Fault propagation analysis method fusing bogie rotating mechanical system fault mode library and causal chain explanatory structure model

Abstract

The invention provides a fault propagation analysis method fusing a rotating mechanical system fault mode library and a causal chain explanatory structure model, and belongs to the field of fault diagnosis and reliability engineering, and the method comprises the steps: constructing a fault propagation mode library, calculating a causal weight matrix, and constructing the causal chain explanatory structure model. Constructing a fault propagation mode library to extract the characteristics of the rotating mechanical system so as to reduce the error influence of physical connection characteristics on a causal weight matrix; a direct influence matrix, a forward propagation matrix and a reverse propagation matrix need to be calculated for calculating the causal weight matrix so as to help to identify a fault propagation path in the system; a causal chain explanatory structure model is constructed, a dynamic causal chain network is constructed, causal relationship strength in a fault propagation path is quantified, and a key propagation path and a feedback loop are identified, so that nonlinear interaction of multiple fault modes in the rotating machinery is more accurately modeled.

Description

Technical Field

[0001] The present invention belongs to the technical field of fault diagnosis and propagation in rotating mechanical devices, and in particular relates to a study on the propagation analysis of mechanical faults in bogies. Background Art

[0002] A fault propagation analysis method that integrates the fault mode library of bogie rotating mechanical system and the causal chain explanatory structural model is proposed. Through bidirectional rule fusion and environmental coupling modeling of the causal chain, forward propagation and reverse feedback are combined to identify the fault propagation path between components and can be applied to various industrial systems.

[0003] Currently, the Interpretive Structural Model (ISM) is a powerful tool for studying the propagation analysis of rotating machinery faults. This method analyzed the fault propagation path of a large generator set and revealed the hierarchical relationship of fault propagation between mechanical components. This research provides new ideas for the analysis of rotating machinery fault propagation. The ISM model can also be used to analyze the fault propagation laws of high-speed train power systems. The research results show that the ISM method has a good effect in revealing the relationship of fault propagation in complex systems. In rotating machinery systems, faults do not occur in isolation, but will propagate through the connection and interaction between components. ISM is mainly used to analyze the structural relationship between different faults in the system to help understand the path and hierarchy of fault propagation.

[0004] By constructing adjacency and reachability matrices, ISM can only describe the static hierarchical structure of a system (e.g., potential fault propagation paths) but cannot quantify dynamic propagation probabilities or path weights. In experiments studying rotating machinery fault propagation, ISM can indicate that "bearing wear may cause gear eccentricity," but cannot answer questions such as "What is the probability of bearing wear causing gear eccentricity" or "Which path propagates the fastest?" Furthermore, ISM can only identify key nodes (e.g., the fault source at the top of the hierarchy) but cannot quantify the global influence of a node in the propagation network. It cannot distinguish the importance of a core component that causes failures in ten subsystems from a component that only affects two subsystems. Summary of the Invention

[0005] The present invention provides a fault propagation analysis method that integrates a bogie rotating machinery system fault mode library and a causal chain explanatory structural model to overcome the limitations of the traditional ISM method and break through the limitations of the traditional ISM in complex system analysis.

[0006] The present invention relates to a fault propagation analysis method that integrates a bogie rotating machinery system fault mode library and a causal chain explanatory structural model, comprising the following steps: A. To create a fault propagation pattern library, we first need to build a fault propagation model. Based on the fault propagation model, we introduce the vibration characteristic signal of the rotating machinery to build a fault propagation pattern library, so as to reduce the error impact of the physical connection characteristics on the causal weight matrix. Collect and integrate equipment fault data from multiple channels such as operation manuals, technical documents, expert feedback, sensor records, etc., and classify and organize them according to the cause, manifestation and related variables of the fault. Establish a reusable "fault-phenomenon" correspondence set to obtain the signal data of the bogie rotating machinery, including the vibration signals collected by the sensor to jointly build the fault propagation pattern library. B. Assuming that the bogie rotating machinery system W is composed of N components, introduce three parameters in the fault propagation system to build a fault propagation model: a. Interaction frequency Fi refers to the frequency of interaction between different components in the rotating machinery system W, reflecting the degree of interdependence between them. A higher interaction frequency increases the likelihood of mutual influence between components and, consequently, the probability of fault propagation. b. Connection ratio Cc indicates the strength or reliability of the connections between components in the rotating machinery system W. A higher value indicates a tighter connection. A higher connection ratio means a greater likelihood of fault propagation between components. c. Failure probability Pf indicates the probability of a fault occurring within each component in the rotating machinery system W. For example, a value of 0.05 indicates a 5% probability of failure. v represents the degree of the node, and N is the total number of nodes. The higher the value of V, the stronger the connectivity of the system, and the greater the risk of fault propagation. Pf is the initial probability of internal failure of the component. Define the fault propagation function: The total fault propagation probability is weighted by the fault propagation function, and the total fault propagation probability can be obtained by combining them. C. Calculate the comprehensive weight matrix. The initial weight values ​​can be directly calculated by the similarity of the rotating machinery's characteristic vibration signals in the fault propagation pattern library. The forward propagation weight matrix is ​​calculated by adjusting the weights based on the physical connection strength and the layered attenuation factor. The reverse feedback weight matrix is ​​calculated by quantifying the reverse impact through differential equations and expert rules. Finally, the calculated values ​​are weighted and fused to balance data and knowledge-driven approaches.

