Ship physical test-bed-integrated platform two-way communication method based on test-bed agent

Abstract

This invention discloses a bidirectional communication method and system for a ship's physical test bench and integrated platform based on a test bench agent. The method includes: establishing a bidirectional communication channel between the physical test bench and the integrated platform using a test bench agent as a relay hub, and receiving standardized instructions with verification from the integrated platform. The test bench agent encodes the unique identifier of the test bench into a finite-field projective geometric point set, generating dedicated projective coordinates to achieve collision-free addressing and precise distribution of parallel instructions. Simultaneously, it extracts the topologically invariant features of the original serial port data, constructs a topological adjacency graph, completes out-of-order frame rearrangement through topological isomorphism determination, and completes missing frames based on completeness constraints, achieving timing recovery without external clocks or threshold weights. After protocol conversion, multi-layer verification, and manual confirmation, the instructions are forwarded, and the processed return data is encapsulated and uploaded, forming a complete communication loop.

Description

Technical Field

[0001] This invention belongs to the field of ship bench proxy testing technology, and more specifically, relates to a two-way communication method and system for ship physical bench-integrated platform based on bench proxy. Background Technology

[0002] Currently, communication between ship physical test benches and integrated platforms generally suffers from protocol heterogeneity. The incompatible communication protocols used by the two lead to barriers in data interaction, making it impossible to achieve smooth two-way communication, which in turn affects the efficiency of command issuance and data feedback.

[0003] Meanwhile, in existing communication methods, parallel instruction addressing is prone to conflicts. When multiple devices are running in parallel, instruction distribution is easily confused, making it impossible to accurately distinguish the target device, resulting in instruction execution deviations and affecting the normal operation of ship equipment. In addition, during serial port transmission, data is prone to problems such as frame out-of-order delivery and frame loss. Traditional timing recovery methods rely on external clocks or manual intervention, which are not only cumbersome to operate but also prone to recovery deviations, resulting in a lack of data integrity.

[0004] Furthermore, existing technologies lack reliable structural constraints and mathematical support, making them prone to data loss and mistransmission of commands during communication. This prevents the formation of a complete communication loop, severely impacting the management efficiency of shipboard physical test benches and hindering the intelligent and efficient development of shipboard equipment. Therefore, there is an urgent need for a communication method that can resolve issues such as protocol heterogeneity, addressing conflicts, and data timing recovery, enabling stable, accurate, and efficient two-way communication between the physical test bench and the integrated platform. This will ensure the reliability and continuity of shipboard equipment management, promote the upgrading and optimization of communication technologies for shipboard equipment, and has significant practical implications and application value. Summary of the Invention

[0005] This invention aims to solve problems such as protocol heterogeneity, parallel instruction addressing conflicts, and out-of-order serial data frame loss in communication between ship physical test benches and integrated platforms, and to build a stable and efficient two-way communication system to achieve accurate instruction distribution and reliable data transmission, thereby improving the overall operational efficiency and safety of ship equipment management.

[0006] To address the aforementioned deficiencies or improvement needs of existing technologies, as a first aspect of this invention, the present invention provides a bidirectional communication method between a ship's physical test bench and an integrated platform based on a test bench agent, comprising: S1. Using the bench agent as a transit hub, a bidirectional communication channel for heterogeneous protocol adaptation is established between the physical bench and the integration platform. The integration platform sends standardized instructions carrying verification information to the bench agent. S2. The bench agent encodes the unique identifier of the bench as a set of projective geometric points in a finite field. It generates instruction-specific projective coordinates through a projective transformation matrix. It utilizes the non-intersecting property of the projective space to achieve collision-free addressing of parallel instructions, automatically distinguishes the target bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from a mathematical structure perspective. S3. The benchtop agent extracts the topology-invariant features of the data stream of the original serial port data, constructs a data frame topology adjacency graph, completes the optimal rearrangement of out-of-order frames through topology isomorphism determination, and performs structural completion of missing frames based on topology space completeness constraints, thereby achieving data timing recovery under the conditions of no external clock and no threshold weight. S4. After completing protocol conversion, multi-layer verification, and manual confirmation, the bench agent forwards the instruction to the physical bench for execution and uniformly encapsulates and synchronously uploads the processed return data to the integration platform, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

[0007] Furthermore, the process of constructing the bidirectional communication channel in S1 is as follows: The rack agent is deployed on the host computer of the physical rack. The rack agent establishes a communication link with the physical rack that is compatible with multiple serial port protocols. The rack agent establishes a full-duplex persistent connection with the integration platform. The rack agent completes identity registration and topic subscription with the message middleware. The bench agent generates a unique identifier during the connection establishment process. This unique identifier is directly used for finite field projective geometry encoding. The bench agent monitors the connection status in real time and initiates autonomous reconstruction when interrupted, thereby building a bidirectional communication channel that can support subsequent addressing and data processing.

