A method for simulating damage of aviation arc fault

Abstract

The present invention relates to the field of simulation computing technology, and discloses a method for simulating damage caused by an aviation arc fault, comprising the following steps: collecting physical quantity distribution data, medium property data, and arc data of the current time step and each previous time step through multi-physics field simulation software; encoding and combining the collected data into first graph structure data and first sequence data; inputting the first graph structure data and the first sequence data into a first neural network, and outputting the spatial coordinates of the arc at the next time step; connecting the arc end point coordinates output by the first neural network at all time steps to obtain the arc movement trajectory; judging the damage degree of the aircraft according to the volume of the temperature of the position of the aircraft where the arc passes exceeding the melting point of the material at that position. The present invention provides a method for simulating damage caused by an aviation arc fault, which solves the technical problems of high experimental cost and high difficulty in reproducing experimental scenes in the fault arc damage verification experiment.

Description

Technical Field

[0001] The present invention relates to the field of simulation computing technology, and more specifically, to a method for simulating aviation arc fault damage. Background Art

[0002] With the rapid development of the modern aviation industry, aircraft are equipped with more and more onboard equipment, resulting in a sharp increase in the use of aircraft cable harnesses. During aircraft operation, cables are easily affected by air turbulence, causing loose connections or physical damage to the insulation layer due to contact with metal parts. In addition, the cables are located in high-corrosion areas. Due to long-term environmental erosion or aging such as temperature, chemical erosion and humidity, the insulation of the cable insulation layer is greatly reduced. These factors undoubtedly increase the safety risk of triggering fault arcs.

[0003] Fault arc is a continuous plasma discharge phenomenon after the insulation medium is broken down. It has the characteristics of strong conductivity, concentrated energy, and high temperature. It is very easy to cause direct and indirect damage to important aircraft components (such as surrounding cable harnesses, various pipelines, fuel tank wall panels, and windshields), and even induce catastrophic failure of the aircraft. In addition, due to the high randomness and concealment of fault arcs, it is difficult to detect fault arcs in real time. Therefore, fault arcs have always been an important safety hazard to the safety of aircraft electrical interconnection systems (EWIS).

[0004] The research on damage prediction of aviation fault arcs has gradually become a key focus of aircraft safety design and airworthiness certification. However, the current research on aviation fault arcs is mainly based on the nature of the fault arc itself, and the fault arc is identified by analyzing the differences in the waveform characteristics of different fault arcs, or by building a simplified model through simulation software to simulate and analyze the real-time temperature field and electromagnetic field changes of the fault arc. Due to the randomness and complexity of the fault arc, experiments to simulate fault arcs in aircraft are not only costly, but also may bring safety risks. Summary of the invention

[0005] The present invention provides an aviation arc fault damage simulation method to solve the technical problems of high experimental cost and great difficulty in reproducing experimental scenes in fault arc damage verification experiments.

[0006] The present invention provides an aviation arc fault damage simulation method, comprising the following steps:

[0007] Step 100, collecting physical quantity distribution data, medium property data, and arc data of the current time step and each previous time step through multi-physics field simulation software;

[0008] Step 200, the data collected at the current time step is encoded and combined into first graph structure data, and the data collected at the current time step and each previous time step is encoded and combined into first sequence data;

[0009] Step 300, inputting the first graph structure data and the first sequence data into the first neural network, outputting the spatial coordinates of the arc at the next time step, the first neural network comprising a first module, a second module, a third module and an output module;

[0010] Step 400, connecting the arc endpoint coordinates output by the first neural network at all time steps to obtain the arc movement trajectory;

[0011] Step 500, judging the damage degree of the aircraft according to the volume of the position of the aircraft where the arc passes through which the temperature exceeds the melting point of the material at the position.

[0012] In a preferred embodiment, the physical quantity distribution data includes: temperature field distribution and electric field distribution;

[0013] The medium attribute data include: density, conductivity and dielectric constant of various surrounding media;

[0014] Arc data includes: arc spatial coordinates, voltage and current.

[0015] In a preferred embodiment, first graph structure data is generated based on the collected data, the first graph structure data takes the spatial coordinates of the current arc as the central node, the first graph structure data includes nodes and edges, a node contains a set of first data as node features, the first data includes: the spatial coordinates, temperature, electric field vector of the position, and the density, conductivity and dielectric constant of the medium; there is an edge between two adjacent nodes, the edge represents the distance between the spatial coordinates of the two nodes, the first graph structure data includes the node where the spatial coordinates of the collected arc are located and its adjacent nodes and the edges between these nodes.

