KKS code-based power plant raw coal bunker equipment whole life cycle tracing system

Abstract

The application discloses a power plant raw coal bunker equipment full life cycle tracing system based on KKS coding, comprising a physical sensing layer, an edge computing layer and a logic evolution and control layer, the physical sensing layer applies an alternating excitation magnetic field to a device entity, obtains a magnetic Barkhausen noise signal and converts the signal into a digital signal, the edge computing layer performs filtering on the digital signal and extracts an initial physical feature vector, performs a hash operation to generate an intrinsic entity identifier, the logic evolution and control layer constructs a directed acyclic graph structure, calculates a characteristic expected drift compensation tensor by using a space environment stress spectrum, completes device addressing matching by means of Mahalanobis distance, and converts a characteristic drift residual into a material characteristic degradation index to dynamically update node connection weights. The application realizes the binding of physical entities and digital identifiers, guarantees the integrity of device cross-node circulation records, and effectively drives predictive maintenance and early warning.

Description

Technical Field

[0001] This invention relates to the field of power plant equipment asset management and traceability technology, specifically to a full lifecycle traceability system for power plant raw coal bunker equipment based on KKS coding. Background Technology

[0002] With the continuous expansion of thermal power plant scale, the refined management of equipment assets has become a key link in ensuring the safe and stable operation of the units. As the core hub of the pulverized coal preparation system, the reliability of the raw coal bunker and its related equipment directly affects the boiler combustion efficiency and the unit's operating status. In order to achieve scientific operation and maintenance of raw coal bunker equipment, the industry has generally introduced a full life cycle traceability system, which aims to record data and track the status of equipment throughout all stages from warehousing, installation, service, to decommissioning and dismantling.

[0003] In existing power plant equipment traceability systems, the KKS coding standard is typically used to construct a ledger of equipment's spatial location and functional logic. In the specific physical mapping stage, managers primarily rely on attaching external physical tags, such as QR code nameplates, barcodes, or RFID electronic tags, to the equipment surface to establish a link between the physical equipment and the ledger data in the information system. During routine inspections or equipment transfers, on-site personnel use handheld terminals to scan and identify these external physical tags to obtain equipment identification and update the equipment's location and maintenance status in the information system, thereby maintaining the flow of traceability data.

[0004] However, existing traceability methods have the following drawbacks in practical applications: The raw coal storage area is exposed to harsh industrial conditions characterized by high dust, strong vibration, and corrosiveness. Traditional external physical tags attached to equipment surfaces are prone to wear, detachment, or damage from thick coal dust, preventing on-site terminals from properly reading tag information and causing a break in the link between the physical equipment and its digital identifier. More critically, traditional traceability systems often use static relational data tables to record equipment status. When raw coal storage equipment undergoes complex cross-domain transfers such as dismantling or relocation between different functional tag numbers, the system can only record simple changes in current location, making it difficult to trace the complete transfer topology of the equipment across multiple lifecycle stages. This vulnerability in physical identification and the logical gap in record-keeping ultimately prevent the full lifecycle traceability system from providing accurate data for equipment transfer tracking and scientific operation and maintenance. Summary of the Invention

[0005] The technical problem to be solved by this invention is that in the existing power plant raw coal bunker equipment traceability system, the external physical tags attached to the equipment surface are easily worn and fail under harsh working conditions, resulting in the interruption of the reading chain. Furthermore, the traditional static relational data table is difficult to trace the complete logical topology of complex flow processes such as equipment cross-tag dismantling and transfer. As a result, the full life cycle traceability system cannot provide an accurate basis for the scientific operation and maintenance of the equipment.

[0006] This invention provides a full lifecycle traceability system for power plant raw coal bunker equipment based on KKS coding, comprising: a physical sensing layer, an edge computing layer, and a logical evolution and control layer.

[0007] The physical sensing layer is deployed at the raw coal bunker equipment site to perform excitation and signal acquisition of the microscopic physical characteristics of the material. The physical sensing layer includes a portable magnetic field excitation and detection device and the raw coal bunker equipment entity. During periodic inspections or equipment transfer verification, the portable magnetic field excitation and detection device applies an alternating excitation magnetic field to a preset reference surface of the raw coal bunker equipment entity to obtain the original voltage signal of magnetic Barkhausen noise generated by the movement of magnetic domain walls inside the raw coal bunker equipment entity, and converts the original voltage signal of magnetic Barkhausen noise into a digital signal.

[0008] The edge computing layer establishes a communication connection with the physical sensing layer to perform signal processing and entity identification calculation on the acquired digital signals. The edge computing layer includes a signal preprocessing module, a feature vector construction module, and a hash calculation module. The signal preprocessing module receives digital signals and performs bandpass filtering. The feature vector construction module extracts time-domain and frequency-domain feature parameters based on the filtered signal to generate an initial physical feature vector. The hash calculation module performs a hash operation on the initial physical feature vector using a one-way hash algorithm to generate the corresponding intrinsic entity identifier.

[0009] The logical evolution and control layer establishes a data connection with the edge computing layer to construct the device flow topology and perform node state updates. The logical evolution and control layer includes a topology network construction module, a stress environment configuration module, an adaptive addressing matching module, and a state evolution control module.

[0010] The topology network construction module constructs a directed acyclic graph structure. The global state topology network model is defined as consisting of a set of nodes and a set of directed edges. The set of directed edges is used to record the temporal and spatial binding relationships of nodes. The topology network construction module extracts the spatial location identifiers of the raw coal bunker process flow according to the power plant KKS coding rules and uses them as node identifiers to divide the node set into a subset of static functional nodes representing spatial locations and a subset of entity nodes representing raw coal bunker equipment entities. The union of the static functional node subset and the entity node subset is the node set, and the intersection of the static functional node subset and the entity node subset is an empty set.

