Edge cloud collaborative electro-hydraulic actuator digital twin modeling method and system

Abstract

The invention discloses a side cloud collaborative electro-hydraulic actuator digital twin modeling method and system, and the method comprises the steps: building a high-fidelity simulation model of a target electro-hydraulic actuator through simulation software, and generating simulation data through the simulation model; constructing a proxy model of the operation characteristics of the target electro-hydraulic actuator by using a reduced-order model and deep learning, and deploying the proxy model to edge equipment and a cloud server according to the computing power demand and the input parameter quantity; acquiring operation state data of the target electro-hydraulic actuator in real time; inputting the operation state data into an agent model to perform real-time model reasoning on the edge device and the cloud server so as to predict the physical field operation characteristics of the target electro-hydraulic actuator under the current actual working condition; and carrying out data fusion and visualization processing on the operating characteristics of the physical field to construct a digital twin reflecting the operating state of the target electro-hydraulic actuator. The method has real-time performance and high modeling precision.

Description

Technical Field

[0001] This application belongs to the technical field of electro-hydraulic actuator health monitoring, and specifically relates to a method and system for modeling a digital twin of an electro-hydraulic actuator with edge-cloud collaboration. Background Art

[0002] The electro-hydraulic actuator is a key execution unit in the hydraulic system, which has the advantages of high response speed, high-precision control, and high dynamic performance. It has been widely used in fields such as aerospace, construction machinery, and robotics. Its health status will directly affect the working efficiency and safety of the hydraulic system.

[0003] Traditional health monitoring methods rely on offline data or single sensors, suffering from problems such as insufficient real-time performance and low data utilization rate. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for modeling a digital twin of an electro-hydraulic actuator with edge-cloud collaboration to achieve efficient integration of sensor data and real-time virtual-real interaction under the edge-cloud collaboration architecture.

[0005] According to the first aspect of the embodiments of this application, a method for modeling a digital twin of an electro-hydraulic actuator with edge-cloud collaboration is provided. The modeling method may include:

[0006] Using simulation software to construct a high-fidelity simulation model of the target electro-hydraulic actuator, and generating simulation data using the simulation model;

[0007] Using a reduced-order model and deep learning to construct a surrogate model for the operating characteristics of the target electro-hydraulic actuator, and deploying it on edge devices and cloud servers according to the computing power requirements and the number of input parameters;

[0008] Real-time collecting the operating state data of the target electro-hydraulic actuator;

[0009] Inputting the operating state data into the surrogate model to perform real-time model inference on edge devices and cloud servers to predict the physical field operating characteristics of the target electro-hydraulic actuator under the current actual working conditions;

[0010] Performing data fusion and visualization processing on the physical field operating characteristics to construct a digital twin reflecting the operating state of the target electro-hydraulic actuator.

[0011] In some alternative embodiments of this application, using simulation software to construct a high-fidelity simulation model of the target electro-hydraulic actuator includes:

[0012] Using simulation software to construct a structural stress simulation model and an electromagnetic simulation model of the target electro-hydraulic actuator;

[0013] Using simulation software to construct a fluid simulation model of the target electro-hydraulic actuator;

[0014] Use simulation software to construct a system-level simulation model of the target electro-hydraulic actuator.

[0015] In some alternative embodiments of the present application, a surrogate model of the operating characteristics of the target electro-hydraulic actuator is constructed using a reduced-order model and deep learning, including:

[0016] Use the Latin hypercube sampling method to generate sample points of the simulation working condition variables, and use the simulation model to output the simulation results of the sample points of the simulation working condition variables;

[0017] Use the proper orthogonal decomposition method to construct a reduced-order model for the simulation results, so as to obtain the low-dimensional output corresponding to each sample and form a training data set;

[0018] And use the training data set to train the surrogate model to obtain a trained surrogate model.