[0016] In a preferred embodiment, medium property data of the spatial coordinates of the arc obtained at the current time step and each previous time step are collected, and the collected data are encoded into a first sequence data, the first sequence data includes n sequence units, and the tth sequence unit represents the medium property data of the spatial coordinates of the arc obtained at the tth time step.

[0017] In a preferred embodiment, the calculation formula of the first module is as follows:

[0018]

[0019] in represents the first hidden state of the i-th node, represents the state weight parameter, and Represent the node features of the i-th and j-th nodes respectively, represents the set of nodes directly connected to the i-th node, express The total number of nodes in , tanh represents the hyperbolic tangent function, Indicates adjustable parameters; The default value of is 1 / N, where N is the dimension of the node feature. Represents a sigmoid function.

[0020] In a preferred embodiment, the calculation formula of the second module is as follows:

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Represent the first, second, and third intermediate states respectively, where represents the tth sequence unit of the first sequence data, and They represent the t-th and t-1-th second hidden states respectively, n≥t≥1, n represents the total number of sequence units of the first sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

[0026] In a preferred embodiment, the calculation formula of the third module is as follows:

[0027] ;

[0028] in is the nth second hidden state output by the second module, is the first hidden state of the i-th node output by the first module, represents the third hidden state, and N is the set of all nodes in the first graph structure data.

[0029] In a preferred embodiment, the calculation formula of the output module is as follows:

[0030] ;

[0031] Where y represents the first vector of the output, the cth component of the first vector represents the probability that the arc endpoint coordinates at the next time step belong to the cth spatial coordinate, and m represents the third hidden state. is the fourth weight parameter, is the fourth bias parameter, Represents a sigmoid function.

[0032] In a preferred embodiment, the step of calculating the volume of the position of the aircraft where the arc passes through and the temperature exceeds the melting point of the material at the position comprises:

[0033] Step 101, determining arc waveform, arc movement trajectory and arc duration;

[0034] Step 102, confirming the size parameters of the location where the aviation fault arc occurs, including: arc starting wire length, wire diameter, insulation layer thickness, fuel line thickness, line radius, line length, and the distance between the arc and the aluminum alloy line;

[0035] Step 103, integrating the arc waveform in step 101 to calculate the fault arc energy;

[0036] Step 104, based on the size parameters of the aviation fault arc occurrence location confirmed in step 102, a model is constructed using multi-physics field simulation software, and boundary conditions are set, wherein the heat source value is set based on the arc energy result calculated in step 103;

[0037] Step 105, meshing the model using a custom meshing method, determining the arc fault occurrence area based on the arc movement trajectory and setting it to refinement, and setting the remaining areas to normal;

[0038] Step 106, setting the solver to transient calculation, taking the arc duration in step 101 as the model calculation duration, setting the step length according to the calculation duration, and starting the calculation;

[0039] Step 107, determine whether the calculation result converges. If the iterative calculation error value is less than the set value, it is determined to be converged and the calculation is completed. Combined with the arc movement trajectory, the temperature of the position of the aircraft where the arc passes exceeds the volume of the melting point of the material at that position. If the model calculation does not converge, the model automatically reports an error, then return to steps 104 and 105, reset the boundary conditions and grid size until convergence.

[0040] The present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned aviation arc fault damage simulation method is implemented.

[0041] The beneficial effects of the present invention are:

[0042] The present invention can consider multiple factors such as temperature field, electric field, arc path, dielectric material, etc. and construct a simulation model through multi-physics field simulation software. The damage volume of the aluminum alloy pipeline is obtained through the simulation results, which solves the technical problems of high cost and difficulty in reproducing experimental scenarios in the fault arc damage verification experimental method. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of a method for simulating aviation arc fault damage according to the present invention;

[0044] Figure 2 It is a visualization diagram of the calculation results of the damage caused by the arc to the surrounding components. DETAILED DESCRIPTION

[0045] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0046] At least one embodiment of the present invention discloses a method for simulating aviation arc fault damage, such as Figure 1 As shown, the following steps are included:

[0047] Step 100, collecting physical quantity distribution data, medium property data, and arc data of the current time step and each previous time step through multi-physics field simulation software;

[0048] The physical quantity distribution data include: temperature field distribution and electric field distribution;

[0049] The medium attribute data include: density, conductivity and dielectric constant of various surrounding media;

[0050] Arc data include: arc spatial coordinates, voltage and current;

[0051] In one embodiment of the present invention, initial physical quantity distribution data, medium property data, and arc data are obtained through experimental measurements.