[0011] The stress environment configuration module configures a spatial environment stress spectrum matrix for each node in the static functional node subset. The spatial environment stress spectrum matrix includes the vibration stress amplitude, dominant frequency vibration characteristics, and environmental temperature parameters of the corresponding node's spatial location.

[0012] The adaptive addressing and matching module receives the current physical feature vector uploaded by the edge computing layer. The adaptive addressing and matching module retrieves the spatial environment stress spectrum matrix and entity service time parameters preset by the corresponding node, and calculates the feature expected drift compensation tensor in combination with the material degradation model. The adaptive addressing and matching module updates the initial physical feature vector using the feature expected drift compensation tensor and generates the expected comparison feature vector. The adaptive addressing and matching module calculates the Mahalanobis distance between the current physical feature vector and the expected comparison feature vector, and determines whether the Mahalanobis distance is less than the preset addressing and matching threshold. When the Mahalanobis distance is less than the addressing and matching threshold, the adaptive addressing and matching module determines that the addressing is successful, and establishes a mapping relationship between the raw coal bunker equipment entity and the corresponding node in the static functional node subset in the directed acyclic graph structure constructed by the topology network construction module.

[0013] The state evolution control module calculates the feature drift residual vector between the current physical feature vector and the initial physical feature vector based on the matching result of the adaptive addressing matching module. The state evolution control module transforms the feature drift residual vector into a material feature degradation index through a multidimensional fatigue mapping function, and updates the weight values ​​of the corresponding edges in the directed edge set E using the material feature degradation index. When the weight value of the corresponding edge reaches a preset critical threshold, the state evolution control module triggers and outputs a device maintenance early warning command. When a device dismantling command is received and the current topology directed edge connection is released, the topology network construction module solidifies the released topology directed edge into a static historical snapshot edge, and retains the flow record of the entity node in the global state topology network model to form a historical edge subgraph set, realizing the full life cycle traceability of the device's cross-node flow.

[0014] Preferably, the topology network construction module extracts the equipment hierarchical association relationship of the power plant's raw coal bunker process flow and maps the equipment hierarchical association relationship to the global state topology network model; the topology network construction module initializes each spatial location index in the static functional node subset according to the power plant facility configuration ledger and the KKS coding rules, each spatial location index corresponds to an independent running process node, and the attributes of the running process node include spatial coordinate parameters and system classification parameters to which the process belongs; after the raw coal bunker equipment entity completes the registration of intrinsic entity identifier, the topology network construction module generates equipment object nodes that map the raw coal bunker equipment entity in the entity node subset.

[0015] Preferably, the stress environment configuration module extracts historical operating condition data for the corresponding spatial physical location according to the spatial physical location index of each element in the static functional node subset; the stress environment configuration module quantifies the extracted historical operating condition data, calculates and obtains the average vibration stress amplitude, dominant frequency vibration characteristic parameter, and extreme environmental temperature parameter for the corresponding spatial physical location, and retrieves the rated mechanical load parameter, coal powder friction coefficient measure, and environmental corrosion index factor for the corresponding spatial physical location from the preset physical property mapping database through a lookup matching command based on the rated operating parameters and coal quality characteristics of the raw coal bunker equipment; the stress environment configuration module constructs the quantified parameters into a two-dimensional feature matrix according to a preset dimension to generate a spatial environmental stress spectrum matrix for the corresponding spatial physical location.

[0016] Preferably, the signal preprocessing module within the edge computing layer performs envelope detection processing on the filtered signal sequence, extracts the outer contour features of the filtered signal sequence, and generates an envelope signal sequence; the feature vector construction module calculates the root mean square value of the filtered signal sequence within one magnetization period as the root mean square value parameter, retrieves the maximum amplitude point in the envelope signal sequence as the Barkhausen jump peak parameter, and performs discrete accumulation integration on the envelope signal sequence in the time domain to obtain the envelope integral area parameter; the feature vector construction module performs discrete Fourier transform calculation on the filtered signal sequence to obtain the power spectral density peak energy parameter, and concatenates and combines the extracted root mean square value parameter, Barkhausen jump peak parameter, envelope integral area parameter, and power spectral density peak energy parameter to construct an initial physical feature vector.

[0017] Preferably, the hash calculation module extracts each feature parameter from the initial physical feature vector and performs fixed-point numerical truncation processing on each feature parameter to generate a fixed-point numerical sequence; the hash calculation module converts the fixed-point numerical sequence into a binary data stream, retrieves the secure hash algorithm model from the system memory, and uses the secure hash algorithm model to perform hash mapping calculation on the binary data stream to generate intrinsic entity identifiers; the topology network construction module allocates node storage units in the entity node subset of the global state topology network model, writes the intrinsic entity identifier as the index primary key into the node storage unit to generate entity nodes, and writes the initial physical feature vector and the device entry time parameter as node attribute data.

[0018] Preferably, the topology network construction module receives the equipment installation confirmation instruction, extracts the intrinsic entity identifier and target installation location index carried in the equipment installation confirmation instruction, and retrieves and determines the matching source entity node and the matching target functional node respectively; the topology network construction module generates a topology directed edge with the source entity node as the starting point and the target functional node as the ending point, and writes the current time parameter as the installation timestamp attribute into the data structure of the topology directed edge; the topology network construction module extracts the historical connection edge record parameters recorded in the attribute data of the source entity node to determine the service attribute, obtains the residual fatigue weight parameter, and assigns the obtained residual fatigue weight parameter to the weight attribute field of the topology directed edge, completing the initial binding of action connection edge and state attribute.