[0008] Furthermore, the process of generating instruction-specific projective coordinates through the projective transformation matrix in S2 is as follows: Using a preset finite field of prime powers as the basis for the overall projective operation, the bench agent determines the feature vector dimension based on the maximum number of benches that the system can support, and generates the feature vector of the corresponding dimension by binary digit encoding the unique identifier of the bench. The benchtop agent constructs a full-rank projective transformation matrix of the same dimension within the finite field. The construction method is to select linearly independent column vectors in the finite field, and combine them after orthogonalization to form a matrix, ensuring that the determinant of the matrix is ​​non-zero. This projective transformation matrix is ​​generated during the system initialization phase and used globally. The initial projective coordinates are obtained by multiplying the eigenvectors by the projective transformation matrix. Then, the initial projective coordinates are orthogonalized and normalized in a finite field. Combined with the generated historical addressing coordinates, iterative calculations are completed to finally obtain the final addressing coordinates that meet the linear independence requirement.

[0009] Furthermore, the process of achieving collision-free addressing of parallel instructions using the non-intersecting property of projective spaces in S2 is as follows: During the system initialization phase, the benchtop agent, based on the defined finite field... With the dimension of the feature vector ,Will The finite field projective space is decomposed into orthogonal subspaces according to the number of racks, dividing the entire projective space into a set of orthogonal subspaces that correspond one-to-one with the racks, with each subspace corresponding to the instruction addressing domain of a rack. The partitioning of subspaces satisfies orthogonality and mutual exclusivity, meaning that any two different pedestals correspond to the same subspace. and They satisfy an orthogonal relation And the intersection is empty. This ensures that there are no overlapping addressing regions between subspaces; The rack agent will provide the final addressing coordinates for each rack. Mapped to a dedicated subspace, each bench's instructions are transmitted and recognized only within its assigned subspace. When the integration platform issues parallel instructions, the bench agent extracts the final addressing coordinates carried by the instructions. The target subspace is determined by spatial location, and then the command is forwarded to the physical platform corresponding to that subspace. Since the subspaces are orthogonal and do not overlap, instructions from different racks have independent transmission channels in the projective space, avoiding cross-domain interference and addressing conflicts, and ensuring the stable distribution and execution of parallel instructions from multiple racks from the spatial structure level.

[0010] Furthermore, the process of constructing the data frame topology adjacency graph in S3 is as follows: After the benchtop agent performs frame splitting and format verification on the raw serial port data, it maps the various core features of each frame of data to a finite field of a preset prime power. Based on this, it constructs a fixed-dimensional topological feature vector corresponding to each data frame. The benchtop agent uses each topological feature vector as a node to form a node set. Based on the cyclic shift invariance of finite fields, a fixed-order unit cyclic matrix with the function of cyclic right shift of vectors is defined. The topological adjacency relationship between nodes is constructed through this matrix and its power transformation. For any two nodes, if there exists a unique positive integer as the number of cyclic shift steps, such that one vector is completely identical to the other vector after the corresponding number of cyclic shift transformations, then it is determined that the data frames corresponding to the two vectors have a topological adjacency relationship and an undirected edge is established, thus forming an edge set. The final data frame topology adjacency graph is composed of the node set and the edge set.

[0011] Furthermore, the process of achieving optimal rearrangement of out-of-order frames through topological isomorphism determination in S3 is as follows: The benchtop agent is based on the constructed data frame topology adjacency graph. ,in For node set, It is a set of edges, where each node corresponds to one edge. 3D topological eigenvectors ,and , For a predefined finite field, It is a prime power; First, extract the topological invariants of the graph, including the node degree sequence. Adjacency matrix and characteristic spectrum ;Node degree sequence The adjacency matrix is ​​composed of the number of adjacent edges of each node. Defined in Above, the element values ​​represent the adjacency relationships between nodes, and the feature spectrum... Adjacency matrix The set of eigenvalues, the above invariant set, uniquely characterizes the structural features of the topological graph and is not affected by the order of node arrangement; Benchtop agent pre-stores the baseline topology adjacency graph corresponding to the standard data stream , and its benchmark invariant set Topology graph constructed from out-of-order data frames Iterate through all possible permutations and substitutions of nodes. Construct the permutation matrix And calculate the adjacency matrix after the permutation. ,in For corresponding permutations The permutation matrix, dimension and Consistent; If a unique permutation exists , so that the adjacency matrix after the permutation Adjacency matrix with reference Completely identical, and characteristic spectrum Compared with the benchmark characteristic spectrum If a match is found, the topology graph is determined. Compared with the baseline topology diagram Topological isomorphism, this permutation This refers to the optimal rearrangement order of out-of-order frames; the gantry proxy is based on the permutation... The out-of-order data frames are rearranged to restore the original out-of-order frame sequence to a standard sequence consistent with the reference data stream, thus completing the optimal rearrangement of out-of-order frames.