[0019] In some alternative embodiments of the present application, use the proper orthogonal decomposition method to construct a reduced-order model for the simulation results, so as to obtain the low-dimensional output corresponding to each sample and form a training data set, including:

[0020] Perform proper orthogonal decomposition on the simulation results;

[0021] Select the singular values and the corresponding singular vectors to construct a reduced-order model;

[0022] Use the reduced-order model to reduce the dimension of the simulation results to obtain a low-dimensional output;

[0023] Construct a training data pair from the simulation results and the corresponding low-dimensional output to obtain a training data set.

[0024] In some alternative embodiments of the present application, the operating state data of the target electro-hydraulic actuator is collected in real time, including:

[0025] Obtain the original time step used when training the surrogate model;

[0026] Use physical sensors to collect the operating state data of the target electro-hydraulic actuator in real time at the original time step.

[0027] In some alternative embodiments of the present application, the operating state data is input into the surrogate model for real-time model inference on the edge device and the cloud server, including:

[0028] Use the edge device to perform inference on the electro-hydraulic actuator system performance model;

[0029] Use the edge device to perform inference on the piston rod structure stress model;

[0030] Use the edge device to perform inference on the force motor electromagnetic model;

[0031] Using a cloud server for hermetic leakage model inference;

[0032] Using a cloud server for solenoid valve fluid model inference.

[0033] In some alternative embodiments of the present application, data fusion and visualization processing are performed on the operating characteristics of the physical field to construct a digital twin reflecting the operating state of the target electro-hydraulic actuator, including:

[0034] Loading a 3D model file and a corresponding vertex information file;

[0035] Loading the inference result of the surrogate model and a corresponding node information file;

[0036] Using a spatial matching algorithm to pair and combine the vertices in the 3D model with the nodes in the inference result of the surrogate model;

[0037] Mapping and normalizing the inference result of the surrogate model to the model vertices to obtain a result value;

[0038] Mapping the result value to a color gradient and recoloring the model using vertex coloring method.

[0039] According to the second aspect of the embodiments of the present application, there is provided an edge-cloud collaborative electro-hydraulic actuator digital twin modeling system, which is characterized by including: a processor, a memory, and a program or instruction stored on the memory and executable on the processor, and when the program or instruction is executed by the processor, it implements the steps of the edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to any one of the embodiments of the first aspect.

[0040] The above technical solution of the present application has the following beneficial technical effects:

[0041] The method of the embodiments of the present application has real-time performance and high modeling accuracy. Through the edge-cloud collaborative computing technology, combining the low-latency processing of edge devices and the powerful computing power of cloud servers, real-time high-precision modeling of the state information of electro-hydraulic actuators is achieved, enabling health monitoring to be completed in a shorter time, improving the response speed and prediction accuracy; and this method has high efficiency and realizes resource optimization. By adopting model reduction and surrogate model technologies, the dependence on traditional simulation models is reduced, and the computational complexity is lowered; at the same time, edge devices and cloud servers work together to optimize the allocation of computing power resources and improve the overall operation efficiency of the monitoring system; in addition, this method has intuitiveness and realizes the construction and visualization of a comprehensive digital twin of electro-hydraulic actuators, intuitively showing the operating state of the device, which is convenient for monitoring and analysis. Description of the Drawings

[0042] Figure 1It is a schematic flowchart of the method for modeling the digital twin of an electro-hydraulic actuator with edge-cloud collaboration in an exemplary embodiment of the present application;

[0043] Figure 2 It is a flowchart of the digital twin modeling method in another exemplary embodiment of the present application;

[0044] Figure 3 It is a flowchart of the surrogate model training method in an exemplary embodiment of the present application;

[0045] Figure 4 It is a flowchart of the surrogate model inference method in an exemplary embodiment of the present application;

[0046] Figure 5 It is a flowchart of the three-dimensional visualization method for inference results in an exemplary embodiment of the present application;

[0047] Figure 6 It is a block diagram of the digital twin operation system in an exemplary embodiment of the present application. Detailed implementation manners

[0048] To make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the detailed implementation manners and with reference to the accompanying drawings. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present application. In addition, in the following descriptions, the descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts of the present application.