[0019] Preferably, the adaptive addressing matching module calculates the expected drift compensation tensor of features in conjunction with the material degradation model. The expected drift compensation tensor of features is obtained by linearly superimposing the product of the drift sensitivity coefficient matrix and the space environment stress spectrum matrix, plus the product of the service time parameter and the time drift compensation coefficient. The adaptive addressing matching module performs vector addition on the initial physical feature vector and the expected drift compensation tensor of features to obtain the expected comparison feature vector. The adaptive addressing matching module extracts the drift tolerance covariance matrix contained in the attributes of the matched static functional nodes. The drift tolerance covariance matrix is ​​statistically generated by the system based on the historical discrete drift sample set of the corresponding spatial physical location. The adaptive addressing matching module calculates the Mahalanobis distance based on the current physical feature vector, the expected comparison feature vector, and the drift tolerance covariance matrix. The Mahalanobis distance is obtained by calculating the square root of the product of the transpose of the difference vector between the current physical feature vector and the expected comparison feature vector, the inverse matrix of the drift tolerance covariance matrix, and the difference vector.

[0020] Preferably, the state evolution control module calculates a feature drift residual vector by subtracting the initial physical feature vector from the current physical feature vector; the state evolution control module uses the feature drift residual vector as an input parameter to import it into a multidimensional fatigue mapping function, and calculates the L2 norm squared value and the multidimensional gradient feature parameter of the feature drift residual vector respectively; the state evolution control module multiplies the L2 norm squared value by a preset first conversion coefficient constant to obtain a first weighted parameter, multiplies the multidimensional gradient feature parameter by a preset second conversion coefficient constant to obtain a second weighted parameter, and adds the first weighted parameter and the second weighted parameter to obtain the material feature degradation index.

[0021] Preferably, the state evolution control module retrieves and determines the topological directed edge connecting the current entity node and the matching static functional node from the set of directed edges in the global state topology network model, and extracts the state evolution weight parameter recorded at the initial binding time of the topological directed edge; the state evolution control module performs numerical accumulation operation on the state evolution weight parameter and the calculated material characteristic degradation index to obtain the dynamically updated weight value; the state evolution control module writes the dynamically updated weight value into the weight attribute field of the topological directed edge, and updates the current system time parameter into the observation timestamp attribute of the topological directed edge.

[0022] Preferably, the state evolution control module compares the dynamically updated weight value with a preset failure critical weight threshold. When the dynamically updated weight value is greater than or equal to the failure critical weight threshold, the maintenance warning state parameter is set to one; when the dynamically updated weight value is less than the failure critical weight threshold, the maintenance warning state parameter is set to zero. The topology network construction module receives the device disassembly confirmation command, obtains the disassembly timestamp attribute, and updates the Boolean value of the connection between the corresponding source entity node and the target functional node to zero in the adjacency matrix of the directed acyclic graph data structure to remove the topology directed edge connection. The edges are solidified as static historical snapshot edges; the topology network construction module reads the final dynamic update weight value corresponding to the disassembly timestamp attribute, sets the final dynamic update weight value as the current residual fatigue weight parameter of the source entity node, and writes the current residual fatigue weight parameter as a historical connection edge record parameter into the node attribute data structure of the source entity node, restores the connection state parameter of the source entity node to the initial independent state, so that the static historical snapshot edges together with the flow records of each time of the source entity node are stored in the global state topology network model, forming a set of historical connection edge subgraphs starting from the source entity node in terms of data structure.

[0023] The present invention, by adopting the above technical solution, can bring the following beneficial effects:

[0024] 1. This invention uses magnetic Barkhausen noise acquisition technology in the physical sensing layer, combined with the extraction of material micro-feature parameters and feature identification registration in the edge computing layer, to achieve implicit binding between the physical entity of the equipment and the intrinsic digital identifier. This solves the problems of easy failure, difficulty in reading, and easy information chain breakage of external physical tags in harsh working conditions such as high dust and strong corrosion in the raw coal bunker area, and ensures the reliability of physical anchoring for equipment life cycle traceability in complex industrial environments.

[0025] 2. This invention constructs a dynamic topology based on a directed acyclic graph through logical evolution and control layers, and introduces spatial environmental stress spectrum and material degradation model into feature drift compensation calculation. This enables logical tracing of complex behaviors such as equipment installation, dismantling, and relocation. It solves the technical limitation of traditional relational databases in not being able to retain a complete life assessment genealogy when processing equipment cross-placement transfers, and effectively improves the integrity of full life cycle status data and the scientific management level of equipment operation and maintenance.

[0026] 3. This invention transforms the microscopic magnetic feature drift residual into a material feature degradation index and dynamically updates it to the topological directed edge weights, thereby achieving synchronous recording of the spatiotemporal correlation between equipment fatigue state and functional tag number. This solves the problem that the static management mode in traditional predictive maintenance cannot accurately quantify the fatigue degree of equipment service. While meeting the requirements of equipment full life cycle traceability, it drives the generation of predictive maintenance early warning through topological evolution data, significantly enhancing the system's safety monitoring capability for key equipment components. Attached Figure Description

[0027] Figure 1 This is a module diagram of the power plant raw coal storage equipment full life cycle traceability system based on KKS coding according to the present invention;

[0028] Figure 2 This is a flowchart of the power plant raw coal storage equipment full life cycle traceability method of the present invention;

[0029] Figure 3 This is a schematic diagram of the global state topology network structure based on a directed acyclic graph (DAG) according to the present invention;

[0030] Figure 4 This is a flowchart of the magnetic Barkhausen noise signal processing and initial physical feature instantiation process of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Example 1

[0033] See attached document Figures 1-2 This invention provides a full lifecycle traceability system for power plant raw coal bunker equipment based on KKS coding, comprising: a physical sensing layer 100, an edge computing layer 200, and a logic evolution and control layer 300.