[0052] In one embodiment of the present invention, the entire experimental space is mean discretized to obtain multiple spatial coordinates, and the temperature field distribution and electric field distribution as well as the density, conductivity and dielectric constant of various surrounding media refer to the temperature, electric field vector and the density, conductivity and dielectric constant of the medium at each spatial coordinate.

[0053] Step 200, the data collected at the current time step is encoded and combined into first graph structure data, and the data collected at the current time step and each previous time step is encoded and combined into first sequence data;

[0054] In one embodiment of the present invention, first graph structure data is generated based on collected data, the first graph structure data takes the spatial coordinates of the current arc as the central node, the first graph structure data includes nodes and edges, a node contains a set of first data as node features, the first data includes: the spatial coordinates, temperature, electric field vector of the position, and the density, conductivity and dielectric constant of the medium; there is an edge between two adjacent nodes, the edge represents the distance between the spatial coordinates of the two nodes, the first graph structure data includes the node where the spatial coordinates of the collected arc are located and its adjacent nodes and the edges between these nodes.

[0055] In one embodiment of the present invention, medium property data of the space coordinates of the electric arc obtained at the current time step and each previous time step are collected, and the collected data are encoded into a first sequence data, the first sequence data includes n sequence units, and the tth sequence unit represents the medium property data of the space coordinates of the electric arc obtained at the tth time step;

[0056] Step 300, inputting the first graph structure data and the first sequence data into the first neural network, outputting the spatial coordinates of the arc at the next time step, the first neural network comprising a first module, a second module, a third module and an output module;

[0057] The first module is used to take the first graph structure data as input and output the first hidden state of each node in the current time step;

[0058] The second module is used to take the first sequence data as input and output the second hidden state of the current time step;

[0059] The third module is used to sum the first hidden state and the second hidden state, and output the third hidden state of each node in the current time step;

[0060] In one embodiment of the present invention, the summing process is to sum the hidden state of each node of the first graph structure data with the second hidden state respectively;

[0061] The output module is used to take the third hidden state as input and output the spatial coordinates of the arc at the next time step;

[0062] In one embodiment of the present invention, the calculation formula of the first module is as follows:

[0063]

[0064] in represents the first hidden state of the i-th node, represents the state weight parameter, and Represent the node features of the i-th and j-th nodes respectively, represents the set of nodes directly connected to the i-th node, express The total number of nodes in , tanh represents the hyperbolic tangent function, Indicates adjustable parameters; The default value of is 1 / N, where N is the dimension of the node feature. Represents a sigmoid function.

[0065] In one embodiment of the present invention, the calculation formula of the second module is as follows:

[0066] ;

[0067] ;

[0068] ;

[0069] ;

[0070] in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Represent the first, second, and third intermediate states respectively, where represents the tth sequence unit of the first sequence data, and They represent the t-th and t-1-th second hidden states respectively, n≥t≥1, n represents the total number of sequence units of the first sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

[0071] In one embodiment of the present invention, the calculation formula of the third module is as follows:

[0072] ;

[0073] in is the nth second hidden state output by the second module, is the first hidden state of the i-th node output by the first module, represents the third hidden state, and N is the set of all nodes in the first graph structure data.

[0074] In one embodiment of the present invention, the calculation formula of the output module is as follows:

[0075] ;

[0076] Where y represents the first vector of the output, the cth component of the first vector represents the probability that the arc endpoint coordinates at the next time step belong to the cth spatial coordinate, and m represents the third hidden state. is the fourth weight parameter, is the fourth bias parameter, Represents a sigmoid function.

[0077] Step 400, connecting the arc endpoint coordinates output by the first neural network at all time steps to obtain the arc movement trajectory;

[0078] Step 500, judging the damage degree of the aircraft according to the volume of the position of the aircraft where the arc passes through which the temperature exceeds the melting point of the material at the position.

[0079] For example, if the insulation layer at the cable location melts and exposes the core wire, it is judged that the cable is short-circuited. The degree of damage is then assessed based on the impact of the cable short-circuit on the operation of the aircraft.

[0080] In one embodiment of the present invention, the second module is independently pre-trained. During the pre-training, the second module is connected to the first output layer, and the first output layer outputs the final medium properties representing the coordinate position of the arc end point at the current time step. The final medium properties refer to the medium properties after the medium is affected by the arc passing through.