[0012] Furthermore, the process of structural completion of missing frames based on topological space completeness constraints in S3 is as follows: The benchtop agent uses the rearranged data frame topology adjacency graph as its structural basis, within a finite domain. Under constraints, perform a completeness check on the topological space by traversing the topological feature vectors corresponding to all nodes. The system identifies structural breakpoints in the adjacency matrix that contain empty connections, discontinuous degree sequences, and unclosed feature spectra, and marks the location of these breakpoints as the location of missing data frames. For each marked structural breakpoint, the benchtop agent extracts the feature vectors of the existing adjacent nodes on both sides of the breakpoint. and Based on the rules of linear space operations in finite fields, the topological feature vector of the missing frame is calculated. The operation process follows The closure property of addition guarantees a unique and stable result. Based on the calculated topological feature vector The benchtop agent reverse maps and generates the corresponding data frame structure, including frame header identifier, data length, checksum field and business association number, fully restores all valid information of the missing frame, and inserts the completed node into the corresponding position in the topology adjacency relationship graph. After completion, the benchtop agent updates the adjacency matrix and feature spectrum to ensure that the overall topology meets the completeness condition. This enables the complete structure recovery of the data stream under conditions without external clock or weight threshold, effectively solving the problem of incomplete data caused by frame loss and frame breakage in serial transmission, and ensuring that the uplink data timing is continuous and the logic is reliable.

[0013] Furthermore, the process by which the S3 platform agent completes the protocol conversion is as follows: The benchtop agent parses the protocol fields of the complete data stream after timing recovery and structural completion, and rearranges and aligns the non-standard data format transmitted by the serial port physical layer according to the unified communication specifications of the integration platform. benchtop agent based on finite domain The data payload is uniformly quantized and mapped, and the multi-type protocol fields of different physical platforms are converted into standardized data units of the same dimension, so that all kinds of heterogeneous data have a consistent transmission format and verification basis. During the conversion process, the benchtop agent retains the original topology feature identifier and business association information of the data frame without changing the meaning of the data itself. It only completes the protocol format normalization process. The converted data meets the reception standards of the integration platform and can be directly used for encapsulation and uploading.

[0014] As a second aspect of the present invention, a two-way communication system for a ship physical test bench-integrated platform based on a test bench agent is also provided, comprising: The bidirectional communication channel construction unit is used to establish a bidirectional communication channel between the physical test bench and the integration platform with the test bench agent as the relay hub. The integration platform sends standardized instructions carrying verification information to the test bench agent. The conflict-free parallel distribution unit is used by the test bench agent to encode the unique identifier of the test bench into a set of finite field projective geometric points, generate instruction-specific projective coordinates through the projective transformation matrix, and realize parallel instruction collision-free addressing by utilizing the non-intersecting property of the projective space. It automatically distinguishes the target test bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from the mathematical structure. The topology timing recovery and completion unit is used to extract the topology invariant features of the data stream of the original serial port data by the benchtop agent, construct the topology adjacency relationship graph of the data frame, complete the optimal rearrangement of out-of-order frames through topology isomorphism determination, and perform structural completion of missing frames based on topology space completeness constraints, so as to realize data timing recovery under the conditions of no external clock and no threshold weight. The end-to-end communication unit is used to forward instructions to the physical benchtop after the benchtop agent completes protocol conversion, multi-layer verification, and manual confirmation. It also encapsulates the processed return data and uploads it to the integration platform in a unified and time-synchronized manner, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

[0015] As a third aspect of the invention, a computer-readable storage medium is also provided, on which a computer program is stored, which is executed by a processor, according to any one of the claims, a two-way communication method for a ship physical bench-integrated platform based on bench agent.

[0016] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: 1. The present invention provides a bidirectional communication method between a ship's physical test bench and an integrated platform based on a test bench agent. This method establishes a bidirectional communication channel between the physical test bench and the integrated platform, adapting to heterogeneous protocols, by using a test bench agent as a relay hub. The integrated platform sends standardized commands carrying verification information to the test bench agent. Simultaneously, the test bench agent generates a unique identifier during connection establishment, monitors the connection status in real time, and initiates autonomous reconstruction upon interruption, ensuring the stability and continuity of the communication channel. This setup effectively solves the communication barrier problem caused by protocol heterogeneity between the physical test bench and the integrated platform, realizing bidirectional data interaction between the two. It provides a reliable transmission foundation for subsequent command distribution, data processing, and feedback, ensuring smooth command issuance and data return, and avoiding communication interruptions or data loss due to protocol incompatibility.

[0017] 2. The present invention provides a bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench proxy. This method encodes the unique identifier of each test bench as a set of finite-field projective geometric points through a test bench proxy. It generates instruction-specific projective coordinates using a projective transformation matrix and leverages the non-intersecting nature of projective spaces to achieve collision-free addressing of parallel instructions. It automatically distinguishes target test benches and completes instruction distribution. Based on the mathematical properties of finite-field projective geometry, this technology eliminates conflicts and addressing errors in parallel instructions from a mathematical structure perspective, solves the instruction interference problem in multi-test bench parallel communication, ensures that multiple physical test benches can simultaneously receive and execute corresponding instructions, improves the efficiency and accuracy of instruction distribution, and guarantees communication stability during multi-test bench parallel operation.