[0007] The present invention provides a fault propagation analysis method that integrates a fault pattern library for bogie rotating machinery systems with a causal chain explanatory structural model. By combining the fault pattern library and a causal weight matrix, this method systematically analyzes the causal relationships between various fault modes in a rotating machinery system. This correlation analysis makes fault diagnosis more precise, helping to identify potential sources of equipment failure and secondary faults, thereby improving diagnostic accuracy and efficiency. Using the causal chain explanatory structural model (CCISM), the propagation paths between different fault modes can be simulated, revealing how faults propagate from one subsystem or component to another. This fault propagation analysis can proactively identify potential chain reactions, helping engineers take preventive measures before faults spread, thereby reducing the impact of faults on the system. The fault pattern library and causal chain analysis provide more scientific decision support for equipment maintenance. This method can formulate targeted maintenance and preventive measures based on the probability and propagation trends of different fault modes, reducing blind spots and improving the efficient use of maintenance resources. Through scientific fault propagation analysis, this method helps improve system reliability. By dynamically monitoring and analyzing the operating status of rotating machinery systems in real time, potential faults can be detected early and intervened promptly, thereby improving overall system reliability and safety. This technology is not only applicable to single-machine fault analysis but can also be extended to complex mechanical systems and different types of rotating machinery, demonstrating strong compatibility and adaptability. By flexibly adjusting the fault mode library and causal weight matrix, the method can address equipment maintenance needs under varying operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A flow chart of equipment fault propagation in accordance with the present invention, which is a method for analyzing fault propagation by integrating a fault pattern library of a bogie rotating machinery system with a causal chain explanatory structural model; Figure 2 A vibration signal time domain diagram, frequency domain diagram, and power spectrum density comparison diagram of a fault propagation analysis method that integrates a bogie rotating machinery system fault mode library and a causal chain explanatory structural model according to the present invention; Figure 3 This is a thermal weight matrix diagram of a fault propagation analysis method that integrates a bogie rotating machinery system fault mode library and a causal chain explanatory structural model according to the present invention; Figure 4 This is a thermal weight matrix diagram comparing the ISM and CCISM weights of a fault propagation analysis method that integrates the bogie rotating machinery system fault mode library and the causal chain explanatory structural model of the present invention. DETAILED DESCRIPTION

[0009] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0010] Please refer to Figures 1 to 4 The first embodiment of the present invention proposes a fault propagation analysis method that integrates a bogie rotating machinery system fault mode library and a causal chain explanatory structural model, including the following steps: A. To create a fault propagation pattern library, we first need to build a fault propagation model. Based on the fault propagation model, we introduce the vibration characteristic signal of the rotating machinery to build a fault propagation pattern library, so as to reduce the error impact of the physical connection characteristics on the causal weight matrix. Collect and integrate equipment fault data from multiple channels such as operation manuals, technical documents, expert feedback, sensor records, etc., and classify and organize them according to the cause, manifestation and related variables of the fault. Establish a reusable "fault-phenomenon" correspondence set to obtain the signal data of the bogie rotating machinery, including the vibration signals collected by the sensor to jointly build the fault propagation pattern library. B. Assuming that the bogie rotating machinery system W is composed of N components, introduce three parameters in the fault propagation system to build a fault propagation model: a. Interaction frequency Fi refers to the frequency of interaction between different components in the rotating machinery system W, reflecting the degree of interdependence between them. A higher interaction frequency increases the likelihood of mutual influence between components and, consequently, the probability of fault propagation. b. Connection ratio Cc indicates the strength or reliability of the connections between components in the rotating machinery system W. A higher value indicates a tighter connection. A higher connection ratio means a greater likelihood of fault propagation between components. c. Failure probability Pf indicates the probability of a fault occurring within each component in the rotating machinery system W. For example, a value of 0.05 indicates a 5% probability of failure. v represents the degree of the node, and N is the total number of nodes. The higher the value of V, the stronger the connectivity of the system, and the greater the risk of fault propagation. Pf is the initial probability of internal failure of the component. Define the fault propagation function: The total fault propagation probability is weighted by the fault propagation function, and the total fault propagation probability can be obtained by combining them. C. Calculate the comprehensive weight matrix. The initial weight values ​​can be directly calculated by the similarity of the rotating machinery's characteristic vibration signals in the fault propagation pattern library. The forward propagation weight matrix is ​​calculated by adjusting the weights based on the physical connection strength and the layered attenuation factor. The reverse feedback weight matrix is ​​calculated by quantifying the reverse impact through differential equations and expert rules. Finally, the calculated values ​​are weighted and fused to balance data and knowledge-driven approaches.