[0081] The calculation formula of the first output layer is as follows:

[0082] ;

[0083] in represents the pre-trained output vector, one component of which represents the value of a medium attribute, represents the t-th second hidden state, is the fifth weight parameter, is the fifth bias parameter, Represents a sigmoid function.

[0084] In one embodiment of the present invention, the loss function of the second module pre-training is a mean square error loss function, and the calculation formula is as follows:

[0085] ;

[0086] in, is the predicted value generated by the input data at the sth time step, is the exact value of the sth time step measured through a specific experimental process, and n1 is the total number of time steps.

[0087] In order to obtain the volume of the position of the aircraft where the arc passes and the temperature exceeds the melting point of the medium at the position, the present invention provides a method for calculating the damage of the arc to the surrounding components, comprising:

[0088] Step 101, determining arc waveform, arc movement trajectory and arc duration;

[0089] Step 102, confirming the size parameters of the location where the aviation fault arc occurs, including: arc starting wire length, wire diameter, insulation layer thickness, fuel line thickness, line radius, line length, and the distance between the arc and the aluminum alloy line;

[0090] Step 103, integrating the arc waveform in step 101 to calculate the fault arc energy;

[0091] Step 104, based on the size parameters of the aviation fault arc occurrence location confirmed in step 102, a model is constructed using multi-physics field simulation software, and boundary conditions are set, wherein the heat source value is set based on the arc energy result calculated in step 103;

[0092] Step 105, meshing the model using a custom meshing method, determining the arc fault occurrence area based on the arc movement trajectory and setting it to refinement, and setting the remaining areas to normal;

[0093] Step 106, setting the solver to transient calculation, taking the arc duration in step 101 as the model calculation duration, setting the step length according to the calculation duration, and starting the calculation;

[0094] In one embodiment of the present invention, the solver model includes an energy model and a radiation model, wherein the energy model equation includes:

[0095] ;

[0096] ;

[0097] ;

[0098] in is the air density, is the velocity vector, is the viscous stress, is the momentum source term, is the specific total enthalpy, is the thermal conductivity, is the energy source term.

[0099] The radiation model equations include:

[0100] ;

[0101] in, is the position vector, s is the stroke length, S is the direction vector, is the absorption coefficient, is the scattering coefficient, is the radiation intensity, is the blackbody emission intensity, is the frequency, is the temperature, is the phase function, Ω is the spatial solid angle, and S is the radiation source.

[0102] Step 107, determine whether the calculation result converges. If the iterative calculation error value is less than the set value, it is determined to be converged and the calculation is completed. Combined with the arc movement trajectory, the temperature of the position of the aircraft where the arc passes exceeds the volume of the melting point of the material at that position. If the model calculation does not converge, the model automatically reports an error, then return to steps 104 and 105, reset the boundary conditions and grid size until convergence.

[0103] It should be noted that the iterative calculation error value refers to the difference between the previous calculation result and the next calculation result.

[0104] In one embodiment of the present invention, the arc voltage is 28V DC, the arc current is 15A, the arc starting wire bundle length is 10mm, the wire bundle diameter is 0.813mm, the insulation layer thickness is 0.4mm, the pipeline thickness is 1mm, the pipeline radius is 8.08mm, the pipeline length is 100mm, and the distance between the arc and the aluminum alloy pipeline is 1mm.

[0105] like Figure 2 Shown is a visualization of the calculation results of the arc damage to surrounding components.

[0106] The present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned aviation arc fault damage simulation method is implemented.

[0107] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. A method for simulating aviation arc fault damage, characterized in that: The following steps are involved: Step 100, collecting physical quantity distribution data, medium property data, and arc data of the current time step and each previous time step through multi-physics field simulation software; Step 200, the data collected at the current time step is encoded and combined into first graph structure data, and the data collected at the current time step and each previous time step is encoded and combined into first sequence data; Step 300, inputting the first graph structure data and the first sequence data into the first neural network, outputting the spatial coordinates of the arc at the next time step, the first neural network comprising a first module, a second module, a third module and an output module; The first module is used to take the first graph structure data as input and output the first hidden state of each node in the current time step; The second module is used to take the first sequence data as input and output the second hidden state of the current time step; The third module is used to sum the first hidden state and the second hidden state, and output the third hidden state of each node in the current time step; The output module is used to take the third hidden state as input and output the spatial coordinates of the arc at the next time step; Step 400, connecting the arc endpoint coordinates output by the first neural network at all time steps to obtain the arc movement trajectory; Step 500, judging the damage degree of the aircraft according to the volume of the position of the aircraft where the arc passes through which the temperature exceeds the melting point of the material at the position.