[0049] The schematic diagram of the layer structure according to the embodiment of the present application is shown in the accompanying drawings. These figures are not drawn to scale, where for the purpose of clarity, some details are enlarged and some details may be omitted. The shapes of various regions and layers shown in the figures, as well as their relative sizes and positional relationships, are merely exemplary. In practice, there may be deviations due to manufacturing tolerances or technical limitations, and those skilled in the art can design regions / layers with different shapes, sizes and relative positions according to actual needs.

[0050] Obviously, the described embodiments are some but not all of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the scope of protection of the present application.

[0051] In the description of the present application, it should be noted that the terms "first", "second", and "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0052] In addition, the technical features involved in different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0053] Digital twin technology realizes precise simulation and prediction by constructing virtual digital twins, which can reflect the operating state of equipment in real time and become an important means of fault warning. However, constructing and updating digital twins requires processing a large amount of real-time data, and it is difficult for a single edge terminal or cloud platform to meet the requirements of both low latency and high performance at the same time.

[0054] Edge-cloud collaborative computing technology can integrate the low-latency advantage of the edge side and the powerful data processing ability of the cloud, realize real-time high-precision modeling of the state information of electro-hydraulic actuators, and provide a new technical path for health monitoring, fault prediction and system optimization. At present, how to efficiently integrate sensor data under the edge-cloud collaborative architecture, realize real-time virtual-real interaction, and use artificial intelligence algorithms for fault warning still faces challenges.

[0055] Next, in combination with the accompanying drawings, through specific embodiments and their application scenarios, a method and system for modeling a digital twin of an electro-hydraulic actuator with edge-cloud collaboration provided by the embodiments of the present application will be described in detail.

[0056] As Figure 1 shown, in the first aspect of the embodiments of the present application, a method for modeling a digital twin of an electro-hydraulic actuator with edge-cloud collaboration is provided. The modeling method may include:

[0057] S110: Use simulation software to construct a high-fidelity simulation model of the target electro-hydraulic actuator, and generate simulation data using the simulation model;

[0058] S120: Use a reduced-order model and deep learning to construct a surrogate model of the operating characteristics of the target electro-hydraulic actuator, and deploy it on the edge device and the cloud server according to the computing power requirements and the number of input parameters;

[0059] S130: Collect the operating state data of the target electro-hydraulic actuator in real time;

[0060] S140: Input the operating state data into the surrogate model to perform real-time model inference on the edge device and the cloud server to predict the physical field operating characteristics of the target electro-hydraulic actuator under the current actual working conditions;

[0061] S150: Perform data fusion and visualization processing on the physical field operating characteristics to construct a digital twin reflecting the operating state of the target electro-hydraulic actuator.

[0062] The method of this embodiment has real-time performance and high modeling accuracy. Through edge-cloud collaborative computing technology, combining the low-latency processing of edge devices and the powerful computing capabilities of cloud servers, it realizes real-time and high-precision modeling of the state information of electro-hydraulic actuators, enabling health monitoring to be completed in a shorter time, improving the response speed and prediction accuracy; and this method has high efficiency and realizes resource optimization. By using model order reduction and surrogate model technology, it reduces the dependence on traditional simulation models and lowers the computational complexity; at the same time, edge devices and cloud servers work together to optimize the allocation of computing power resources and improve the overall operation efficiency of the monitoring system; in addition, this method has intuitiveness and realizes the construction and visualization of a comprehensive digital twin of electro-hydraulic actuators, intuitively displaying the operating state of the device, which is convenient for monitoring and analysis.

[0063] Specifically, as Figure 2 shown, the digital twin modeling method may include:

[0064] Construct an electro-hydraulic actuator simulation model using simulation software;

[0065] Generate simulation data using the simulation model, construct a surrogate model for the operating characteristics of electro-hydraulic actuators using model order reduction and deep learning techniques, and selectively deploy it on edge devices and cloud servers according to its computing power requirements and the number of input parameters;

[0066] Use physical sensors to collect real-time operation state data of the electro-hydraulic actuator entity;

[0067] Input the sensor data into the surrogate model, perform real-time model inference on the edge device or cloud server, and predict the operating characteristics of each physical field of the electro-hydraulic actuator under the current actual working conditions;

[0068] Perform data fusion and visualization processing on the inference results of the surrogate model on the cloud server side, construct a comprehensive digital twin that comprehensively reflects the key operating states of the electro-hydraulic actuator, and embed it in the monitoring platform.

