Failure chain construction method and system for complex electromechanical systems

Abstract

The present invention provides a method and system for constructing a failure chain of a complex electromechanical system, which belongs to the technical field of electromechanical system operation and maintenance, and constructs a topological network model of a complex electromechanical system; lists risk points based on the minimum maintainable and computable unit, and expresses the state of the risk points based on a mathematical characterization method; calculates and measures the correlation coefficients between the state parameters of the risk points, and constructs a direct influence matrix between the risk points; calculates the global influence matrix between the risk points, and obtains the quantitative degree of dependence between the risk points; predicts the failure sequence of the complex electromechanical system, and constructs a failure propagation chain of the complex electromechanical system. The present invention takes into account the factors that the failure propagation of the internal components of the complex electromechanical system leads to the overall failure of the complex electromechanical system, and provides a method and basis for the early prediction and later maintenance of the failure of the complex electromechanical system; deeply analyzes the interaction mechanism between components, and reveals the influence of local failure on system failure.

Description

Technical Field

[0001] The present invention relates to the technical field of electromechanical system operation and maintenance, and in particular to a method and system for constructing a failure chain of a complex electromechanical system taking dependencies and uncertainties into consideration. Background Art

[0002] The components of complex mechatronic systems are interconnected through mechanical, electrical and information types to achieve different functions of "component-subsystem-system", and the degree of interaction and dependency between components are different. However, the existing system safety analysis theory ignores the coupling strength between components and cannot accurately describe the functional behavior determined by the constitutive performance of components and the triple coupling relationship. As a result, the current system failure cause mechanism is unclear and risk classification control cannot be closed-loop. As a result, the current maintenance mode is still mainly event-driven fault repair mode and time-driven planned repair mode, which cannot deal with the operational safety hazards caused by component failure or failure and the safety accidents caused by the overall failure of the system due to under-maintenance and over-maintenance. Summary of the invention

[0003] The object of the present invention is to provide a method and system for constructing a failure chain of a complex electromechanical system taking into account dependencies and uncertainties, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] In one aspect, the present invention provides a method for constructing a failure chain of a complex electromechanical system, comprising:

[0006] Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed;

[0007] The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters;

[0008] Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, and the direct impact matrix between risk points is constructed. The matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points.

[0009] Calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0010] Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0011] Preferably, a topological network model of a complex electromechanical system is constructed, including: abstracting autonomous entities that have specific functions, are independent, indivisible, and interact with other units as nodes; abstracting the mechanical connections, electrical connections, and information connections between components in the complex electromechanical system as edges.

[0012] Preferably, risk points are selected based on the minimum maintainable and computable unit, combined with fault data and the current research status of the internal component structure of complex electromechanical systems; the risk point status is expressed through three state parameters and design indicators: mechanical, electrical, and information.

[0013] Preferably, based on the uncertainty method, the correlation coefficients between the risk point state parameters are calculated and measured, a direct impact matrix between the risk points is constructed, and the matrix is ​​visualized to obtain a weighted directed graph of the risk point dependency, including:

[0014] Combined with the data of complex electromechanical system failure cases, a risk point state parameter matrix was constructed;

[0015] Calculate the correlation coefficient of each risk point state parameter in the risk point state parameter matrix, measure the correlation coefficient using uncertainty, and construct the direct impact matrix between risk points;

[0016] Combined with the failure case data of complex electromechanical systems, a failure sequence frequency matrix is ​​constructed. According to the distribution of elements in the matrix in the fault sequence set, the cutoff value for distinguishing the correlation and causal relationship between risk points is obtained.

[0017] Preferably, the global impact matrix between risk points is calculated to obtain the quantitative degree of dependence between risk points, including: using the direct impact matrix and the scaling matrix to calculate the global impact matrix between risk points, the global impact matrix reveals the indirect impact between risk points; based on the global impact matrix between risk points, the weighted out-degree and weighted in-degree of each risk point are calculated.

[0018] Preferably, the failure sequence of the complex electromechanical system is predicted and the failure propagation chain of the complex electromechanical system is constructed, including: drawing a line graph of the change of state parameters of the risk point in combination with the historical fault data of the complex electromechanical system; using the global influence matrix between the risk points as the training data of the ANN artificial neural network to predict the complete failure chain of the complex electromechanical system; based on the component influence relationship of the complex electromechanical system and the direct influence matrix between the risk points, using the ANN artificial neural network to predict the failure chain of the complex electromechanical system with different time steps caused by the failure of the risk point; based on the time step of the failure of the risk point causing the failure of other risk points leading to the failure of the complex electromechanical system, calculating the shortest time step; using the direct influence matrix between the risk points as the training data of the ANN artificial neural network to predict the failure chain of the complex electromechanical system with different time steps.