2. The method for simulating aviation arc fault damage according to claim 1, characterized in that: The physical quantity distribution data include: temperature field distribution and electric field distribution; The medium attribute data include: density, conductivity and dielectric constant of various surrounding media; Arc data includes: arc spatial coordinates, voltage and current.

3. The method for simulating aviation arc fault damage according to claim 1, characterized in that: The first graph structure data is generated based on the collected data. The first graph structure data takes the spatial coordinates of the current arc as the central node. The first graph structure data includes nodes and edges. A node contains a set of first data as node features. The first data includes: the spatial coordinates, temperature, electric field vector of the position, and the density, conductivity and dielectric constant of the medium; there is an edge between two adjacent nodes, and the edge represents the distance between the spatial coordinates of the two nodes. The first graph structure data includes the node where the spatial coordinates of the collected arc are located and its adjacent nodes as well as the edges between these nodes.

4. The method for simulating aviation arc fault damage according to claim 1, characterized in that: The medium property data of the space coordinates of the arc obtained in the current time step and each previous time step are collected, and the collected data are encoded into a first sequence data, the first sequence data includes n sequence units, and the tth sequence unit represents the medium property data of the space coordinates of the arc obtained in the tth time step.

5. The method for simulating aviation arc fault damage according to claim 1, characterized in that: The calculation formula for the first module is as follows: ； in represents the first hidden state of the i-th node, represents the state weight parameter, and Represent the node features of the i-th and j-th nodes respectively, represents the set of nodes directly connected to the i-th node, express The total number of nodes in , tanh represents the hyperbolic tangent function, Indicates adjustable parameters; The default value of is 1 / N, where N is the dimension of the node feature. Represents a sigmoid function.

6. The method for simulating aviation arc fault damage according to claim 5, characterized in that: The calculation formula for the second module is as follows: ； ； ； ； in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Represent the first, second, and third intermediate states respectively, where represents the tth sequence unit of the first sequence data, and They represent the t-th and t-1-th second hidden states respectively, n≥t≥1, n represents the total number of sequence units of the first sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

7. The method for simulating aviation arc fault damage according to claim 6, characterized in that: The calculation formula of the third module is as follows: ； in is the nth second hidden state output by the second module, is the first hidden state of the i-th node output by the first module, represents the third hidden state, and N is the set of all nodes in the first graph structure data.

8. The method for simulating aviation arc fault damage according to claim 7, characterized in that: The calculation formula for the output module is as follows: ； Where y represents the first vector of the output, the cth component of the first vector represents the probability that the arc endpoint coordinates at the next time step belong to the cth spatial coordinate, and m represents the third hidden state. is the fourth weight parameter, is the fourth bias parameter, Represents a sigmoid function.

9. The method for simulating aviation arc fault damage according to claim 1, characterized in that: The steps of calculating the volume of the location of the plane through which the arc passes where the temperature exceeds the melting point of the material at that location include: Step 101, determining arc waveform, arc movement trajectory and arc duration; Step 102, confirming the size parameters of the location where the aviation fault arc occurs, including: arc starting wire length, wire diameter, insulation layer thickness, fuel line thickness, line radius, line length, and the distance between the arc and the aluminum alloy line; Step 103, integrating the arc waveform in step 101 to calculate the fault arc energy; Step 104, based on the size parameters of the aviation fault arc occurrence location confirmed in step 102, a model is constructed using multi-physics field simulation software, and boundary conditions are set, wherein the heat source value is set based on the arc energy result calculated in step 103; Step 105, meshing the model using a custom meshing method, determining the arc fault occurrence area based on the arc movement trajectory and setting it to refinement, and setting the remaining areas to normal; Step 106, setting the solver to transient calculation, taking the arc duration in step 101 as the model calculation duration, setting the step length according to the calculation duration, and starting the calculation; Step 107, determine whether the calculation result converges. If the iterative calculation error value is less than the set value, it is determined to be converged and the calculation is completed. Combined with the arc movement trajectory, the temperature of the position of the aircraft where the arc passes exceeds the volume of the melting point of the material at that position. If the model calculation does not converge, the model automatically reports an error, then return to steps 104 and 105, reset the boundary conditions and grid size until convergence.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for simulating aviation arc fault damage as claimed in claim 1 is implemented.

Images