[0069] Finally, deploy the entire operating system in the on-site monitoring environment to achieve real-time synchronization of the digital twin and the physical device through edge-cloud collaboration.

[0070] In some embodiments, constructing a high-fidelity simulation model of the target electro-hydraulic actuator using simulation software includes:

[0071] Construct a structural stress simulation model and an electromagnetic simulation model of the target electro-hydraulic actuator using simulation software;

[0072] Construct a fluid simulation model of the target electro-hydraulic actuator using simulation software;

[0073] Construct a system-level simulation model of the target electro-hydraulic actuator using simulation software.

[0074] For the structural stress simulation model and electromagnetic simulation model of the electro-hydraulic actuator, simulation software such as ANSYS and ABAQUS can be used to establish them. For the fluid simulation model of the electro-hydraulic actuator, fluid simulation software or modules such as PumpLinx and ANSYS Fluent can be used to establish it. For the system-level simulation model of the electro-hydraulic actuator, AMESim simulation software can be used to establish it. The geometric characteristics and parameter settings of each simulation model must be consistent with the actual electro-hydraulic actuator to ensure the high fidelity and reliability of the model.

[0075] like Figure 3 As shown, in some embodiments, a proxy model of the operating characteristics of a target electro-hydraulic actuator is constructed using a reduced-order model and deep learning, including:

[0076] Use Latin hypercube sampling method to generate simulation condition variable sample points, and use simulation model to output simulation results of simulation condition variable sample points;

[0077] The simulation results are used to construct a reduced-order model using the intrinsic orthogonal decomposition method, thereby obtaining the low-dimensional output corresponding to each sample and forming a training data set;

[0078] And use the training data set to train the proxy model to obtain a trained proxy model.

[0079] In this embodiment, sample points are divided and Latin hypercube sampling method is used to generate simulation condition variable sample points, and a high-fidelity simulation model is used to output the simulation results under these condition sample points; then model reduction and neural network training are performed, and the original output data is used to construct a reduced-order model using the proper orthogonal decomposition method, so as to obtain the low-dimensional output corresponding to each sample, form a training data set, select a suitable proxy model structure, and train the proxy model; next, model accuracy verification is performed, and the accuracy of the trained proxy model is verified, and the error is calculated; finally, the model is deployed, and the proxy model is run on an edge device or a cloud server according to the complexity of the proxy model, the number of input parameters, and the real-time requirements.

[0080] To ensure that the simulation data is complete and available, the simulation condition sample points need to be evenly distributed in the multidimensional space. Therefore, the Latin hypercube sampling method is used to generate the simulation condition variable sample points. The value range of a certain condition variable is divided, and the width of the divided sub-interval is defined by the following formula:

[0081] Δ=(ba) / N

[0082] Where Δ is the width of each subinterval, N is the total sample size, and a and b are the minimum and maximum values ​​of the operating condition variable.

[0083] Randomly select a number in each subinterval to obtain the sampling point of the operating condition variable in this subinterval:

[0084] x i =a+((i-1)+u i )Δ

[0085] In the formula, x i is the sampling point of the parameter in the i-th subinterval, u i A random number uniformly distributed between 0 and 1.

[0086] Similarly, for multidimensional operating variables, the sampling points are defined by the following formula:

[0087] x ij =a j +[(π j (i)-1)+u ij ](b j -a j ) / N

[0088] In the formula, x ij is the sampling point of the i-th sample in the j-th dimension, a j , b j is the minimum and maximum value of the operating variable in the jth dimension, π j is the random permutation generated (i.e., π j (1),π j (2),…,π j (N) is a random permutation from 1 to N), u ij A random number uniformly distributed between 0 and 1.

