An equipment digital twin dynamic damage assessment method and system based on multi-physics field joint simulation

Abstract

The application provides a kind of equipment digital twin dynamic damage assessment method and system based on multi-physical field joint simulation, related to digital twin and computer simulation technical field.The method comprises: constructing the collaborative simulation middleware of mapping multi-physical field digital twin model;Access real-time telemetry data, drive distributed multi-physical field solver node to execute joint simulation deduction;Monitor node interaction message state, identify the target solver node of message interaction exception;Block the original publishing authority of the target solver node and redirect the output pointer to the isolated memory area;Broadcast rollback instruction to the remaining normally operating nodes to backtrack to the nearest steady-state synchronization checkpoint;Load virtual degradation agent process;Bind the virtual degradation agent process to the communication topic channel defined for the target solver node in the interface control file, output degraded substitute data, maintain joint simulation deduction and output evaluation results.The application ensures the continuity of equipment damage cascade assessment deduction.

Claims

1. A method for dynamic damage assessment of equipment based on multiphysics co-simulation digital twins, characterized in that, include: Parse the interface control files of the equipment to be evaluated and construct a co-simulation middleware that maps a multi-physics digital twin model; The collaborative simulation middleware accesses real-time telemetry data and drives distributed multiphysics solver nodes to perform joint simulation and deduction. Monitor the interaction message status of the multiphysics solver nodes in the co-simulation middleware, and identify the target solver nodes with abnormal message interactions; In the co-simulation middleware, the original publishing permissions of the target solver node are blocked, and the output pointer of the target solver node is redirected to an isolated memory region; A rollback command is broadcast to the remaining multiphysics solver nodes that are operating normally, driving the remaining multiphysics solver nodes to backtrack to the most recent steady-state synchronization checkpoint; In the co-simulation middleware, a virtual degradation proxy process matching the target solver node is dynamically loaded; The virtual degradation agent process is bound to the communication topic channel defined for the target solver node in the interface control file. The degradation replacement data is output to the remaining multiphysics solver nodes through the virtual degradation agent process, maintaining the joint simulation and outputting the cascade damage assessment results.

2. The equipment digital twin dynamic damage assessment method based on multiphysics co-simulation according to claim 1, characterized in that, The interface control file of the equipment to be evaluated is parsed, and a co-simulation middleware for constructing a multiphysics digital twin model is built, including: Extract the sensor parameter identifiers, transmit / receive direction attributes, and bus communication protocol types defined in the interface control file; Allocate a global data exchange space in the system memory of the host server, and divide the global data exchange space into multiple topic channels based on the data distribution service protocol; Configure the underlying transmission mechanism of the multiple topic channels using the bus communication protocol type, and register the sensor parameter identifier and the transmit / receive direction attribute to the corresponding topic channel to generate the collaborative simulation middleware.

3. The equipment digital twin dynamic damage assessment method based on multiphysics co-simulation according to claim 2, characterized in that, The step of accessing real-time telemetry data through the collaborative simulation middleware and driving distributed multiphysics solver nodes to perform joint simulation deduction includes: The original communication data frames sent by the equipment to be evaluated are captured using a network bus adapter, and the original communication data frames are unpacked according to the interface control file to obtain the numerical real-time telemetry data. Write the real-time telemetry data into the topic channel corresponding to the collaborative simulation middleware; The multiphysics solver node includes at least one or more combinations of structural dynamics solvers, heat conduction solvers, and embedded control logic solvers. The multiphysics solver node subscribes to and reads the real-time telemetry data from the associated topic channel and performs step-by-step calculations based on a multi-threaded concurrency mechanism.

4. The equipment digital twin dynamic damage assessment method based on multiphysics co-simulation according to claim 3, characterized in that, The process of performing the joint simulation also includes: A circular snapshot buffer queue is set up within the co-simulation middleware; After each preset number of step calculation cycles, the set of internal state parameters output by the multiphysics solver node is collected; Calculate the checksum of the internal state parameter set, and use the internal state parameter set with timestamp and checksum as the most recent steady-state synchronization checkpoint, and push it into the circular snapshot buffer queue for rolling storage.

5. The method for dynamic damage assessment of equipment digital twins based on multiphysics co-simulation according to claim 1, characterized in that, The monitoring of the interaction message status of the multiphysics solver nodes in the co-simulation middleware, and the identification of target solver nodes with abnormal message interactions, includes: Bypass sniffing plugins are mounted on each topic channel of the co-simulation middleware; The bypass sniffing plugin is used to read the payload data and arrival time interval of the interactive messages; When the load data is read as a non-abnormal value for a preset number of consecutive preset number of times, or when the arrival time interval exceeds the preset heartbeat timeout threshold, the corresponding multiphysics solver node is determined to be in an abnormal state. The solver node associated with the abnormal status is marked as the target solver node.