[0019] In a second aspect, the present invention provides a complex electromechanical system failure chain construction system, comprising:

[0020] The first building module is used to build a complex electromechanical system topology network model based on the complex electromechanical system structure;

[0021] A characterization module, used to list risk points based on minimum maintainable and computable units, and to express the risk point status based on a mathematical characterization method; wherein the risk point status parameters include mechanical status parameters, electrical status parameters, and information status parameters;

[0022] The second building module is used to calculate and measure the correlation coefficients between the risk point state parameters based on the uncertainty method, build the direct impact matrix between the risk points, and visualize the matrix to obtain a weighted directed graph of the risk point dependency;

[0023] The calculation module is used to calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0024] The prediction module is used to predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0025] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the method for constructing a failure chain of a complex electromechanical system as described above is implemented.

[0026] In a fourth aspect, the present invention provides a computer program product, comprising a computer program, wherein when the computer program is executed on one or more processors, the computer program is used to implement the method for constructing a failure chain of a complex electromechanical system as described above.

[0027] In a fifth aspect, the present invention provides an electronic device, comprising: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the complex electromechanical system failure chain construction method as described above.

[0028] The beneficial effects of the present invention are as follows: the factors that cause the overall failure of the complex electromechanical system due to the failure propagation of the internal components of the complex electromechanical system are taken into consideration, and a method and basis are provided for the early prediction and subsequent maintenance of the failure of the complex electromechanical system; the interaction mechanism between components is deeply analyzed, and the influence of local failure on system failure is revealed.

[0029] Additional advantages of the present invention will be more clearly given in the following description or learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0031] Figure 1 It is a schematic diagram of a topological network model of a typical complex electromechanical system (urban rail train braking system) described in an embodiment of the present invention.

[0032] Figure 2 It is a weighted directed influence relationship diagram between risk points of a typical complex electromechanical system (urban rail train braking system) described in an embodiment of the present invention.

[0033] Figure 3 It is a weighted directed graph of the dependency relationships between risk points of a typical complex electromechanical system (urban rail train braking system) described in an embodiment of the present invention.

[0034] Figure 4 It is a distribution histogram of elements of the failure sequence frequency matrix of a typical complex electromechanical system (urban rail train braking system) described in an embodiment of the present invention.

[0035] Figure 5 It is a line graph of risk point state parameter changes of a typical complex electromechanical system (urban rail train braking system) described in an embodiment of the present invention.

[0036] Figure 6 It is a bar graph of complete failure chain propagation rate of a typical complex electromechanical system (urban rail train braking system) according to an embodiment of the present invention.

[0037] Figure 7 It is a feasibility comparison line graph of the complete failure chain propagation rate of a typical complex electromechanical system (urban rail train braking system).

[0038] Figure 8 This is a probability bar chart of the impact of failure risk points in the failure chain of a typical complex electromechanical system (urban rail train braking system) on brake cylinder failure at different time steps.

[0039] Fig. 9 The feasibility comparison line graph of failure chain propagation rate at different time steps for a typical complex electromechanical system (urban rail train braking system). DETAILED DESCRIPTION

[0040] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below by the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0041] It should be understood by those skilled in the art that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0042] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with that in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0043] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or groups thereof.

[0044] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0045] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings, and the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0046] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0047] Example 1

[0048] This embodiment 1 provides a complex electromechanical system failure chain construction system, including:

[0049] The first building module is used to build a complex electromechanical system topology network model based on the complex electromechanical system structure;

[0050] A characterization module, used to list risk points based on minimum maintainable and computable units, and to express the risk point status based on a mathematical characterization method; wherein the risk point status parameters include mechanical status parameters, electrical status parameters, and information status parameters;

[0051] The second building module is used to calculate and measure the correlation coefficients between the risk point state parameters based on the uncertainty method, build the direct impact matrix between the risk points, and visualize the matrix to obtain a weighted directed graph of the risk point dependency;

[0052] The calculation module is used to calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0053] The prediction module is used to predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0054] In this embodiment 1, a method for constructing a failure chain of a complex electromechanical system is implemented using the above system, including:

[0055] Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed;

[0056] The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters;

[0057] Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, and the direct impact matrix between risk points is constructed. The matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points.