[0018] 3. The present invention provides a bidirectional communication method for a ship physical test bench-integrated platform based on a test bench agent. This method extracts the topology-invariant features of the raw serial port data through a test bench agent, constructs a data frame topology adjacency graph, performs optimal rearrangement of out-of-order frames through topology isomorphism determination, and completes the structure of missing frames based on topology space completeness constraints. This achieves data timing recovery under conditions without external clocks or threshold weights. After protocol conversion, multi-layer verification, and manual confirmation, the method completes command forwarding and data encapsulation and uploading, forming a closed-loop communication process. This technology effectively solves the problems of data timing disorder and incomplete data caused by out-of-order frames and frame loss in serial port transmission, achieving accurate data recovery and reliable transmission. Furthermore, the closed-loop design ensures the integrity of command issuance, execution, and feedback, improving the overall reliability and security of communication. Attached Figure Description

[0019] Figure 1 This is a flowchart of a two-way communication method between a ship physical test bench and an integrated platform based on a test bench agent, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the overall functional architecture of the agent program according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the system units in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0021] Example 1 Please refer to Figure 1 This embodiment 1 provides a two-way communication method between a ship's physical test bench and an integrated platform based on a test bench agent, including: S1. Using the bench agent as a transit hub, a bidirectional communication channel for heterogeneous protocol adaptation is established between the physical bench and the integration platform. The integration platform sends standardized instructions carrying verification information to the bench agent. S2. The bench agent encodes the unique identifier of the bench as a set of projective geometric points in a finite field. It generates instruction-specific projective coordinates through a projective transformation matrix. It utilizes the non-intersecting property of the projective space to achieve collision-free addressing of parallel instructions, automatically distinguishes the target bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from a mathematical structure perspective. S3. The benchtop agent extracts the topology-invariant features of the data stream of the original serial port data, constructs a data frame topology adjacency graph, completes the optimal rearrangement of out-of-order frames through topology isomorphism determination, and performs structural completion of missing frames based on topology space completeness constraints, thereby achieving data timing recovery under the conditions of no external clock and no threshold weight. S4. After completing protocol conversion, multi-layer verification, and manual confirmation, the bench agent forwards the instruction to the physical bench for execution and uniformly encapsulates and synchronously uploads the processed return data to the integration platform, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

[0022] Please refer to Figure 2 This embodiment 1 further elaborates on the above steps.

[0023] (1) Construction of bidirectional communication channel To address the challenges of protocol heterogeneity, unstable connections, and difficulties in unified addressing between shipboard physical test benches and the integration platform, this embodiment uses a test bench agent as the core relay node to construct a bidirectional communication channel adaptable to multiple scenarios. The test bench agent is deployed in the host computer environment of the physical test bench, serving as a unified entry point for data interaction between the two ends. It first establishes a communication link with the physical test bench that is compatible with multiple serial communication protocols to meet the access requirements of different test bench models. Simultaneously, it establishes a full-duplex persistent connection with the integration platform to ensure real-time bidirectional transmission of commands and data.

[0024] During the connection establishment phase, the benchtop agent registers its identity and subscribes to communication topics with the system message middleware, achieving message routing alignment with the platform. To ensure the uniqueness and stability of subsequent addressing encoding, the agent automatically generates its own unique identifier during connection initialization. This identifier directly serves as the basis for subsequent finite field projective geometric encoding, requiring no additional configuration.

[0025] During channel operation, the rack agent continuously monitors the link status between the physical rack side and the integration platform side. Once a connection interruption, timeout, or abnormal disconnection is detected, the autonomous reconstruction mechanism is immediately activated to quickly restore the communication link. Through the above steps of deployment, connection establishment, identity registration, link monitoring, and autonomous recovery, a stable and reliable bidirectional communication channel that can support subsequent parallel addressing and timing data processing is ultimately formed, providing a fundamental guarantee for the overall communication process.

[0026] (2) Conflict-free parallel distribution In scenarios where multiple shipboard physical test benches are simultaneously connected to an integrated platform, parallel instruction addressing is prone to address overlap, signal interference, and target recognition errors, directly affecting the stability of collaborative equipment operation. To address this issue from a mathematical and spatial structure perspective, this embodiment relies on projective geometry and finite field operations to achieve conflict-free addressing. The system uses a preset prime power finite field as the unified number field basis for all projective operations. The test bench agent first determines the feature vector dimension based on the maximum number of test benches supported by the system, and then generates the corresponding feature vector by binary digitizing the unique identifier of the test bench. That is, the finite field adopts... , The prime power serves as the field foundation for the entire projective operation; the gantry agent determines the dimension of the eigenvector based on the maximum number of gantry units supported by the system. The unique identifier of the test bench is formed by binary digit encoding. 3D feature vector, denoted as , ; Furthermore, the rack agent in above generated The full-rank projective transformation matrix of order 1 is denoted as . Projective transformation matrix It is generated by constructing random orthogonal matrices over a finite field. The construction process is as follows: Random selection The linearly independent column vectors, after being orthogonalized by Gram-Schmidt, are combined into a matrix to ensure that the projective transformation matrix... Satisfy determinant Projective transformation matrix Generated and globally fixed during system initialization, all benchtop agents share the same projective transformation matrix. ; Projective coordinates are derived from eigenvectors With projective transformation matrix The product is directly generated, and the calculation formula is: ,in For the initially generated projective coordinates, ; benchtop agent Perform Gram-Schmidt orthogonalization within a finite field; the calculation formula is as follows: ,in For the final projective coordinates of the generated history, The inner product operation is defined within a finite field. To obtain the corresponding norm, the final addressing coordinates that satisfy the condition of pairwise linear independence are finally obtained. , ; Due to the projective transformation matrix For a full-rank matrix, this transformation is performed in The above satisfies injectivity, and different eigenvectors They must correspond to different initial projective coordinates After orthogonalization and normalization, all final projective coordinates In a finite field projective space, there is no overlap between any two nodes, which completely eliminates addressing conflicts and cross-interference of parallel instructions from a mathematical structure perspective, ensuring the distribution of parallel instructions across multiple racks.