[0034] The physical sensing layer 100 is deployed at the raw coal bunker equipment site to perform excitation and signal acquisition of the microscopic physical characteristics of the material. The physical sensing layer 100 includes a portable magnetic field excitation and detection device 110 and a raw coal bunker equipment entity 120. The portable magnetic field excitation and detection device 110 applies an alternating excitation magnetic field to a preset reference surface of the raw coal bunker equipment entity 120. The portable magnetic field excitation and detection device 110 acquires the original voltage signal of magnetic Barkhausen noise generated by the movement of magnetic domain walls inside the raw coal bunker equipment entity 120, and converts the original voltage signal of magnetic Barkhausen noise into a digital signal.

[0035] The edge computing layer 200 establishes a communication connection with the physical sensing layer 100 to perform signal processing and entity identification calculation on the acquired digital signals. The edge computing layer 200 includes a signal preprocessing module 210, a feature vector construction module 220, and a hash calculation module 230. The signal preprocessing module 210 receives digital signals and performs bandpass filtering. The feature vector construction module 220 extracts the root mean square value and Barkhausen jump peak parameter based on the filtered signal to generate an initial physical feature vector. The hash calculation module 230 performs a hash operation on the initial physical feature vector using a one-way hash algorithm to generate the corresponding intrinsic entity identifier.

[0036] The logic evolution and control layer 300 establishes a data connection with the edge computing layer 200 to construct the device flow topology and perform node state updates. The logic evolution and control layer 300 includes a topology network construction module 310, a stress environment configuration module 320, an adaptive addressing matching module 330, and a state evolution control module 340.

[0037] Topology network construction module 310 constructs a directed acyclic graph structure, and sets the global state topology network model as follows: ,in For a set of nodes, Let be a set of directed edges. To record the temporal and spatial binding relationships of nodes, the topology network construction module 310 extracts the spatial location identifiers of the raw coal bunker process flow according to the power plant's KKS coding rules, and uses these identifiers as node identifiers to group the nodes. Divided into a subset of static functional nodes representing spatial locations A subset of entity nodes representing entity 120 of the raw coal bunker equipment. static functional node subset subset of entity nodes Satisfying the relation as well as ;

[0038] Stress environment configuration module 320 is a subset of static functional nodes. Each node in the system is configured with a spatial environment stress spectrum matrix, which includes the vibration stress amplitude, dominant frequency vibration characteristics, and environmental temperature parameters of the corresponding node's spatial location.

[0039] The adaptive addressing and matching module 330 receives the current physical feature vector uploaded by the edge computing layer 200. It retrieves the pre-set spatial environment stress spectrum matrix and entity service time parameters of the corresponding node, calculates the feature expectation drift compensation tensor using the material degradation model, updates the initial physical feature vector using the feature expectation drift compensation tensor, and generates an expected comparison feature vector. The adaptive addressing and matching module 330 calculates the Mahalanobis distance between the current physical feature vector and the expected comparison feature vector, and determines whether the Mahalanobis distance is less than a preset addressing and matching threshold. If the Mahalanobis distance is less than the addressing and matching threshold, the adaptive addressing and matching module 330 determines that the addressing is successful and establishes the raw coal bunker equipment entity 120 and a subset of static functional nodes in the directed acyclic graph structure constructed by the topology network construction module 310. The mapping relationship between the corresponding nodes is determined by the state evolution control module 340, which calculates the feature drift residual vector between the current physical feature vector and the initial physical feature vector based on the matching result of the adaptive addressing matching module 330.

[0040] The state evolution control module 340 transforms the feature drift residual vector into a material feature degradation index through a multidimensional fatigue mapping function, and updates the directed edge set using the material feature degradation index. The weight value of the corresponding edge is set in the state evolution control module 340. When the weight value of the corresponding edge reaches the preset critical threshold, the state evolution control module 340 triggers and outputs a device maintenance early warning command.

[0041] Furthermore, based on the aforementioned KKS-coded power plant raw coal storage equipment lifecycle traceability system, the lifecycle traceability of power plant raw coal storage equipment includes the following steps:

[0042] In step S101, the topology network construction module 310 in the logic evolution and control layer 300 establishes a global state topology network model. The topology network construction module 310 generates a static functional node subset in the node set of the global state topology network model. The stress environment configuration module 320 obtains the environmental data of the measuring points at the raw coal bunker site and configures the spatial environment stress spectrum matrix for each node in the static functional node subset.

[0043] In step S102, during the raw coal storage equipment entity 120 storage stage, the portable magnetic field excitation and detection device 110 in the physical sensing layer 100 applies an alternating excitation magnetic field to the preset reference surface of the raw coal storage equipment entity 120. The portable magnetic field excitation and detection device 110 collects the original voltage signal of the magnetic Barkhausen noise of the raw coal storage equipment entity 120 and converts the original voltage signal of the magnetic Barkhausen noise into a digital signal. The signal preprocessing module 210 in the edge computing layer 200 performs a bandpass filtering operation on the digital signal. The feature vector construction module 220 extracts the feature parameters of the filtered digital signal and generates an initial physical feature vector. The hash calculation module 230 performs a hash operation on the initial physical feature vector to generate an intrinsic entity identifier. The topology network construction module 310 generates the corresponding entity node in the node set according to the intrinsic entity identifier and generates a directed edge with timestamp information between the corresponding node in the static functional node subset and the entity node to complete the initial mapping of the equipment state.