Claims

1. A fault propagation analysis method integrating a bogie rotating machinery system fault mode library and a causal chain explanatory structural model, characterized by comprising the following steps: A. Creating a fault propagation pattern library first requires building a fault propagation model. Based on this model, we incorporate rotating machinery vibration signatures to construct the fault propagation pattern library, thereby mitigating the impact of physical connection characteristics on the causal weight matrix. Equipment fault data from multiple sources, including operating manuals, technical documentation, expert feedback, and sensor data, is collected and integrated. This data is categorized and organized based on the cause, manifestation, and associated variables of the fault. A reusable "fault-phenomenon" correspondence set is established to acquire signal data from the bogie rotating machinery, including vibration signals collected by sensors, to construct the fault propagation pattern library. B. Assuming that the bogie rotating machinery system W consists of N components, three parameters are introduced into the fault propagation system to construct the fault propagation model: a. Interaction frequency Fi, which refers to the frequency of interaction between different components in the rotating machinery system W and reflects the degree of dependence between them. The higher the interaction frequency, the greater the likelihood of mutual influence between components, and the higher the probability of fault propagation. b. Connection ratio Cc, which indicates the strength or reliability of the connections between the components of the rotating machinery system W. A higher value indicates a tighter connection. A higher connection ratio means a greater likelihood of fault propagation between components. c. Failure probability Pf, which indicates the probability of a fault occurring within each component of the rotating machinery system W. For example, a value of 0.05 indicates a 5% probability of failure. v represents the degree of the node, and N is the total number of nodes. The higher the value of V, the stronger the connectivity of the system, and the greater the risk of fault propagation. Pf is the initial probability of internal failure of the component. Define the fault propagation function: The total fault propagation probability is weighted by the fault propagation function, and the total fault propagation probability can be obtained by combining them. C. Calculate the comprehensive weight matrix. The initial weight values ​​can be directly calculated by the similarity of the rotating machinery's characteristic vibration signals in the fault propagation pattern library. The forward propagation weight matrix is ​​calculated by adjusting the weights based on the physical connection strength and the layered attenuation factor. The reverse feedback weight matrix is ​​calculated by quantifying the reverse impact through differential equations and expert rules. Finally, the calculated values ​​are weighted and fused to balance data and knowledge-driven approaches.

2. A fault propagation analysis method integrating a bogie rotating machinery system fault mode library and a causal chain explanatory structural model according to claim 1, characterized in that: Constructing a fault propagation pattern library involves using the influence weights between physical connection characteristics as edges to construct a fault propagation pattern graph. In a causal chain model, the fault propagation pattern library must typically be converted into a direct influence matrix to calculate a causal weight matrix, which in turn helps identify fault propagation paths in the system.

3. The fault propagation analysis method integrating the fault mode library of the bogie rotating machinery system and the causal chain explanatory structural model according to claim 1 is characterized in that: Calculating the causal weight matrix involves extracting a representative set of multidimensional feature vectors based on the bogie rotating machinery vibration signals under typical operating conditions, as covered in a fault propagation pattern library. For each pair of fault nodes (i, j), the cosine similarity of their feature data across multiple sample dimensions is calculated to form an information structure matrix.

4. A fault propagation analysis method integrating a bogie rotating machinery system fault mode library and a causal chain explanatory structural model according to claim 1, characterized in that: The causal weight matrix of the bogie subsystem of the structure is 6*6, where the bogie system includes lubrication system, mechanical rotation system, compression system, electrical system, cooling system, and oil injection system.

Images