[0027] During the addressing phase, the benchtop agent, in the system initialization phase, bases its work on the determined finite field. With the dimension of the feature vector ,Will The finite field projective space is decomposed into orthogonal subspaces according to the number of racks, dividing the entire projective space into a set of orthogonal subspaces that correspond one-to-one with the racks, with each subspace corresponding to the instruction addressing domain of a rack. The partitioning of subspaces satisfies orthogonality and mutual exclusivity, meaning that any two different pedestals correspond to the same subspace. and They satisfy an orthogonal relation And the intersection is empty. This ensures that there are no overlapping addressing regions between subspaces; The rack agent will provide the final addressing coordinates for each rack. Mapped to a dedicated subspace, each bench's instructions are transmitted and recognized only within its assigned subspace. When the integration platform issues parallel instructions, the bench agent extracts the final addressing coordinates carried by the instructions. The target subspace is determined by spatial location, and then the command is forwarded to the physical platform corresponding to that subspace. Since the subspaces are orthogonal and do not overlap, instructions from different racks have independent transmission channels in the projective space, avoiding cross-domain interference and addressing conflicts, and ensuring the stable distribution and execution of parallel instructions from multiple racks from the spatial structure level.

[0028] (3) Topology timing recovery and completion During serial data transmission on a ship's physical test bench, issues such as out-of-order data frames, frame loss, and timing errors can occur due to transmission distance, electromagnetic interference, and link jitter. Traditional recovery methods relying on external clocks or manually set thresholds lack stability and cannot guarantee the continuity of data logic. To address this problem, the test bench agent, based on topology analysis and finite field operation rules, achieves autonomous data timing recovery without relying on external clocks or weighted thresholds.

[0029] First, the benchtop agent performs frame splitting and format verification on the raw serial port data, mapping the frame header identifier, data segment length, verification field, and instruction association number of each frame to a preset finite field. Above, among which It is a prime power, serving as the unified number field foundation for topological operations; structure 3D topological eigenvectors , ,in For data frame sequence number, For a fixed feature dimension, For the first Frame number One feature in The quantization value within; Using each topological feature vector To form a node set for nodes Then, based on the cyclic shift invariance over a finite field, we construct the adjacency relationship: define the cyclic shift matrix. ,in for An identity cyclic matrix of order 1 satisfies the transformation property of cyclically shifting vector elements one bit to the right. For any two nodes , If there exists a unique positive integer satisfy ,in Circular shift matrix of The exponent represents a continuous sequence. The next cyclic shift transformation determines whether two frames are topologically adjacent and establishes an undirected edge. , forming an edge set ,in, This represents the number of cyclic shift steps, corresponding to the logical interval between data stream frames. Finally, the data frame topology adjacency graph is obtained. This graph maintains its structure under cyclic shift transformation and can uniquely determine the inter-frame logical order without the need for thresholds, weights, or external clocks.

[0030] In the out-of-order processing stage, the benchtop agent uses the pre-built data frame topology adjacency graph. ,in For node set, It is a set of edges, where each node corresponds to one edge. 3D topological eigenvectors ,and , For a predefined finite field, It is a prime power; First, extract the topological invariants of the graph, including the node degree sequence. Adjacency matrix and characteristic spectrum ;Node degree sequence The adjacency matrix is ​​composed of the number of adjacent edges of each node. Defined in Above, the element values ​​represent the adjacency relationships between nodes, and the feature spectrum... Adjacency matrix The set of eigenvalues, the above invariant set, uniquely characterizes the structural features of the topological graph and is not affected by the order of node arrangement; Benchtop agent pre-stores the baseline topology adjacency graph corresponding to the standard data stream , and its benchmark invariant set Topology graph constructed from out-of-order data frames Iterate through all possible permutations and substitutions of nodes. Construct the permutation matrix And calculate the adjacency matrix after the permutation. ,in For corresponding permutations The permutation matrix, dimension and Consistent; If a unique permutation exists , so that the adjacency matrix after the permutation Adjacency matrix with reference Completely identical, and characteristic spectrum Compared with the benchmark characteristic spectrum If a match is found, the topology graph is determined. Compared with the baseline topology diagram Topological isomorphism, this permutation This refers to the optimal rearrangement order of out-of-order frames; the gantry proxy is based on the permutation... The out-of-order data frames are rearranged to restore the original out-of-order frame sequence to a standard sequence consistent with the reference data stream, thus completing the optimal rearrangement of out-of-order frames.

