Research And Implementation Of Data Center Capacity Early Warning System Based On UPS Life Management

With the continuous development of the era of big data and the gradual rise of data centers,simple human management can no longer meet the needs of data center management,and the demand for intelligence and systematization is gradually increasing.The capacity management of the data center is one of the focuses of onsite operation and maintenance management,and plays a vital role in the stable operation of the data center.According to the characteristics of air-cooled data center infrastructure equipment,this paper analyzes the capacity calculation and early warning of different configuration modes,proposes fault prediction technology under different capacities,establishes a relatively complete capacity early warning architecture,and develops and implements the capacity early warning system.The main research contents of this paper are as follows:(1)Taking the domestic air-cooled data center as the research object,analyze the capacity data of the data center under different configuration modes of the ACTransformer-UPS-Array cabinet-Cabinet five-layer power supply equipment are analyzed,and add the analysis of the HVAC data,developed a capacity early warning system to meet the individualized capacity analysis and early warning requirements of different data centers;(2)According to the collected historical fault data of the data center,the life curve of the factors affecting the life of the equipment is calculated by discrete curve fitting and LM optimization algorithm,and the life characteristic model of UPS equipment is built by combining the K-means clustering algorithm and the AHPentropy method;(3)Based on the BiLSTM neural network,the optimized DBSCAN algorithm is added for data processing,combined with the UPS equipment life characteristic model,according to the multi-eigenvalue analysis,to complete the prediction of the failure time of power distribution equipment under different capacities,and in the capacity early warning system to give warnings for possible abnormal situations.