[0058] Calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0059] Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0060] Specifically, in this embodiment 1, the method for constructing a failure chain of a complex electromechanical system based on the above-mentioned system implementation is a method for constructing a failure chain of a complex electromechanical system that takes into account dependencies and uncertainties, studies the mathematical characterization method of the risk status of components of a complex electromechanical system, comprehensively considers the status of the risk point and the topological functional attribute set and the functional attribute set of the edge, constructs the failure chain of the complex electromechanical system after the state of the risk point changes, explores the failure mechanism of "local failure → chain failure → system functional failure", and forms a set of failure propagation paths of the complex electromechanical system.

[0061] Specifically, the method comprises the following steps:

[0062] Step S1: According to the composition relationship of the complex electromechanical system and the coupling mechanism, a multi-correlated constitutive model of the complex electromechanical system is constructed. The specific steps are as follows:

[0063] Step S101: abstract autonomous entities that have specific functions, are independent, indivisible, and interact with other units into nodes. V is a collection of nodes in the multi-association constitutive model of a complex electromechanical system.

[0064] Step S102: Abstract the mechanical connections, electrical connections and information connections between components in the complex electromechanical system as edges. E is a set of edges in the multi-association constitutive model of the complex electromechanical system.

[0065] Step S2: By analyzing the principles and functions of components in complex electromechanical systems, some nodes are selected as risk points based on the minimum maintainable and computable units, and the status of the risk points is expressed through state parameters.

[0066] Step S201: Select risk points based on the minimum maintainable and computable unit, combined with fault data and the current research status of the internal component structure of the complex electromechanical system.

[0067] Step S202: Express the risk point status through three state parameters and design indicators: mechanical, electrical, and information.

[0068] Step S3: By analyzing the data of complex electromechanical system failure cases, a measurement operator based on multiple dependencies between risk points with uncertainty is proposed.

[0069] Step S4: A failure sequence prediction method for complex electromechanical systems based on ANN and dynamic risk is proposed to form a failure path set.

[0070] Among them, a measurement operator based on multiple dependencies between risk points with uncertainty proposed in step S3 considers using uncertainty method to quantify the degree of dependency between risk points. The specific operation steps are as follows:

[0071] Step S301: Based on the correlation coefficient and uncertainty calculation, construct a direct impact matrix between risk points to distinguish the dependencies between risk points. Including:

[0072] Step S3011: Construct a risk point state parameter matrix. Combined with complex electromechanical system failure case data, a risk point state parameter matrix is ​​constructed.

[0073] Step S3012: construct a direct impact matrix between risk points. Calculate the correlation coefficient of each risk point state parameter in the risk point state parameter matrix, measure the correlation coefficient using uncertainty, and construct a direct impact matrix between risk points.

[0074] Step S3013: Determine the cutoff value for distinguishing the dependency between risk points. Combined with the complex electromechanical system failure case data, a failure sequence frequency matrix is ​​constructed, and the cutoff value for distinguishing the correlation and causal relationship between risk points is obtained according to the distribution of the elements in the matrix in the failure sequence set.

[0075] Step S302: Calculate the degree of dependency between risk points based on the direct impact matrix between risk points. This includes:

[0076] Step S3021: construct a global impact matrix between risk points. The global impact matrix between risk points is calculated using the direct impact matrix and the scaling matrix. The global impact matrix reveals the indirect impact between risk points.

[0077] Step S3022: Calculate the weighted degree of the risk point. Based on the global impact matrix between the risk points, calculate the weighted out-degree and weighted in-degree of each risk point.

[0078] In step S4, a failure sequence prediction method for complex electromechanical systems based on ANN and dynamic risk is proposed to construct a failure chain of complex electromechanical systems. It includes:

[0079] Step S401: Based on the global impact matrix between the complex electromechanical system failure cases and risk points, the ANN artificial neural network is used to predict the complete failure chain of the complex electromechanical system failure caused by the failure of the risk point. It includes:

[0080] Step S4011: Construct a line graph of risk point state parameter changes. Based on a complex electromechanical system failure case, draw a line graph of risk point state parameter changes.

[0081] Step S4012: Constructing a complete failure chain of a complex electromechanical system. The global influence matrix between risk points is used as training data for the ANN artificial neural network to predict the complete failure chain of the complex electromechanical system.

[0082] Step S402: Based on the direct influence matrix between the component influence relationship and risk points of the complex electromechanical system, the ANN artificial neural network is used to predict the failure chain of different time steps that will lead to the failure of the complex electromechanical system due to the failure of the risk point. This includes:

[0083] Step S4021: Calculate the time step length of the failure of the risk point leading to the failure of the complex electromechanical system. Calculate the shortest time step length based on the time step length of the failure of the risk point leading to the failure of the complex electromechanical system due to the failure of other risk points.

