correctionwarning temperature
万方数据
维普