[0089] Finally, the sampling points of all dimensions are combined into a multidimensional sample vector:

[0090] X i =(x i1 ,x i2 ,...,x im )

[0091] Where, X i is the combined multidimensional sample vector, the dimension of the m-dimensional sample.

[0092] In some embodiments, the simulation results are used to construct a reduced-order model using the intrinsic orthogonal decomposition method, thereby obtaining a low-dimensional output corresponding to each sample to form a training data set, including:

[0093] Perform intrinsic orthogonal decomposition on the simulation results;

[0094] Select singular values ​​and corresponding singular vectors to construct a reduced-order model;

[0095] Use the reduced-order model to reduce the dimension of the simulation results and obtain low-dimensional output;

[0096] The simulation results and the corresponding low-dimensional outputs constitute training data pairs to obtain a training data set.

[0097] Specifically, the sample vector is used to obtain the original output data of the simulation model to form the original output matrix:

[0098] Y=[y 1 y 2 ...y N ] T ∈R N×d

[0099] Where Y is the original output matrix, y is the original output data of the simulation model, and d is the dimension of the high-dimensional output.

[0100] The original output is subjected to intrinsic orthogonal decomposition, that is, the matrix Y is subjected to singular value decomposition:

[0101] Y=U∑V T

[0102] Where U∈R N×N is an orthogonal matrix, ∑∈R N×d is the singular value matrix, V∈R d×d is the right singular vector matrix.

[0103] Select several main singular values ​​and corresponding singular vectors to construct a reduced-order model, thereby reducing the high-dimensional output data to a low-dimensional representation. The low-dimensional output representation of each sample is:

[0104] y i,reduced =[U k ∑ k ] i,: ∈R k ,k=d

[0105] In the formula, y i,reduced is the low-dimensional output of the i-th sample, k is the number of selected main singular values, U k ∈R N×k is the left singular vector matrix, ∑ k ∈R k×k is the singular value matrix, V k ∈R d×k is the right singular vector matrix.

[0106] Input sample And the corresponding low-dimensional output Construct training data pairs, select appropriate proxy model structures, construct mapping relationships and perform interpolation.

[0107] In some embodiments, a long short-term memory network may be used to construct a mapping relationship, which is particularly suitable for situations where real-time prediction is required.

[0108] In other embodiments, a radial basis function network may be used to construct a mapping relationship, which is suitable for situations with a large amount of data and low real-time requirements.

[0109] like Figure 3 As shown, in the proxy model training process, one or more representative working condition samples that do not belong to the sample vector can be selected, and these samples can be input into the trained reduced-order proxy model to obtain the corresponding prediction output; the same working condition samples can be input into the original simulation model to obtain its original simulation results; the prediction output of the proxy model can be compared with the simulation results of the original simulation model to calculate the mean square error and average error.

[0110] In some embodiments, proxy models with a large number of input parameters or high real-time requirements are deployed in edge devices, and proxy models with high computational complexity, large computing power requirements or reliance on complex cloud technology stacks are deployed in cloud servers.

[0111] In some embodiments, real-time acquisition of operating status data of a target electro-hydraulic actuator includes:

[0112] Get the original time step used when training the proxy model;

[0113] Physical sensors are used to collect the operating status data of the target electro-hydraulic actuator in real time at the original time step.

[0114] In order to meet the input requirements of the proxy model, the sampling frequency of the data signal of a certain actual operating state of the electro-hydraulic actuator should be consistent with the inverse of the original time step used in the training of the corresponding proxy model:

[0115] f i =1 / Δt i

[0116] In the formula, f i is the sampling frequency of the i-th state data signal, Δt i The time step used when training the proxy model corresponding to the state data, in seconds.

[0117] like Figure 4 As shown, in some embodiments, the operating state data is input into the agent model to perform real-time model reasoning on the edge device and the cloud server, including:

[0118] Using edge devices to reason about the performance model of electro-hydraulic actuator systems;

[0119] Using edge devices to infer the piston rod structure stress model;

[0120] Use edge devices to perform electromagnetic model inference on force motors;

[0121] Using cloud servers to perform seal leakage model inference;

[0122] Use cloud server to infer the solenoid valve fluid model.