[0084] Step S4022: construct failure chains of complex electromechanical systems at different time steps. Use the direct impact matrix between risk points as training data for the ANN artificial neural network to predict failure chains of complex electromechanical systems at different time steps.

[0085] In summary, in this embodiment 1, the failure of the complex electromechanical system is studied from the perspective of the failure propagation of the internal components of the complex electromechanical system leading to the overall failure of the complex electromechanical system, the interaction mechanism between the components is deeply analyzed, and the influence of local failure on system failure is revealed, which provides methods and basis for the early prediction and subsequent maintenance of the failure of the complex electromechanical system.

[0086] Example 2

[0087] In order to provide a theoretical basis for the failure of complex electromechanical systems, in this embodiment 2, a method for constructing a failure chain of a complex electromechanical system considering dependencies and uncertainties is provided. The method considers the coupling mechanism between components of the complex electromechanical system, constructs a multi-correlated constitutive model of the complex electromechanical system, selects risk points based on the minimum maintainable and computable units, expresses the state of the risk points, proposes a measurement operator for multiple dependencies between risk points based on uncertainty, calculates the correlation coefficient and uncertainty between the state parameters of the risk points, constructs a direct influence matrix between the risk points, and then uses matrix operations to calculate the global influence matrix between the risk points, quantifies the degree of dependence between the risk points, and finally proposes a failure sequence prediction method for a complex electromechanical system based on ANN and dynamic risk to construct a failure chain of a complex electromechanical system.

[0088] The method for risk point selection and state parameter expression based on the minimum maintainable and calculable unit includes the following steps:

[0089] Step S1: According to the composition relationship of the complex electromechanical system and the coupling mechanism, a multi-correlated constitutive model of the complex electromechanical system is constructed. The specific steps are as follows:

[0090] Step S101: abstract autonomous entities that have specific functions, are independent, indivisible, and interact with other units into nodes. V is a collection of nodes in the multi-association constitutive model of a complex electromechanical system.

[0091] Step S102: Abstract the mechanical connections, electrical connections and information connections between components in the complex electromechanical system as edges. E is the set of edges in the multi-association constitutive model of the complex electromechanical system:

[0092] in, Indicates the mechanical connection relationship between components; Indicates the electrical connection relationship between components; Represents the information connection relationship between components.

[0093] Step S2: By analyzing the principles and functions of components in complex electromechanical systems, some nodes are selected as risk points based on the minimum maintainable and computable units, and the status of the risk points is expressed through state parameters.

[0094] Step S201: Select risk points. Based on the minimum maintainable and computable unit, combined with the composition, working principle and fault data of the complex electromechanical system, select risk points v i , including key risk points and weak risk points.

[0095] Step S202: The risk point status is expressed through mechanical, electrical, and information status parameters and design indicators. P represents the risk point status parameter, P = 1 represents normal operation of the risk point, P = 0 represents a risk point failure, m represents the mechanical status parameter, e represents the electrical status parameter, and i represents the information status parameter.

[0096] Among them, P represents the risk point state parameter; m represents the mechanical state parameter; m0 represents the mechanical design index range; e represents the electrical state parameter; e0 represents the electrical design index range; i represents the information state parameter; i0 represents the information design index range.

[0097] Step S3: By analyzing the data of complex electromechanical system failure cases, a measurement operator based on multiple dependencies between risk points with uncertainty is proposed. It includes:

[0098] Step S301: Construct a risk point state parameter matrix. Let P = [P ij ] n×n is the risk point state parameter matrix, where the element P in the matrix ij Represents the state parameter P of risk point j at the i-th time step j , n is the number of risk points selected. Construct the risk point state parameter matrix P, P = [P ij ] n×n As shown in Table 1.

[0099] Table 1 Risk point status parameter matrix

[0100]

[0101] Step S302: construct a direct impact matrix between risk points. The direct impact matrix reveals the direct impact relationship and degree between risk points.

[0102] After expressing the risk point state parameters, the correlation coefficient between the risk point state parameters is calculated and measured using uncertainty. Uncertainty refers to the degree of uncertainty of the measured value due to the error in the measurement process. The smaller the uncertainty, the greater the measurement accuracy.

[0103] According to the obtained correlation coefficient, the direct impact matrix between risk points is constructed, and the direct impact matrix D = [d ij ] n×n , where the element d in the matrix ij represents the influence of risk point i on risk point j, and n is the number of selected risk points. Construct the risk point state parameter matrix D, D = [d ij ] n×n As shown in Table 2. The constructed direct impact matrix D is visualized to obtain the weighted directed impact relationship diagram between risk points.