[0044] In step S103, during the service of the raw coal bunker equipment entity 120, the portable magnetic field excitation and detection device 110 collects the current physical feature vector of the raw coal bunker equipment entity 120. The edge computing layer 200 transmits the current physical feature vector to the logic evolution and control layer 300. The adaptive addressing matching module 330 extracts the corresponding node's preset spatial environment stress spectrum matrix and service time parameters. The adaptive addressing matching module 330 calculates the feature expected drift compensation tensor in conjunction with the material degradation model. The feature expected drift compensation tensor satisfies the formula:

[0045]

[0046] In the formula, For the characteristic expected drift compensation tensor, The stress spectrum matrix of the space environment. For service time parameters, This is the drift sensitivity coefficient matrix. This is the time drift compensation coefficient;

[0047] The adaptive addressing matching module 330 updates the initial physical feature vector using the feature expectation drift compensation tensor to obtain the expected comparison feature vector. The adaptive addressing matching module 330 calculates the Mahalanobis distance between the current physical feature vector and the expected comparison feature vector. The Mahalanobis distance satisfies the formula:

[0048]

[0049] In the formula, The Mahalanobis distance, This is the current physical feature vector. To compare the feature vectors, The drift tolerance covariance matrix is ​​used to determine the identity authentication of the raw coal bunker equipment entity 120 when the Mahalanobis distance is less than the addressing matching threshold.

[0050] Step S104: After successful identity authentication, the state evolution control module 340 calculates the feature drift residual vector between the current physical feature vector and the initial physical feature vector. The state evolution control module 340 inputs the feature drift residual vector into the multidimensional fatigue mapping function to calculate and obtain the material feature degradation index. The material feature degradation index satisfies the formula:

[0051]

[0052] In the formula, The material characteristic degradation index, For the feature drift residual vector, The first conversion coefficient constant, This is the second conversion coefficient constant. For L2 norm operations, As a multidimensional gradient operator, the state evolution control module 340 uses the material characteristic deterioration index as a weight value and updates it to the corresponding directed edge. The state evolution control module 340 monitors the change in the weight value of the directed edge. When the weight value reaches the preset critical threshold, the state evolution control module 340 outputs an equipment maintenance early warning command. When the equipment dismantling action occurs, the topology network construction module 310 removes the current directed edge and records the current weight value. The disconnected topology directed edge is solidified as a static historical snapshot edge, and the historical flow record of the raw coal bunker equipment entity 120 is retained in the global state topology network model to form a set of historical connected subgraphs, realizing the full life cycle traceability of equipment flow across nodes.

[0053] Example 2

[0054] See attached document Figures 3-4 Based on Embodiment 1, and in conjunction with the aforementioned disclosed physical perception layer 100, edge computing layer 200, and logic evolution and control layer 300, the topology network construction module 310 within the logic evolution and control layer 300 performs the initial construction of the directed acyclic graph structure. The topology network construction operation includes:

[0055] Step S201: The topology network construction module 310 extracts the equipment hierarchical relationships in the process flow of the power plant's raw coal bunker, and maps these relationships to a global state topology network model. The global state topology network model is represented using a directed acyclic graph data structure and defined as graph variables. In the directed acyclic graph data structure, Defined as a set of nodes containing the identifiers of all tracked objects in the system. Defined as a set of directed edges that record and identify time-bound relationships;

[0056] Step S202, Topology network construction module 310 pairs of node sets The node classification operation is performed. The topology network construction module 310 extracts the spatial location identifiers of the raw coal bunker process flow according to the power plant KKS coding rules, and uses them as node identifiers to generate a subset of static functional nodes. and entity node subset static functional node subset subset of entity nodes Satisfying the relation:

[0057]

[0058]

[0059] In the formula, For a set of nodes, A subset of static functional nodes. A subset of entity nodes For the union operation, For intersection operations, The empty set operator, a subset of static function nodes. elements in This represents a subset of entity nodes representing specific spatial physical location indexes defined under the power plant coding standard. elements in Entity 120 represents the raw coal bunker equipment in the life cycle transition stage;

[0060] Step S203, the topology network construction module 310 constructs the static functional node subset according to the power plant facility configuration ledger and the KKS coding rules. The initialization of each spatial location index corresponds to an independent running process node. The attributes of the running process node include spatial coordinate parameters and system classification parameters to which the process belongs. After the original coal bunker equipment entity 120 completes the registration of the intrinsic entity identifier, the topology network construction module 310 registers the entity node subset. Generate a device object node that maps the raw coal bunker equipment entity 120;

[0061] Step S204, the topology network construction module 310 configures the set of directed edges. The generation rules stipulate that when the raw coal bunker equipment entity 120 is installed at a specific spatial physical location, the topology network construction module 310 will generate the directed edge set. Establish a subset of entity nodes The corresponding element in the middle points to a subset of static functional nodes. For the directed connections of corresponding elements, the topology network construction module 310 assigns initial timestamp variables to the established directed connections. A directed connection satisfies the formula:

[0062]

[0063] In the formula, For the initial timestamp variable Directed connections established at any time A subset of entity nodes The corresponding element in A subset of static functional nodes The corresponding element in This represents the mapping relationship between the start and end points of a directed connection;

[0064] In step S205, the topology network construction module 310 initializes the state evolution weight parameters of the established directed connections, and the initialized state evolution weight parameters satisfy the formula:

[0065]

[0066] In the formula, For directed connections, the initial timestamp variable The state evolution weight parameter at time step The residual fatigue weight parameter is accumulated in the historical circulation record of the current raw coal storage equipment entity 120. For the raw coal storage equipment entity 120 that is registered for the first time, the topology network construction module 310 sets the value of the residual fatigue weight parameter to zero.