[0031] For missing data frames, the benchtop agent uses the rearranged data frame topology adjacency graph as its structural basis, within a finite domain. Under constraints, perform a completeness check on the topological space by traversing the topological feature vectors corresponding to all nodes. The system identifies structural breakpoints in the adjacency matrix that contain empty connections, discontinuous degree sequences, and unclosed feature spectra, and marks the location of these breakpoints as the location of missing data frames. For each marked structural breakpoint, the benchtop agent extracts the feature vectors of the existing adjacent nodes on both sides of the breakpoint. and Based on the rules of linear space operations in finite fields, the topological feature vector of the missing frame is calculated. The operation process follows The closure property of addition guarantees a unique and stable result. Based on the calculated topological feature vector The benchtop agent reverse maps and generates the corresponding data frame structure, including frame header identifier, data length, checksum field and business association number, fully restores all valid information of the missing frame, and inserts the completed node into the corresponding position in the topology adjacency relationship graph. After completion, the benchtop agent updates the adjacency matrix and feature spectrum to ensure that the overall topology meets the completeness condition. This enables the complete structure recovery of the data stream under conditions without external clock or weight threshold, effectively solving the problem of incomplete data caused by frame loss and frame breakage in serial transmission, and ensuring that the uplink data timing is continuous and the logic is reliable.

[0032] (4) End-to-end communication Shipboard physical test benches and integrated platforms often employ different communication protocols and data formats for data transmission. Different models of physical test benches, due to variations in manufacturers and functional requirements, exhibit differences in non-standard data formats transmitted via serial ports. However, the integrated platform, for unified management, needs to receive standardized and regulated data. This protocol heterogeneity directly leads to problems such as field mismatches, parsing failures, and data misinterpretation during data interaction, severely impacting the accuracy of command execution and data feedback. This becomes a critical bottleneck in establishing a smooth communication link between the physical test benches and the integrated platform. To effectively address this issue and break down the barriers to heterogeneous data transmission, the test bench agent, after completing data timing recovery and structural completion to ensure data flow integrity and logical continuity, specifically performs standardized protocol conversion processing to achieve normalized adaptation of heterogeneous data.

[0033] The benchtop agent first performs protocol field parsing on the complete data stream after timing recovery and structural completion, and rearranges and aligns the non-standard data format transmitted by the serial port physical layer according to the unified communication specifications of the integration platform. benchtop agent based on finite domain The data payload is uniformly quantized and mapped, and the multi-type protocol fields of different physical platforms are converted into standardized data units of the same dimension, so that all kinds of heterogeneous data have a consistent transmission format and verification basis. During the conversion process, the benchtop agent retains the original topology feature identifier and business association information of the data frame without changing the meaning of the data itself, and only completes the protocol format normalization process; The data stream after protocol conversion fully meets the receiving standards of the integration platform. It can be directly encapsulated by the benchtop agent without additional adjustments, which fully prepares for subsequent multi-layer verification, manual confirmation relay, and finally synchronous upload of the returned data to the integration platform. This effectively breaks down the communication barriers of heterogeneous protocols and ensures the smooth and seamless communication process.

[0034] Example 2 Please refer to Figure 3 This embodiment 2 provides a two-way communication system between a ship's physical test bench and an integrated platform based on a test bench agent, including: The bidirectional communication channel construction unit is used to establish a bidirectional communication channel between the physical test bench and the integration platform with the test bench agent as the relay hub. The integration platform sends standardized instructions carrying verification information to the test bench agent. The conflict-free parallel distribution unit is used by the test bench agent to encode the unique identifier of the test bench into a set of finite field projective geometric points, generate instruction-specific projective coordinates through the projective transformation matrix, and realize parallel instruction collision-free addressing by utilizing the non-intersecting property of the projective space. It automatically distinguishes the target test bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from the mathematical structure. The topology timing recovery and completion unit is used to extract the topology invariant features of the data stream of the original serial port data by the benchtop agent, construct the topology adjacency relationship graph of the data frame, complete the optimal rearrangement of out-of-order frames through topology isomorphism determination, and perform structural completion of missing frames based on topology space completeness constraints, so as to realize data timing recovery under the conditions of no external clock and no threshold weight. The end-to-end communication unit is used to forward instructions to the physical benchtop after the benchtop agent completes protocol conversion, multi-layer verification, and manual confirmation. It also encapsulates the processed return data and uploads it to the integration platform in a unified and time-synchronized manner, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

[0035] Example 3 This embodiment 3 also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement any step of a two-way communication method between a ship physical bench and an integrated platform based on a bench agent.

[0036] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0037] For a description of the computer-readable storage medium provided in this application, please refer to the above method embodiments; further details will not be repeated here.