[0123] This embodiment collects key operating status data such as oil pressure, oil temperature, pipeline flow, controller current and voltage of the electro-hydraulic actuator according to the sampling frequency set by the physical sensor in the previous embodiment. At the same time, a connection is established with MQTTBroker to receive current operating status data such as control parameters.

[0124] For proxy models deployed on edge devices, the corresponding collected data is directly used as input parameters for model inference. For proxy models deployed on cloud servers, the corresponding collected data is reported to the MQTT Broker and then inferred by the corresponding proxy model. For proxy models with mutually coupled input and output dependencies, the inference order needs to be determined.

[0125] The system performance model is derived from the AMESim simulation software. The proxy model structure is based on the long short-term memory network, saved in onnx format, and inferred in the edge device through onnxruntime.

[0126] The input of the model is the control parameters of the electro-hydraulic actuator motion amplitude, frequency, number of cycles and external load force, and the output is the displacement of the electro-hydraulic actuator cylinder, the displacement of the solenoid valve core, the driving current of the electromagnetic motor, the inner cavity pressure, the outer cavity pressure, and the force acting on the solenoid valve core.

[0127] The piston rod structural stress model is derived from the ANSYS simulation software. The proxy model structure is based on a radial basis function network, saved in pkl format, and inferred in the edge device through the joblib library.

[0128] The input of the model is the external load force in the control parameters, as well as the displacement of the electro-hydraulic actuator cylinder and the inner cavity pressure. The output is the stress of the internal node of the piston rod, the stress of two measuring points on the cylinder wall and the stress at the piston rod earring.

[0129] The electromagnetic model of the force motor is derived from the electromagnetic module of the ANSYS simulation software. The proxy model structure is based on the radial basis function network, saved in pkl format, and inferred in the edge device through the joblib library.

[0130] The input of the model is the collected magnetic flux of the permanent magnet of the linear force motor, the driving current, the external load force in the control parameters, and the displacement of the actuator cylinder of the electro-hydraulic actuator. The output is the magnetic induction intensity of the internal nodes of the linear force motor and the output force of the linear force motor.

[0131] The seal leakage model is derived from the MATLAB Simulink simulation software. The surrogate model structure is based on a radial basis function network and is saved in the fmu format. Inference is performed through the fmpy library on a cloud server.

[0132] The inputs of this model are the amplitude, frequency, number of cycle periods of the electro-hydraulic actuator movement in the control parameters, the external load force, and the external cavity pressure output by the surrogate model. The outputs are the dynamic seal wear amount and the leakage rate.

[0133] The solenoid valve fluid model is derived from the ANSYS Fluent simulation module. The surrogate model structure is based on a radial basis function network. Its weight matrix, coordinate matrix, and correction matrix are saved in the csv format. Inference is performed through a specific JavaScript script on a cloud server.

[0134] The inputs of this model are the amplitude, frequency, number of cycle periods of the electro-hydraulic actuator movement in the control parameters, and the collected oil temperature. The outputs are the internal node pressure and the internal node temperature of the solenoid valve.

[0135] As Figure 5 shown, in some embodiments, data fusion and visualization processing are performed on the operating characteristics of the physical field to construct a digital twin reflecting the operating state of the target electro-hydraulic actuator, including:

[0136] Loading the 3D model file and the corresponding vertex information file;

[0137] Loading the inference results of the surrogate model and the corresponding node information file;

[0138] Using a spatial matching algorithm to pair and combine the vertices in the 3D model with the nodes in the inference results of the surrogate model;

[0139] Mapping and normalizing the inference results of the surrogate model to the model vertices to obtain the result values;

[0140] Mapping the result values to a color gradient and recoloring the model using vertex coloring method.

[0141] In this embodiment, model files in formats such as fbx, stl, and gltf can be used. The corresponding vertex information files can be saved in formats such as csv and json. The vertex information file contains the three-dimensional spatial coordinates of all vertices of the 3D model.