[0067] Furthermore, combined with the stress environment configuration module 320 within the logic evolution and control layer 300, the configuration and quantification of the space environment stress spectrum matrix can be performed using the following steps:

[0068] Step S301: The stress environment configuration module 320 establishes a communication connection with the on-site sensor network of the raw coal bunker. The stress environment configuration module 320 configures the static functional node subset. The spatial physical location index of each element is used to extract the historical operating condition data of the corresponding spatial physical location. The historical operating condition data includes mechanical environment state parameters and physical environment state parameters. The mechanical environment state parameters include continuous vibration waveform data in the time dimension, and the physical environment state parameters include environmental temperature monitoring sequence and environmental humidity monitoring sequence.

[0069] In step S302, the stress environment configuration module 320 quantifies the extracted historical operating condition data, performs a fast Fourier transform on the continuous vibration waveform data to convert the time-domain waveform to the frequency domain, extracts the peak frequency parameter in the spectrum as the main frequency vibration characteristic parameter, calculates the root mean square value of the continuous vibration waveform data, and uses the root mean square value as the average vibration stress amplitude, performs an extreme value extraction operation on the environmental temperature monitoring sequence within a preset time sliding window, extracts the maximum value in the environmental temperature monitoring sequence as the extreme environmental temperature parameter, and retrieves the rated mechanical load parameter, coal powder friction coefficient measure, and environmental corrosion index factor of the corresponding spatial physical location from the preset physical property mapping database according to the rated operating parameters and coal quality characteristics of the raw coal bunker equipment through a lookup table matching instruction.

[0070] In step S303, the stress environment configuration module 320 constructs the quantified parameters into a two-dimensional feature matrix according to a preset dimension, generating a spatial environment stress spectrum matrix corresponding to the spatial physical location. The spatial environment stress spectrum matrix satisfies the formula:

[0071]

[0072] In the formula, A subset of static functional nodes medium elements The corresponding space environment stress spectrum matrix, The average vibration stress amplitude, The characteristic parameter of the dominant frequency vibration. This refers to the extreme ambient temperature parameter. The friction coefficient of pulverized coal is used as a measure. The rated mechanical load parameter, It is an environmental corrosion index factor;

[0073] In step S304, the stress environment configuration module 320 encapsulates the generated spatial environment stress spectrum matrix into a node attribute dataset, and then writes the node attribute dataset into the elements of the global state topology network model. In the storage structure, the node binding of stress environment parameters is completed. The bound spatial environment stress spectrum matrix is ​​used as the prior environment benchmark parameter to provide the calculation basis for the adaptive addressing matching module 330.

[0074] Furthermore, combining the aforementioned disclosed physical sensing layer 100 and edge computing layer 200, the acquisition of the original voltage signal of magnetic Barkhausen noise and the construction of the initial physical feature vector include the following steps:

[0075] In step S401, the portable magnetic field excitation and detection device 110 in the physical sensing layer 100 includes a signal generating device and a detection probe. During the storage and subsequent inspection and verification stages of the raw coal bunker equipment entity 120, the detection probe of the portable magnetic field excitation and detection device 110 is attached to the preset reference surface of the raw coal bunker equipment entity 120. The portable magnetic field excitation and detection device 110 outputs a low-frequency alternating excitation magnetic field to the preset reference surface, which induces the jumping motion of the microscopic magnetic domain walls inside the raw coal bunker equipment entity 120. The portable magnetic field excitation and detection device 110 collects the original voltage signal of the magnetic Barkhausen noise induced by the jumping motion of the microscopic magnetic domain walls through the induction coil in the detection probe. The analog-to-digital conversion unit built into the portable magnetic field excitation and detection device 110 converts the original voltage signal of the magnetic Barkhausen noise into a digital signal in discrete time series form and transmits the digital signal to the edge computing layer 200.

[0076] In step S402, the signal preprocessing module 210 in the edge computing layer 200 receives the digital signal and performs bandpass filtering on the digital signal to filter out the low-frequency excitation fundamental signal and high-frequency environmental interference noise signal, thereby obtaining the filtered signal sequence within the effective frequency band. The signal preprocessing module 210 performs envelope detection processing on the filtered signal sequence to extract the outer contour features of the filtered signal sequence and generate the envelope signal sequence.

[0077] In step S403, the feature vector construction module 220 within the edge computing layer 200 calculates time-domain feature parameters based on the filtered signal sequence. The feature vector construction module 220 calculates the root mean square (RMS) value of the filtered signal sequence within one magnetization period and sets the RMS value as the signal RMS value parameter. The signal RMS value parameter satisfies the formula:

[0078]

[0079] In the formula, The root mean square value parameter of the signal. The total number of sampling points within one magnetization cycle. For discrete time series indexing, For the filtered signal sequence at the index The feature vector construction module 220 retrieves the maximum amplitude point in the envelope signal sequence and extracts the maximum amplitude point as the Barkhausen jump peak parameter. The feature vector construction module 220 performs discrete accumulation integration on the envelope signal sequence in the time domain to obtain the envelope integral area parameter, which satisfies the formula:

[0080]

[0081] In the formula, The integral area parameter of the envelope. For the envelope signal sequence in the index The amplitude at that point, The system sampling time interval;

[0082] In step S404, the feature vector construction module 220 performs a discrete Fourier transform on the filtered signal sequence to calculate and obtain the peak power spectral density parameter in the frequency domain. The feature vector construction module 220 concatenates and combines the extracted root mean square value parameter, Barkhausen jump peak parameter, envelope integral area parameter, and peak power spectral density parameter to construct an initial physical feature vector. The initial physical feature vector satisfies the formula:

[0083]

[0084] In the formula, For the corresponding number The initial physical feature vector of 120 raw coal bunker equipment entities The root mean square value parameter of the signal. For the peak parameter of the Buckhausen jump, The integral area parameter of the envelope. The peak energy parameter of the power spectral density. Representing the matrix transpose operator, the constructed initial physical feature vector serves as the physical feature reference for the raw coal bunker equipment entity 120 and is stored in the system storage space awaiting subsequent feature identification registration operations.