[0104] Table 2 Direct impact matrix between risk points

[0105]

[0106] Step S303: Determine the demarcation point value for distinguishing the impact relationship between risk points. Combined with the complex electromechanical system failure case data, construct the failure sequence frequency matrix W = [w ij ] n×n , where the element w in the matrix ij represents the number of time steps after the risk point i fails and the risk point j fails in all failure cases, and n is the number of risk points selected. Construct the risk point state parameter matrix W, W = [w ij ] n×n As shown in Table 3.

[0107] Table 3 Risk point failure sequence frequency matrix

[0108]

[0109] Define a as the cutoff value for distinguishing the impact relationship between risk points. The choice of a depends on w ij Distribution among the set of fault sequences.

[0110] Step S304: Construct a global impact matrix between risk points. Define the scaling matrix R, which is defined as:

[0111] R=[r ij ] n×n

[0112]

[0113] Where: R represents the scaling matrix; r ij Represents the elements in the scaling matrix, used to convert t ij Scaled to the range of [0,1]; i, j represent risk points; n represents the number of risk points; t ijIt represents the elements in the global impact matrix, representing the indirect impact of risk point i on risk point j, that is, the quantitative degree of dependence.

[0114] Based on the direct impact matrix D and scaling matrix R between risk points, combined with matrix operations, the global impact matrix T between risk points is constructed. ij ] n×n , the calculation formula is:

[0115]

[0116] Where: T represents the global impact matrix between risk points; R represents the scaling matrix; D represents the direct impact matrix between risk points; k represents the time step; n represents the number of risk points.

[0117] The global impact matrix T between risk points reveals the indirect impact between risk points. The size of the elements in the matrix represents the degree of dependence between risk points quantified by mathematical calculation.

[0118] Step S4: A failure sequence prediction method for complex electromechanical systems based on ANN and dynamic risk is proposed to form a failure path set.

[0119] Step S401: Construct a line graph of risk point state parameter changes. Based on the working principle and function of the risk point, the connection relationship and coupling mechanism between the risk points, combined with the complex electromechanical system failure case and the calculated direct impact matrix, a line graph of risk point state parameter changes is drawn. The range of risk point state parameter changes is between {0,1}.

[0120] Step S402: Construct a complete failure chain of a complex electromechanical system. Based on the risk point state parameter change line graph, construct a fault sequence matrix of risk point state parameter changes at each time step after the first risk point fails. The first row of elements in the matrix are all 1, indicating that all risk points are operating normally at this time. The last row of elements in the matrix are all 0, indicating that all risk points have failed at this time and the complex electromechanical system has failed as a whole. The fault sequence matrix and the global influence matrix T are converted into input and output data to train the ANN artificial neural network. The trained ANN predicts the probability of fault propagation and constructs a complete failure chain of a complex electromechanical system.

[0121] Step S5: Based on the direct influence matrix between the component influence relationship and the risk points of the complex electromechanical system, the ANN artificial neural network is used to predict the failure chain of different time steps of the complex electromechanical system caused by the failure of the risk point.

[0122] Step S501: Calculation of the time step for the failure of the complex electromechanical system caused by the failure of the risk point. In the actual working process of complex electromechanical equipment, there are cases where the failure of some risk points leads to the failure of the complex electromechanical system, that is, when the fault has not yet propagated to certain risk points, the complex electromechanical system has already failed. Based on the different functions of risk points in the complex electromechanical system, the time step for the propagation of risk point failure to cause the overall failure of the complex electromechanical system is different. The shortest time step for propagation can be obtained based on the weighted directed influence relationship diagram between risk points.

[0123] Step S502: Construct failure chains of complex electromechanical systems at different time steps. Take the failure of each risk point into consideration, construct a risk point state parameter matrix for each case, convert the matrix and the direct impact matrix D into input and output data to train the ANN artificial neural network. The trained ANN predicts the probability of fault propagation and constructs failure chains of complex electromechanical systems at different time steps.

[0124] Example 3

[0125] like Figures 1 to 9 As shown, in this embodiment 3, taking the urban rail train braking system as an example, the failure chain construction method is specifically introduced using actual fault data as follows, including:

[0126] Step 1: Construct a topological network model of the urban rail train braking system, select risk points and express their status.

[0127] According to the composition relationship of the urban rail transit train braking system and combined with the coupling mechanism, a topological coupling network model of the urban rail transit train braking system is constructed.

[0128] Through the analysis of the component principle and function of the air brake in the braking system of urban rail transit trains, based on the minimum maintainable and calculable unit, combined with the composition and working principle of the braking system of urban rail transit trains, the risk point v is selected. i The key risk points include emergency solenoid valve v1, EP valve v2, relay valve v3, empty and loaded vehicle valve v4, brake cylinder v5 and air source system v6, and the weak risk point is anti-skid system v7. The topological coupling network model of some urban rail transit train braking systems containing risk points is extracted.