[0142] For 3D model files with overly sparse vertices, subdivision surface processing is also required to ensure effective mapping of information. Specifically, the number of vertices of the 3D model file should be greater than the number of nodes corresponding to the output results of the surrogate model.

[0143] The inference results of the proxy model can be from the proxy model deployed on the edge device or from the proxy model deployed on the cloud server. The corresponding node information file can be saved in formats such as csv and json. The node information file contains the three-dimensional spatial coordinates of all grid nodes divided during the simulation.

[0144] If the vertex coordinates of the three-dimensional model and the grid node coordinates are stored in different length units, all coordinate values in the corresponding node information file need to be multiplied by the corresponding scale factor to convert to the same length unit.

[0145] For the vertex information file and the node information file with the same length unit, extract all their coordinate values respectively to create point cloud objects, and use the iterative closest point algorithm to align the node point cloud to the vertex point cloud.

[0146] Use the nearest neighbor search algorithm to match the nearest several nodes for each vertex.

[0147] In some embodiments, the number of matched nodes is 1.

[0148] In some other embodiments, the number of matched nodes is defined by the following formula:

[0149]

[0150] In the formula, n is the number of matched nodes, V is the total number of vertices, N is the total number of nodes, is the ceiling operation.

[0151] According to the matched nodes, map the inference results of the proxy model to all vertices of the three-dimensional model, that is, assign calculated values to all vertices of the three-dimensional model. The specific values are defined by the following formula:

[0152]

[0153] In the formula, Vertex i is the calculated value assigned to the vertex of the three-dimensional model, Node j is the value of the corresponding node of the inference result of the proxy model, n is the number of matched nodes, and i, j are the indices of the vertex and the node.

[0154] Map the calculated values assigned to all vertices to a color gradient. For example, map low values to cold colors and high values to warm colors, and use the vertex coloring method to repaint the model.

[0155] The digital twin operation system consists of monitoring sensors, edge computing devices, and cloud servers in the electro-hydraulic actuator system. The operation system for building the electro-hydraulic actuator digital twin is as Figure 6 .

[0156] In an edge computing node, an edge computing device has modules such as a CPU, network port communication, and WLAN communication to implement communication functions between the edge device and the AD conversion module, monitoring signal processing and proxy model inference functions, and communication functions between the edge device and the cloud server. The cloud server and the edge computing device are connected through the MQTT protocol.

[0157] It should be noted that for the edge-cloud collaborative digital twin modeling method of the electro-hydraulic actuator provided in the embodiments of the present application, the execution subject can be an edge-cloud collaborative digital twin modeling device of the electro-hydraulic actuator, or a control module in the edge-cloud collaborative digital twin modeling device of the electro-hydraulic actuator for executing the edge-cloud collaborative digital twin modeling method of the electro-hydraulic actuator. In the embodiments of the present application, taking the edge-cloud collaborative digital twin modeling device of the electro-hydraulic actuator executing the edge-cloud collaborative digital twin modeling method of the electro-hydraulic actuator as an example, the edge-cloud collaborative digital twin modeling device provided in the embodiments of the present application is described.

[0158] In the second aspect of the embodiments of the present application, an edge-cloud collaborative digital twin modeling system of an electro-hydraulic actuator is provided. The system includes: a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the edge-cloud collaborative digital twin modeling method described in any one of the embodiments of the first aspect are implemented.

[0159] The edge-cloud collaborative digital twin modeling system in the embodiments of the present application can be a device, or a component, an integrated circuit, or a chip in a terminal. The system can be a mobile electronic device or a non-mobile electronic device. Exemplarily, the mobile electronic device can be a mobile phone, a tablet computer, a laptop computer, a handheld computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc. The non-mobile electronic device can be a server, a Network Attached Storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc. The embodiments of the present application do not make specific limitations.

[0160] The edge-cloud collaborative digital twin modeling system in the embodiments of the present application can be a device with an operating system. The operating system can be the Android operating system, the iOS operating system, or other possible operating systems. The embodiments of the present application do not make specific limitations.

[0161] The edge-cloud collaborative digital twin modeling system of the electro-hydraulic actuator provided by the embodiment of the present application can achieve Figure 1 each process implemented by the method embodiment. To avoid repetition, it will not be elaborated here.

