Identification of Unknown Abnormal Conditions in Catalytic Cracking Process Based on Two-Step Clustering Analysis and Signed Directed Graph

There are many unknown abnormal working conditions in industrial production. It is difficult to identify unknown abnormal working conditions because there are few relative sample and experience in this field. To solve this problem, a new identification method combining two-step clustering analysis and signed directed graph (TSCA-SDG) is proposed. Firstly, through correlation analysis and R-type clustering analysis, the variables are effectively selected and extracted. Then, a two-step clustering analysis was carried out on the selected variables to obtain the cluster results. Through the establishment of the signed directed graph (SDG) model, the causes of abnormal working conditions and their mutual influence are deduced from the mechanism. The application of the TSCA-SDG method in the catalytic cracking process shows that this method has good performance for abnormal condition identification.

A Framework for Monitoring and Fault Diagnosis in Nuclear Power Plants Based on Signed Directed Graph Methods

When nuclear power plants (NPPs) are in a state of failure, they may release radioactive material into the environment. The safety of NPPs must thus be maintained at a high standard. Online monitoring and fault detection and diagnosis (FDD) are important in helping NPP operators understand the state of the system and provide online guidance in a timely manner. Here, to mitigate the shortcomings of process monitoring in NPPs, five-level threshold, qualitative trend analysis (QTA), and signed directed graph (SDG) inference are combined to improve the veracity and sensitivity of process monitoring and FDD. First, a three-level threshold is used for process monitoring to ensure the accuracy of an alarm signal, and candidate faults are determined based on SDG backward inference from the alarm parameters. According to the candidate faults, SDG forward inference is applied to obtain candidate parameters. Second, a five-level threshold and QTA are combined to determine the qualitative trend of candidate parameters to be utilized for FDD. Finally, real faults are identified by SDG forward inference on the basis of alarm parameters and the qualitative trend of candidate parameters. To verify the validity of the method, we have conducted simulation experiments, which comprise loss of coolant accident, steam generator tube rupture, loss of feed water, main steam line break, and station black-out. This case study shows that the proposed method is superior to the conventional SDG method and can diagnose faults more quickly and accurately.

A Fault Identification Method in Distillation Process Based on Dynamic Mechanism Analysis and Signed Directed Graph

Due to the complexity of materials and energy cycles, the distillation system has numerous working conditions difficult to troubleshoot in time. To address the problem, a novel DMA-SDG fault identification method that combines dynamic mechanism analysis based on process simulation and signed directed graph is proposed for the distillation process. Firstly, dynamic simulation is employed to build a mechanism model to provide the potential relationships between variables. Secondly, sensitivity analysis and dynamic mechanism analysis in process simulation are introduced to the SDG model to improve the completeness of this model based on expert knowledge. Finally, a quantitative analysis based on complex network theory is used to select the most important nodes in SDG model for identifying the severe malfunctions. The application of DMA-SDG method in a benzene-toluene-xylene (BTX) hydrogenation prefractionation system shows sound fault identification performance.

Research on Fault Diagnosis Based on Hierarchical Signed Directed Graph for Nuclear Power Plants

When nuclear power plants (NPPs) are in failure state, it may release radioactive substance into the environment. Thus, the safety of NPPs is put forward a high standard. Fault detection and diagnosis (FDD) are significant for NPPs to help operator timely know the state of system and provide the online guidance. Fault diagnosis can improve the safety of nuclear power plants, but current fault diagnosis methods pay too much attention to accuracy of diagnostic results. As a complex industrial system, how to explain the causes of faults in NPPs becomes more important. Although there are many studies on the knowledge graph, there is no detailed description on the failure process (consider timing). This paper proposed a three-layer structure for FDD in NPPs. Each layer represents the stage of the accident, it can give the operator a clear cognitive process to faults. The three-layer structure includes “smooth layer”, “threshold layer” and “fault layer”. The three layers indicate the reason of faults, the response of the parameters at each stage, and clearly showing the accident process. The smooth layer uses the stability analysis to analyze whether the current NPP is operating abnormally; the threshold layer uses the thresholds of the NPP to monitor which parameters have exceeded the upper limit or the lower limit; the fault layer reflects what is happening in the current operation and accidents are explained using signed directed graph. This paper takes the loss of coolant accident as an example, three-layer structure is analyzed, which shows the feasibility of the method. The case shows that the proposed method is superior to the conventional SDG method, can diagnose the faults, and give the reason of diagnosis results.

A Fault Diagnosis Method Based on Signed Directed Graph and Correlation Analysis for Nuclear Power Plants

Fault detection and diagnosis (FDD) provides safety alarms and diagnostic functions for a nuclear power plant (NPP), which comprises large and complex systems. NPP has a large number of parameters which make it difficult achieve FDD. Now many diagnosis methods have lack of better explanation for faults and quantitative analysis. Therefore, to overcome the “black box” of FDD based on data-driven methods, this paper adopts signed directed graph (SDG) in knowledge graph for FDD. It can express the cause and effect of accidents through knowledge maps. At same time, this paper uses correlation analysis to conduct a quantitative analysis between parameters and faults. It this paper, SDG is used to explain the reason of faults. In order to quickly achieve FDD, this paper introduces a quantitative analysis method. It combines expert system and correlation analysis method to analyze the weight of each parameter. On this basis, matrix reasoning is used to achieve the FDD, and the reason is shown in SDG model inference. This paper takes loss of coolant accident as the case study, the case shows that the proposed method is superior to the conventional SDG method and can diagnose the faults timely.

From graph topology to ODE models for gene regulatory networks

A gene regulatory network can be described at a high level by a directed graph with signed edges, and at a more detailed level by a system of ordinary differential equations (ODEs). The former qualitatively models the causal regulatory interactions between ordered pairs of genes, while the latter quantitatively models the time-varying concentrations of mRNA and proteins. This paper clarifies the connection between the two types of models.We propose a property, called the constant sign property, for a general class of ODE models. The constant sign property characterizes the set of conditions (system parameters, external signals, or internal states) under which an ODE model is consistent with a signed, directed graph. If the constant sign property for an ODE model holds globally for all conditions, then the ODE model has a single signed, directed graph. If the constant sign property for an ODE model only holds locally, which may be more typical, then the ODE model corresponds to different graphs under different sets of conditions. In addition, two versions of constant sign property are given and a relationship between them is proved.As an example, the ODE models that capture the effect of cis-regulatory elements involving protein complex binding, based on the model in the GeneNetWeaver source code, are described in detail and shown to satisfy the global constant sign property with a unique consistent gene regulatory graph. Even a single gene regulatory graph is shown to have many ODE models of GeneNetWeaver type consistent with it due to combinatorial complexity and continuous parameters.Finally the question of how closely data generated by one ODE model can be fit by another ODE model is explored. It is observed that the fit is better if the two models come from the same graph.