[0129] The calculation formula of the emergency solenoid valve state parameters is:

[0130] If the urban rail train is in emergency braking condition,

[0131]

[0132] If the urban rail train is in non-emergency braking condition and the normal and rapid braking condition is applied, then

[0133]

[0134] Where: P1 represents the state parameter of the emergency solenoid valve, P1 is 1 for health, P1 is 0 for failure; t represents the monitoring time; p represents the pressure value of the relay valve volume chamber.

[0135] The calculation formula of EP valve state parameters is:

[0136]

[0137] Where: P2 represents the EP valve state parameter, P2 = 1 represents health, P2 = 0 represents failure; Ps represents the target pressure; PCV represents the pre-control pressure.

[0138] The expression of the empty and loaded vehicle valve output pressure is:

[0139]

[0140] Where: p CV2 Indicates the empty and loaded vehicle valve output pressure; p CV1 Indicates the empty and loaded vehicle valve input pressure; p CV2(MAX) Indicates the maximum output pressure.

[0141] The calculation formula of empty and loaded vehicle valve state parameters is:

[0142]

[0143] Where: P4 represents the empty and loaded vehicle valve state parameter, P4 1 represents healthy, P4 0 represents fault; Ps represents the target pressure.

[0144] The calculation formula of brake cylinder state parameters is:

[0145] If the urban rail train is under normal braking conditions,

[0146]

[0147] Where: P5 represents the brake cylinder status parameter, P5 is 1 for healthy, P5 is 0 for fault; t represents the monitoring time; p represents the brake cylinder pressure.

[0148] Based on the failure case data, a weighted and directed influence relationship diagram between risk points is constructed.

[0149] Step 2: Construct the direct impact matrix D between risk points

[0150] Calculate the correlation coefficient between the state parameters of the risk points and use uncertainty to measure. Taking the emergency solenoid valve and EP valve as examples, the state parameter P1 of the emergency solenoid valve is input into Matlab and represented by the X1 matrix. The same is true for the EP valve, which is represented by the X2 matrix. The process of calculating the correlation coefficient of the two risk point states is as follows:

[0151] X1=[1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0]

[0152] X2=[1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 0 0 0]

[0153] [r, p] = corrcoef(X1, X2)

[0154] get

[0155] r=1.0000 0.5238 0.5238 1.0000

[0156] p=1.0000 0.0178 0.0178 1.0000

[0157] Where: X1 represents the state parameter matrix of the emergency solenoid valve; X2 represents the state parameter matrix of the EP valve; r represents the returned correlation matrix; p represents the returned significance matrix; corrcoef represents the algorithm used by Matlab to calculate the sample correlation coefficient.

[0158] The confidence calculation formula is:

[0159]

[0160] Where: t represents the critical value of t distribution; r represents the correlation coefficient of the return; n-2 represents the degree of freedom.

[0161] The confidence interval calculation process is:

[0162] [r,p,rlo,rup]=corrcoef(X1,X2)

[0163] r=1.0000 0.5238 0.5238 1.0000

[0164] p=1.0000 0.0178 0.0178 1.0000

[0165] rlo=1.0000 0.1058 0.1058 1.0000

[0166] rup=1.0000 0.7845 0.7845 1.0000

[0167] Where: rlo represents the lower limit of the confidence interval correlation coefficient; rup represents the upper limit of the confidence interval correlation coefficient.

[0168] Taking the significance level as 0.05, the returned significance is 0.0178<0.05, the confidence level is 95%, and the confidence interval is [0.1058, 0.7845].

[0169] According to the correlation coefficient between the risk point state parameters and the description and data of the fault case, the direct impact matrix D between the risk points is constructed. ij ] n×n , where n is the number of risk points selected, n is 7, and the direct impact matrix D between risk points is as follows:

[0170]

[0171] The direct impact matrix D between risk points is visualized, and a weighted directed graph of the dependency relationships between risk points in the urban rail train braking system is obtained.

[0172] Step 3: Identify the dependencies between risk points.

[0173] Combined with the urban rail transit train brake system failure case data, according to w ij The distribution in the fault sequence set determines the cutoff point value a.