[0038] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A two-way communication method between a ship's physical test bench and an integrated platform based on a test bench agent, characterized in that, include: S1. Using the bench agent as a transit hub, a bidirectional communication channel for heterogeneous protocol adaptation is established between the physical bench and the integration platform. The integration platform sends standardized instructions carrying verification information to the bench agent. S2. The bench agent encodes the unique identifier of the bench as a set of projective geometric points in a finite field. It generates instruction-specific projective coordinates through a projective transformation matrix. It utilizes the non-intersecting property of the projective space to achieve collision-free addressing of parallel instructions, automatically distinguishes the target bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from a mathematical structure perspective. S3. The benchtop agent extracts the topology-invariant features of the data stream of the original serial port data, constructs a data frame topology adjacency graph, completes the optimal rearrangement of out-of-order frames through topology isomorphism determination, and performs structural completion of missing frames based on topology space completeness constraints, thereby achieving data timing recovery under the conditions of no external clock and no threshold weight. S4. After completing protocol conversion, multi-layer verification, and manual confirmation, the bench agent forwards the instruction to the physical bench for execution and uniformly encapsulates and synchronously uploads the processed return data to the integration platform, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

2. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent, as described in claim 1, is characterized in that... The process of constructing the bidirectional communication channel in S1 is as follows: The rack agent is deployed on the host computer of the physical rack. The rack agent establishes a communication link with the physical rack that is compatible with multiple serial port protocols. The rack agent establishes a full-duplex persistent connection with the integration platform. The rack agent completes identity registration and topic subscription with the message middleware. The bench agent generates a unique identifier during the connection establishment process. This unique identifier is directly used for finite field projective geometry encoding. The bench agent monitors the connection status in real time and initiates autonomous reconstruction when interrupted, thereby building a bidirectional communication channel that can support subsequent addressing and data processing.

3. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent, as described in claim 1, is characterized in that... The process of generating instruction-specific projective coordinates through the projective transformation matrix in S2 is as follows: Using a preset finite field of prime powers as the basis for the overall projective operation, the bench agent determines the feature vector dimension based on the maximum number of benches that the system can support, and generates the feature vector of the corresponding dimension by binary digit encoding the unique identifier of the bench. The benchtop agent constructs a full-rank projective transformation matrix of the same dimension within the finite field. The construction method is to select linearly independent column vectors in the finite field, and combine them after orthogonalization to form a matrix, ensuring that the determinant of the matrix is ​​non-zero. This projective transformation matrix is ​​generated during the system initialization phase and used globally. The initial projective coordinates are obtained by multiplying the eigenvectors by the projective transformation matrix. Then, the initial projective coordinates are orthogonalized and normalized in a finite field. Combined with the generated historical addressing coordinates, iterative calculations are completed to finally obtain the final addressing coordinates that meet the linear independence requirement.

4. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent, as described in claim 3, is characterized in that... The process of achieving collision-free addressing of parallel instructions using the non-overlapping property of projective spaces in S2 is as follows: During the system initialization phase, the benchtop agent, based on the defined finite field... With the dimension of the feature vector ,Will The finite field projective space is decomposed into orthogonal subspaces according to the number of racks, dividing the entire projective space into a set of orthogonal subspaces that correspond one-to-one with the racks, with each subspace corresponding to the instruction addressing domain of a rack. The partitioning of subspaces satisfies orthogonality and mutual exclusivity, meaning that any two different pedestals correspond to the same subspace. and They satisfy an orthogonal relation And the intersection is empty. This ensures that there are no overlapping addressing regions between subspaces; The rack agent will provide the final addressing coordinates for each rack. Mapped to a dedicated subspace, so that the instructions of each rack are transmitted and recognized only within its assigned subspace; When the integration platform issues parallel instructions, the benchtop agent extracts the final addressing coordinates carried in the instructions. The target subspace is determined by spatial location, and then the command is forwarded to the physical platform corresponding to that subspace. Since the subspaces are orthogonal and do not overlap, instructions from different racks have independent transmission channels in the projective space, avoiding cross-domain interference and addressing conflicts, and ensuring the stable distribution and execution of parallel instructions from multiple racks from the spatial structure level.

5. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent, as described in claim 1, is characterized in that... The process of constructing the data frame topology adjacency graph in S3 is as follows: After the benchtop agent performs frame splitting and format verification on the raw serial port data, it maps the various core features of each frame of data to a finite field of a preset prime power. Based on this, it constructs a fixed-dimensional topological feature vector corresponding to each data frame. The benchtop agent uses each topological feature vector as a node to form a node set. Based on the cyclic shift invariance of finite fields, a fixed-order unit cyclic matrix with vector cyclic right shift function is defined. The topological adjacency relationship between nodes is constructed through this matrix and its power transformation. For any two nodes, if there exists a unique positive integer as the number of cyclic shift steps, such that one vector is completely identical to the other vector after the corresponding number of cyclic shift transformations, then it is determined that the data frames corresponding to the two vectors have a topological adjacency relationship and an undirected edge is established, thus forming an edge set. The final data frame topology adjacency graph is composed of the node set and the edge set.

6. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent according to claim 1, characterized in that, The process of achieving optimal rearrangement of out-of-order frames through topological isomorphism determination in S3 is as follows: The benchtop agent is based on the constructed data frame topology adjacency graph. ,in For node set, It is a set of edges, where each node corresponds to one edge. 3D topological eigenvectors ,and , For a predefined finite field, It is a prime power; First, extract the topological invariants of the graph, including the node degree sequence. Adjacency matrix and characteristic spectrum ;Node degree sequence The adjacency matrix is ​​composed of the number of adjacent edges of each node. Defined in Above, the element values ​​represent the adjacency relationships between nodes, and the feature spectrum... Adjacency matrix The set of eigenvalues, the above invariant set, together uniquely characterizes the structural features of the topological graph and is not affected by the order of node arrangement; Benchtop agent pre-stores the baseline topology adjacency graph corresponding to the standard data stream , and its benchmark invariant set Topology graph constructed from out-of-order data frames Iterate through all possible permutations and substitutions of nodes. Construct the permutation matrix And calculate the adjacency matrix after the permutation. ,in For corresponding permutations The permutation matrix, dimension and Consistent; If a unique permutation exists , so that the adjacency matrix after the permutation Adjacency matrix with reference Completely identical, and characteristic spectrum Compared with the benchmark characteristic spectrum If a match is found, the topology graph is determined. Compared with the baseline topology diagram Topological isomorphism, this permutation This refers to the optimal rearrangement order of out-of-order frames; the gantry proxy is based on the permutation... The out-of-order data frames are rearranged to restore the original out-of-order frame sequence to a standard sequence consistent with the reference data stream, thus completing the optimal rearrangement of out-of-order frames.

7. The bidirectional communication method between a ship physical test bench and an integrated platform based on a test bench agent as described in claim 1, characterized in that, The process of structural completion of missing frames based on topological space completeness constraints in S3 is as follows: The benchtop agent uses the rearranged data frame topology adjacency graph as its structural basis, within a finite domain. Under constraints, perform a completeness check on the topological space by traversing the topological feature vectors corresponding to all nodes. The system identifies structural breakpoints in the adjacency matrix that contain empty connections, discontinuous degree sequences, and unclosed feature spectra, and marks the location of these breakpoints as the location of missing data frames. For each marked structural breakpoint, the benchtop agent extracts the feature vectors of the existing adjacent nodes on both sides of the breakpoint. and Based on the rules of linear space operations in finite fields, the topological feature vector of the missing frame is calculated. The operation process follows The closure property of addition guarantees a unique and stable result. Based on the calculated topological feature vector The benchtop agent reverse maps and generates the corresponding data frame structure, including frame header identifier, data length, checksum field and business association number, fully restores all valid information of the missing frame, and inserts the completed node into the corresponding position in the topology adjacency relationship graph. After completion, the benchtop agent updates the adjacency matrix and feature spectrum to ensure that the overall topology meets the completeness condition. This enables the complete structure recovery of the data stream under conditions without external clock or weight threshold, effectively solving the problem of incomplete data caused by frame loss and frame breakage in serial transmission, and ensuring that the uplink data timing is continuous and the logic is reliable.

8. A two-way communication method between a ship's physical test bench and an integrated platform based on a test bench agent, as described in claim 1, is characterized in that... The process by which the S3 platform agent completes the protocol conversion is as follows: The benchtop agent parses the protocol fields of the complete data stream after timing recovery and structural completion, and rearranges and aligns the non-standard data format transmitted by the serial port physical layer according to the unified communication specifications of the integration platform. benchtop agent based on finite domain The data payload is uniformly quantized and mapped, and the multi-type protocol fields of different physical platforms are converted into standardized data units of the same dimension, so that all kinds of heterogeneous data have a consistent transmission format and verification basis. During the conversion process, the benchtop agent retains the original topology feature identifier and business association information of the data frame without changing the meaning of the data itself. It only completes the protocol format normalization process. The converted data meets the reception standards of the integration platform and can be directly used for encapsulation and uploading.

9. A two-way communication system between a ship's physical test bench and an integrated platform based on a test bench agent, characterized in that, include: The bidirectional communication channel construction unit is used to establish a bidirectional communication channel between the physical test bench and the integration platform with the test bench agent as the relay hub. The integration platform sends standardized instructions carrying verification information to the test bench agent. The conflict-free parallel distribution unit is used by the test bench agent to encode the unique identifier of the test bench into a set of finite field projective geometric points, generate instruction-specific projective coordinates through the projective transformation matrix, and realize parallel instruction collision-free addressing by utilizing the non-intersecting property of the projective space. It automatically distinguishes the target test bench and completes instruction distribution, thus eliminating parallel conflicts and addressing errors from the mathematical structure. The topology timing recovery and completion unit is used to extract the topology invariant features of the data stream of the original serial port data by the benchtop agent, construct the topology adjacency relationship graph of the data frame, complete the optimal rearrangement of out-of-order frames through topology isomorphism determination, and perform structural completion of missing frames based on topology space completeness constraints, so as to realize data timing recovery under the conditions of no external clock and no threshold weight. The end-to-end communication unit is used to forward instructions to the physical benchtop after the benchtop agent completes protocol conversion, multi-layer verification, and manual confirmation. It also encapsulates the processed return data and uploads it to the integration platform in a unified and time-synchronized manner, forming a closed-loop communication process of instruction issuance, verification, confirmation, execution, and feedback.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor as described in any one of claims 1-8: a two-way communication method for a ship physical bench-integrated platform based on bench agent.

Images