The nodes of the green computing power system are interdependent from the underlying architecture to the upper-layer application,the nodes are highly correlated,which shows the characteristics of intelligent linkage and integration.The node risk in the system has the characteristics of propagation and diffusion to the associated nodes,which poses a threat to the stable operation of the entire computing power system.The traceability and suppression of risk points and the analysis and isolation of risk links require high technical and resource costs.Therefore,an economically optimal risk regional response model for green computing power system under resource constraints is proposed.Firstly,the response strategy of power elastic compression of mixed risk nodes and the exit mechanism of overload nodes is designed to alleviate the aggregation degree of nodes in the local area of high-load system.Secondly,from the perspective of resource constraints,the regional risk evaluation index of the green computing power system with the best economy and the highest security is constructed,and the regional emergency response strategy of the green computing power system is designed,and the response model is solved by differential evolution algorithm based on the idea of group adaptive global optimization,so as to realize the suppression of hidden dangers in the system node region under the optimal economic cost.Finally,the simulation empirical analysis is carried out to verify the effectiveness of the proposed method.
关键词
绿色算力/资源约束/系统风险/经济最优/应急策略
Key words
green computing power/resource constraints/system risk/economic optimization/emergency strategy
维普
万方数据