[0085] Furthermore, combining the hash calculation module 230 within the edge computing layer 200 and the topology network construction module 310 within the logical evolution and control layer 300, the feature identifier registration and entity node generation operation includes the following steps:

[0086] In step S501, the hash calculation module 230 receives the initial physical feature vector transmitted by the feature vector construction module 220. The hash calculation module 230 extracts each feature parameter from the initial physical feature vector and performs fixed-point numerical truncation processing on each feature parameter to generate a fixed-point numerical sequence. The hash calculation module 230 converts the fixed-point numerical sequence into a binary data stream. The hash calculation module 230 retrieves the secure hash algorithm model from the system memory and uses the secure hash algorithm model to perform hash mapping calculation on the binary data stream to generate intrinsic entity identifiers. The intrinsic entity identifiers satisfy the formula:

[0087]

[0088] In the formula, For the first The intrinsic entity identifier of 120 raw coal bunker equipment entities. For secure hash algorithm functions, This is the initial physical feature vector;

[0089] In step S502, the hash calculation module 230 uploads the intrinsic entity identifier to the topology network construction module 310. The topology network construction module 310 receives the intrinsic entity identifier and allocates node storage units in the entity node subset of the global state topology network model. The topology network construction module 310 writes the intrinsic entity identifier as the index primary key into the node storage unit and generates the entity node corresponding to the raw coal bunker equipment entity 120 in the entity node subset.

[0090] In step S503, the topology network construction module 310 obtains the initial physical feature vector and equipment entry time parameter of the raw coal bunker equipment entity 120. The topology network construction module 310 writes the initial physical feature vector and equipment entry time parameter as node attribute data into the storage unit of the entity node. The topology network construction module 310 sets the connection state parameter of the entity node to the initial independent state, thus completing the instantiation mapping of the entity node of the raw coal bunker equipment entity 120 in the global state topology network model.

[0091] Furthermore, combining the logic evolution and the topology network construction module 310 within the control layer 300, the operation of action connection generation and initial state mapping includes the following steps:

[0092] In step S601, the topology network construction module 310 receives the equipment installation confirmation instruction sent by the raw coal bunker field terminal. The topology network construction module 310 parses the equipment installation confirmation instruction and extracts the intrinsic entity identifier and the target installation location index carried in the equipment installation confirmation instruction. The topology network construction module 310 searches for and determines the matching source entity node in the entity node subset according to the intrinsic entity identifier. The topology network construction module 310 searches for and determines the matching target functional node in the static functional node subset according to the target installation location index.

[0093] In step S602, the topology network construction module 310 performs a directed connection write operation in the directed edge set of the global state topology network model. Starting from the determined source entity node and ending at the determined target functional node, the topology network construction module 310 generates a topological directed edge connecting the source entity node and the target functional node. The topology network construction module 310 obtains the current time parameter output by the system clock synchronization module and writes the current time parameter as an installation timestamp attribute into the data structure of the topology directed edge. The topology network construction module 310 updates the node connectivity state in the adjacency matrix of the directed acyclic graph data structure. The update of the node connectivity state in the adjacency matrix satisfies the formula:

[0094]

[0095] In the formula, This represents the Boolean value indicating the connectivity between the source entity node and the target functional node in the adjacency matrix. This is the row index parameter of the source entity node in the adjacency matrix. The column index parameter of the target functional node in the adjacency matrix;

[0096] In step S603, the topology network construction module 310 performs an initial mapping operation of the state evolution weights on the generated topology directed edges. The topology network construction module 310 reads the storage unit of the source entity node, extracts the historical connection edge record parameters recorded in the attribute data of the source entity node, and determines the service attributes of the raw coal bunker equipment entity corresponding to the source entity node based on the historical connection edge record parameters. When the historical connection edge record parameters are empty, the topology network construction module 310 determines that the raw coal bunker equipment entity is in the first entry registration state and generates a residual fatigue weight parameter with a value of zero. When the historical connection edge record parameters are not empty, the topology network construction module 310 determines that the raw coal bunker equipment entity is in the historical transfer state and extracts the residual fatigue weight parameter saved when the last disconnection occurred in the attribute data of the source entity node.

[0097] In step S604, the topology network construction module 310 assigns the acquired residual fatigue weight parameters to the weight attribute field of the directed edges of the topology. The topology network construction module 310 constructs the initial state feature set of the directed edges of the topology, and the initial state feature set satisfies the formula:

[0098]

[0099] In the formula, Let be the set of initial state features of the directed edges of the topology at the installation timestamp. For the installation timestamp attribute of the directed edge record in the topology, The connected Boolean value of the adjacency matrix. For residual fatigue weighting parameters, Representing the matrix transpose operator, the topology network construction module 310 submits the initial state feature set to the system database for persistent storage, completing the initial connection record between devices and locations.

[0100] Furthermore, in conjunction with the adaptive addressing matching module 330 within the physical perception layer 100, the edge computing layer 200, and the logic evolution and control layer 300, the prior environment compensation and adaptive addressing matching operation includes the following steps:

[0101] In step S701, during the service inspection phase of the raw coal bunker equipment entity 120, the portable magnetic field excitation and detection device 110 in the physical sensing layer 100 collects the current raw voltage signal of the raw coal bunker equipment entity 120 and converts the current raw voltage signal of the raw coal bunker equipment entity 120 into a digital signal. The feature vector construction module 220 in the edge computing layer 200 processes the digital signal and generates the current physical feature vector. The feature vector construction module 220 sends the current physical feature vector to the adaptive addressing matching module 330 in the logic evolution and control layer 300. The adaptive addressing matching module 330 obtains the inspection location index sent by the field terminal.