[0174] Step 4: Calculate the global impact matrix T between risk points

[0175] According to the definition of the scaling matrix, substituting n=7, the scaling matrix R is as follows:

[0176]

[0177] According to the calculation formula of the global impact matrix T, the total impact matrix T between risk points is obtained as shown below:

[0178]

[0179] Step 5: Construct a line chart of risk point status parameter changes

[0180] Based on the failure case, where the risk point status is 1 for normal and the risk point status is 0 for failure, in this failure case, the system initially fails at risk point 6. As the failure status propagates, it affects risk point 1, and then risk point 2, risk point 4, risk point 3, and risk point 7 are affected by the failure in the order of risk point 2, risk point 4, risk point 3, and risk point 7, and finally propagates to risk point 5, and a line graph of risk point status parameter changes is constructed.

[0181] Step 6: Construct a complete failure chain of the urban rail train braking system

[0182] Based on the line graph of risk point state parameter changes in the fault sequence, the fault sequence matrix X of risk point state parameter changes at each time step after the failure of the first risk point is constructed as shown in the following formula.

[0183]

[0184] The global influence matrix T between the fault sequence and the risk point is converted into input and output data to train the ANN artificial neural network. The trained ANN predicts the fault propagation probability and obtains a bar chart of the failure chain propagation rate of the braking system and a feasibility comparison line chart.

[0185] Define the matrix Y, which represents the failure probability in the predicted fault propagation chain. The matrix Y is shown as follows:

[0186]

[0187] The complete failure chain of the urban rail train braking system is constructed, and the connecting influence relationship is "air source system → emergency solenoid valve → EP valve → empty and loaded vehicle valve → relay valve → anti-skid system → brake cylinder".

[0188] Step 7: Construct failure chains of urban rail train braking system at different time steps

[0189] The state of the brake cylinder is judged based on the direct influence matrix D. Since the pressure output by the brake cylinder is transmitted to the basic brake device, the train is braked under the action of the basic brake device, and the brake cylinder is ranked last in the complete failure propagation chain, the state of the brake cylinder is selected to judge the state of the brake system.

[0190] The time step of failure propagation is determined based on the weighted directed influence relationship diagram between risk points. For example, if risk point 3 fails, the failure can be transmitted to risk point 5 in at least 1 time step, affecting the brake cylinder; while if risk point 6 fails, it takes at least 3 time steps to transmit the failure to risk point 5. At this time, we take the failure of each risk point into consideration and take the failure chain with the shortest time step propagated to the brake cylinder for prediction.

[0191] Taking into account the failure of each risk point, the risk point state parameter matrix M is constructed as shown below:

[0192]

[0193] The risk point state parameter matrix M and the direct influence matrix D between risk points are converted into input and output data to train the ANN artificial neural network. The trained ANN predicts the fault propagation probability and obtains the probability bar chart of the influence of the failure risk point of the braking system on the failure of the brake cylinder and the feasibility comparison line chart.

[0194] Example 4

[0195] Embodiment 4 of the present invention provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, a method for constructing a failure chain of a complex electromechanical system is implemented; the method comprises:

[0196] Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed;

[0197] The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters;

[0198] Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, and the direct impact matrix between risk points is constructed. The matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points.

[0199] Calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0200] Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0201] Example 5

[0202] Embodiment 5 of the present invention provides a computer program (product), including a computer program, wherein when the computer program is run on one or more processors, the computer program is used to implement a method for constructing a failure chain of a complex electromechanical system; the method includes:

[0203] Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed;

[0204] The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters;

[0205] Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, and the direct impact matrix between risk points is constructed. The matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points.

[0206] Calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0207] Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0208] Example 6

[0209] Embodiment 6 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions for implementing a method for building a failure chain of a complex electromechanical system; the method includes:

[0210] Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed;

[0211] The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters;

[0212] Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, and the direct impact matrix between risk points is constructed. The matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points.

[0213] Calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points;

[0214] Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems.