[0162] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms without departing from the purpose of the present application and the scope protected by the claims, and all of them belong to the protection scope of the present application.

Claims

1. A digital twin modeling method for an edge-cloud-coordinated electro-hydraulic actuator, characterized in that: include: Using simulation software to build a high-fidelity simulation model of the target electro-hydraulic actuator, and using the simulation model to generate simulation data; Using reduced-order models and deep learning to build a proxy model of the operating characteristics of the target electro-hydraulic actuator, and deploy it on edge devices and cloud servers according to computing power requirements and input parameter quantities; Collecting the operating status data of the target electro-hydraulic actuator in real time; Inputting the operating status data into the proxy model to perform real-time model reasoning on the edge device and the cloud server to predict the physical field operating characteristics of the target electro-hydraulic actuator under the current actual working conditions; Data fusion and visualization processing are performed on the physical field operation characteristics to construct a digital twin reflecting the operation status of the target electro-hydraulic actuator.

2. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 1 is characterized in that: The method of using simulation software to construct a high-fidelity simulation model of the target electro-hydraulic actuator includes: Use simulation software to build a structural stress simulation model and an electromagnetic simulation model of the target electro-hydraulic actuator; Use simulation software to build a fluid simulation model of the target electro-hydraulic actuator; A system-level simulation model of the target electro-hydraulic actuator is constructed using simulation software.

3. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 1 is characterized in that: The method of constructing a proxy model of the target electro-hydraulic actuator operating characteristics by using a reduced-order model and deep learning includes: Generate simulation operating condition variable sample points using Latin hypercube sampling method, and output simulation results of the simulation operating condition variable sample points using simulation model; The simulation results are used to construct a reduced-order model using the intrinsic orthogonal decomposition method, so as to obtain a low-dimensional output corresponding to each sample to form a training data set; The training data set is used to train the proxy model to obtain a trained proxy model.

4. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 3 is characterized in that: The simulation results are used to construct a reduced-order model using the intrinsic orthogonal decomposition method, thereby obtaining a low-dimensional output corresponding to each sample to form a training data set, including: Performing intrinsic orthogonal decomposition on the simulation results; Select singular values ​​and corresponding singular vectors to construct a reduced-order model; Using the reduced-order model to reduce the dimension of the simulation result to obtain a low-dimensional output; The simulation result and the corresponding low-dimensional output constitute a training data pair to obtain a training data set.

5. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 1 is characterized in that: The real-time acquisition of the operating status data of the target electro-hydraulic actuator includes: Obtaining the original time step used when training the proxy model; The operating state data of the target electro-hydraulic actuator is collected in real time using a physical sensor at the original time step.

6. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 1 is characterized in that: The step of inputting the operating status data into the proxy model to perform real-time model reasoning on the edge device and the cloud server includes: Using the edge device to perform performance model reasoning of an electro-hydraulic actuator system; Using the edge device to infer a piston rod structure stress model; Performing force motor electromagnetic model inference using the edge device; Using the cloud server to perform seal leakage model reasoning; The cloud server is used to perform solenoid valve fluid model reasoning.

7. The edge-cloud collaborative electro-hydraulic actuator digital twin modeling method according to claim 1 is characterized in that: The data fusion and visualization processing of the physical field operation characteristics to construct a digital twin reflecting the operation state of the target electro-hydraulic actuator includes: Load the 3D model file and the corresponding vertex information file; Loading the inference results of the proxy model and the corresponding node information file; Use a spatial matching algorithm to pair and combine vertices in the 3D model with nodes in the proxy model inference result; Map and normalize the proxy model inference results to the model vertices to obtain the result values; The resulting values ​​are mapped to a color gradient and the model is recolored using a vertex coloring method.

8. An edge-cloud collaborative electro-hydraulic actuator digital twin modeling system, characterized in that: include: A processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the edge-cloud collaborative electro-hydraulic actuator digital twin modeling method as described in any one of claims 1 to 7 are implemented.

Images