6. The method for dynamic damage assessment of equipment digital twins based on multiphysics co-simulation according to claim 4, characterized in that, The step of blocking the original publishing permissions of the target solver node in the co-simulation middleware and redirecting the output pointer of the target solver node to an isolated memory region includes: Delete the media access control address binding record corresponding to the target solver node from the routing control table of the co-simulation middleware; Request the allocation of an independent sandbox buffer as the isolated memory region from the virtual memory manager of the host operating system; The underlying write system call issued by the target solver node is intercepted, the target memory address in the call parameters is modified, and the output data of the target solver node is forcibly redirected to the independent sandbox buffer to prevent the output data of the target solver node from entering the global data exchange space.

7. The equipment digital twin dynamic damage assessment method based on multiphysics co-simulation according to claim 4, characterized in that, The step of broadcasting a rollback command to the remaining multiphysics solver nodes that are operating normally, driving the remaining multiphysics solver nodes to backtrack to the most recent steady-state synchronization checkpoint, includes: Obtain the time value at which the anomaly was determined as the anomaly timestamp; Generate the rollback instruction that includes the abnormal timestamp and the global interrupt signal; The rollback instruction is sent to the remaining multiphysics solver nodes by controlling the message queue, thus pausing the current step-by-step solution process. Starting from the tail pointer of the circular snapshot buffer queue, traverse backwards to extract the set of internal state parameters that are located before the abnormal timestamp and whose checksum verification has passed; The extracted set of internal state parameters is overwritten into the runtime memory space of the remaining multiphysics solver nodes.

8. The method for dynamic damage assessment of equipment digital twins based on multiphysics co-simulation according to claim 1, characterized in that, The dynamic loading of a virtual degradation proxy process matching the target solver node in the co-simulation middleware includes: Pre-compile multiple sets of dynamic link library files corresponding to the degradation behavior of different physical components; Extract the component identifier code corresponding to the target solver node, and use the component identifier code to retrieve the matching dynamic link library file of the degradation behavior in the model database; The system's underlying application programming interface is invoked to load the matching dynamic link library file for the degradation behavior into the runtime memory and instantiate it into the controlled virtual degradation agent process; Allocate a new logical handle to the virtual degradation agent process and initiate a node registration request to the co-simulation middleware.

9. The method for dynamic damage assessment of equipment digital twins based on multiphysics co-simulation according to claim 1, characterized in that, Maintaining the joint simulation and outputting the cascade damage assessment results includes: After receiving the downgraded replacement data, the remaining multiphysics solver node continues to perform nonlinear iterative solution based on the changes in associated boundary conditions caused by the downgraded replacement data. Record the performance degradation rate and the list of secondary fault nodes generated by the remaining multiphysics solver nodes after receiving the downgraded replacement data; By combining the initial disconnection state when the target solver node malfunctions, the performance index decay rate, and the list of secondary fault nodes, a three-dimensional damage propagation topology map is generated. Based on the historical data of the damaged nodes in the secondary fault node list and fatigue damage theory, the predicted value of the system's residual working life is calculated. The cascade damage assessment results are output by packaging the three-dimensional damage propagation topology map and the predicted residual working life of the system.

10. A digital twin dynamic damage assessment system for equipment based on multiphysics co-simulation, characterized in that, include: The middleware building module is used to parse the interface control files of the equipment to be evaluated and build a co-simulation middleware that maps a multi-physics digital twin model. The simulation driving module is used to access real-time telemetry data through the collaborative simulation middleware and drive distributed multiphysics solver nodes to perform joint simulation simulation. An anomaly monitoring module is used to monitor the interaction message status of the multiphysics solver nodes in the co-simulation middleware and identify the target solver node with anomaly in message interaction. An isolation redirection module is used to block the original publishing permissions of the target solver node in the co-simulation middleware and redirect the output pointer of the target solver node to an isolated memory region; The synchronous rollback module is used to broadcast rollback instructions to the remaining multiphysics solver nodes that are running normally, driving the remaining multiphysics solver nodes to backtrack to the most recent steady-state synchronization checkpoint. The proxy loading module is used to dynamically load virtual degradation proxy processes that match the target solver node in the co-simulation middleware; The takeover evaluation module is used to bind the virtual degradation agent process to the communication topic channel defined for the target solver node in the interface control file, and output degradation replacement data to the remaining multiphysics solver nodes through the virtual degradation agent process, thereby maintaining the joint simulation and outputting the cascade damage evaluation results.

Images