The Research Of Fault Propagation Characteristic Of Chemical Process AHP-SDG Model Based On Complex Network

In the chemical process system,once a variable or component has an abnormal condition,it is possible to spread,accumulate and enlarge within the system.It can cause serious fault and affect the normal operation of production.To ensure the security and stability of production,reduce the probability of fault,depth study of fault propagation characteristic is very necessary.However,with the increasing complexity and scale of chemical process,the correlation between rhe internal variables of system becomes closer,modeling and fault propagation analysis of process processes are also increasingly complex.Therefore,chemical process is combined with(Analytic Hierarchy Process)AHP-SDG(Signed Directed Graph)model in this study.AHP-SDG model is transformed into network model based on complex network theory.Then the fault propagation characteristic of chemical process system is analyzed from the perspective of network topological structure.Firstly,system SDG model is established and transformed into undirected network after dividing hierarchy structure of chemical process.According to TOPSIS(Technique for Order Preference by Similarity to an Ideal Solution)the multi-attribute decisionmaking method,the subsystems,which contain the key node,are identified by selecting multiple node importance indices,such as degree centrality,closeness centrality.The weight of each index is calculated by principal component analysis.Then the subsystem SDG model is transformed into directed network.Key nodes are found further through LeaderRank algorithm.Secondly,according to the methond of flowchart modeling,complete SDG model is constructed and transformed into network model.Considering the probability of fault propagation and network topology,fault propagation strength is defined and calculated.According to ant colony algorithm,the fault propagation path with the highest risk is found by maximum load attack,which is the weak links in production process.Finally,the result shows that the proposed method is feasible.It can not only reduce the complexity of modeling,but also can improve the comprehensive and accuracy of identification.On the other hand,it can find the weak links efficiently in production process.Thus,it can provide an important basis for fault prevention,selecting monitored variables and risk assessment,improve the security and stability of chemical process.