[0215] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0216] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0217] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0218] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0219] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative work on the basis of the technical solution disclosed in the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for constructing a failure chain of a complex electromechanical system, characterized in that: include: Based on the structure of complex electromechanical systems, a topological network model of complex electromechanical systems is constructed; The risk points are listed based on the minimum maintainable and computable units, and the risk point states are expressed based on a mathematical characterization method; wherein the risk point state parameters include mechanical state parameters, electrical state parameters, and information state parameters; Based on the uncertainty method, the correlation coefficients between the state parameters of risk points are calculated and measured, the direct influence matrix between risk points is constructed, and the matrix is ​​visualized to obtain a weighted directed graph of the dependency relationship of risk points; including: combining the data of complex electromechanical system failure cases to construct a risk point state parameter matrix; calculating the correlation coefficients of the state parameters of each risk point in the risk point state parameter matrix, measuring the correlation coefficients using uncertainty, and constructing a direct influence matrix between risk points; combining the data of complex electromechanical system failure cases to construct a failure sequence frequency matrix, and according to the distribution of the elements in the matrix in the failure sequence set, obtaining the demarcation value for distinguishing the correlation and causal relationship between risk points; Calculate the global impact matrix between risk points to obtain the quantitative degree of dependence between risk points; including: using the direct impact matrix and the scaling matrix to calculate the global impact matrix between risk points, the global impact matrix reveals the indirect impact between risk points; based on the global impact matrix between risk points, calculate the weighted out-degree and weighted in-degree of each risk point; Predict the failure sequence of complex electromechanical systems and construct the failure propagation chain of complex electromechanical systems; including: combining the historical fault data of complex electromechanical systems to draw a line graph of the change of risk point state parameters; using the global influence matrix between risk points as the training data of the ANN artificial neural network to predict the complete failure chain of the complex electromechanical system; based on the component influence relationship of the complex electromechanical system and the direct influence matrix between risk points, using the ANN artificial neural network to predict the failure chain of complex electromechanical systems with different time steps caused by the failure of risk points; based on the time step of the failure of risk points causing the failure of other risk points leading to the failure of complex electromechanical systems, calculate the shortest time step; using the direct influence matrix between risk points as the training data of the ANN artificial neural network to predict the failure chain of complex electromechanical systems with different time steps.

2. The method for constructing a failure chain of a complex electromechanical system according to claim 1, characterized in that: Construct a topological network model of a complex electromechanical system, including: abstracting autonomous entities that have specific functions, are independent, indivisible, and interact with other units as nodes; abstracting the mechanical connections, electrical connections, and information connections between components in the complex electromechanical system as edges.

3. The method for constructing a failure chain of a complex electromechanical system according to claim 1, characterized in that: Based on the minimum maintainable and computable unit, risk points are selected in combination with fault data and the current research status of the internal component structure of complex electromechanical systems; the risk point status is expressed through mechanical, electrical, and information state parameters and design indicators.

4. A complex electromechanical system failure chain construction system, characterized in that: include: The first building module is used to build a complex electromechanical system topology network model based on the complex electromechanical system structure; A characterization module, used to list risk points based on minimum maintainable and computable units, and to express the risk point status based on a mathematical characterization method; wherein the risk point status parameters include mechanical status parameters, electrical status parameters, and information status parameters; The second construction module is used to calculate and measure the correlation coefficients between the state parameters of risk points based on the uncertainty method, construct the direct influence matrix between risk points, and visualize the matrix to obtain a weighted directed graph of the dependency relationship of risk points; including: combining the complex electromechanical system failure case data to construct the risk point state parameter matrix; calculating the correlation coefficients of the state parameters of each risk point in the risk point state parameter matrix, measuring the correlation coefficients using uncertainty, and constructing the direct influence matrix between risk points; combining the complex electromechanical system failure case data to construct the failure sequence frequency matrix, and according to the distribution of the elements in the matrix in the failure sequence set, obtaining the demarcation value for distinguishing the correlation and causal relationship between risk points; The calculation module is used to calculate the global impact matrix between risk points and obtain the quantitative degree of dependence between risk points; including: using the direct impact matrix and the scaling matrix to calculate the global impact matrix between risk points, the global impact matrix reveals the indirect impact between risk points; based on the global impact matrix between risk points, calculate the weighted out-degree and weighted in-degree of each risk point; The prediction module is used to predict the failure sequence of the complex electromechanical system and construct the failure propagation chain of the complex electromechanical system; including: combining the historical fault data of the complex electromechanical system to draw a line graph of the change of the state parameters of the risk point; using the global influence matrix between the risk points as the training data of the ANN artificial neural network to predict the complete failure chain of the complex electromechanical system; based on the component influence relationship of the complex electromechanical system and the direct influence matrix between the risk points, using the ANN artificial neural network to predict the failure chain of the complex electromechanical system with different time steps caused by the failure of the risk point; based on the time step of the failure of the risk point causing the failure of other risk points leading to the failure of the complex electromechanical system, calculate the shortest time step; using the direct influence matrix between the risk points as the training data of the ANN artificial neural network to predict the failure chain of the complex electromechanical system with different time steps.

5. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the method for constructing a failure chain of a complex electromechanical system as described in any one of claims 1 to 3 is implemented.

6. A computer program product, characterized in that The invention comprises a computer program, which is used to implement the method for constructing a failure chain of a complex electromechanical system according to any one of claims 1 to 3 when the computer program is run on one or more processors.

7. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the method for constructing a failure chain of a complex electromechanical system as described in any one of claims 1 to 3.

Images