[0102] In step S702, the adaptive addressing matching module 330 retrieves and determines the matching static functional node in the static functional node subset maintained by the topology network construction module 310 according to the index of the location to be inspected. The adaptive addressing matching module 330 extracts the spatial environment stress spectrum matrix bound to the matching static functional node. The adaptive addressing matching module 330 queries the topology directed edge with the matching static functional node as the endpoint in the directed edge set and reads the installation timestamp attribute recorded in the topology directed edge. The adaptive addressing matching module 330 calls the current system time parameter, calculates the time difference between the current system time parameter and the installation timestamp attribute, and obtains the service time parameter.

[0103] In step S703, the adaptive addressing matching module 330 calls the preset material degradation model, combines the space environment stress spectrum matrix and service time parameters, calculates the feature expected drift compensation tensor, and reads the initial physical feature vector stored in the corresponding entity node through the starting index of the topological directed edge. The adaptive addressing matching module 330 performs vector addition operation on the initial physical feature vector and the feature expected drift compensation tensor to obtain the expected comparison feature vector. The update logic of the expected comparison feature vector satisfies the formula:

[0104]

[0105] In the formula, To compare the feature vectors, This refers to the initial physical feature vector stored within the entity node. To calculate the expected drift compensation tensor of the obtained features;

[0106] In step S704, the adaptive addressing matching module 330 extracts the drift tolerance covariance matrix contained in the attributes of the matched static functional nodes. The drift tolerance covariance matrix is ​​generated by the system based on the historical discrete drift sample set of the corresponding spatial physical location. The adaptive addressing matching module 330 calculates the Mahalanobis distance based on the current physical feature vector, the expected matching feature vector, and the drift tolerance covariance matrix. The adaptive addressing matching module 330 compares the calculated Mahalanobis distance with the addressing matching threshold set by the system and outputs the addressing judgment result. The addressing judgment result satisfies the formula:

[0107]

[0108] In the formula, For the addressing determination result, a Boolean variable. The Mahalanobis distance between the current physical feature vector and the expected alignment feature vector. This is a preset addressing matching threshold;

[0109] In step S705, when the value of the Boolean variable of the addressing determination result is 1, the adaptive addressing matching module 330 determines that the identity authentication of the raw coal bunker equipment entity 120 is successful, and determines that the on-site raw coal bunker equipment entity is consistent with the entity object recorded in the topology directed edge. The adaptive addressing matching module 330 sends a state evolution permission signal to the state evolution control module 340. When the value of the Boolean variable of the addressing determination result is 0, the adaptive addressing matching module 330 determines that the identity authentication of the raw coal bunker equipment entity 120 fails, and the adaptive addressing matching module 330 outputs a device abnormality blocking signal to intercept the system's data update action for the current topology directed edge.

[0110] Furthermore, combining the state evolution control module 340 within the logic evolution and control layer 300, the fatigue quantization mapping and graph evolution edge weight dynamic update operation includes the following steps:

[0111] In step S801, the state evolution control module 340 receives the state evolution permission signal sent by the adaptive addressing matching module 330. In response to the state evolution permission signal, the state evolution control module 340 extracts the initial physical feature vector stored in the entity node and the current physical feature vector transmitted by the edge computing layer 200. The state evolution control module 340 subtracts the initial physical feature vector from the current physical feature vector and calculates and generates the feature drift residual vector.

[0112] In step S802, the state evolution control module 340 retrieves the multidimensional fatigue mapping function from the system storage resources. The multidimensional fatigue mapping function is pre-generated by the system based on the offline physical attenuation calibration experiment of the raw coal bunker equipment entity. The state evolution control module 340 imports the feature drift residual vector as an input parameter into the multidimensional fatigue mapping function. The state evolution control module 340 calculates the L2 norm square value and the multidimensional gradient feature parameter of the feature drift residual vector respectively. The state evolution control module 340 retrieves the preset first conversion coefficient constant and second conversion coefficient constant. The state evolution control module 340 multiplies the L2 norm square value by the first conversion coefficient constant to obtain the first weighted parameter. The state evolution control module 340 multiplies the multidimensional gradient feature parameter by the second conversion coefficient constant to obtain the second weighted parameter. The state evolution control module 340 adds the first weighted parameter and the second weighted parameter to obtain the material characteristic deterioration index.

[0113] In step S803, the state evolution control module 340 retrieves the topological directed edges connecting the current entity node and the matching static functional node from the set of directed edges in the global state topology network model. The state evolution control module 340 reads the weight attribute field of the topological directed edges, extracts the state evolution weight parameters recorded at the initial binding time, and performs a numerical accumulation operation between the state evolution weight parameters and the calculated material characteristic deterioration index to obtain dynamically updated weight values. The dynamically updated weight values ​​satisfy the formula:

[0114]

[0115] In the formula, This refers to the dynamic update of the weight values ​​of the directed edges in the topology under the current system time parameters. Let the weight parameter be the state evolution weight of the directed edge in the topology at the initial installation time. To calculate the material characteristic degradation index;

[0116] In step S804, the state evolution control module 340 writes the dynamically updated weight values ​​into the weight attribute field of the topological directed edge, obtains the current system time parameter, updates the current system time parameter to the observation timestamp attribute of the topological directed edge, and synchronizes the updated topological directed edge data to the system database, thus completing the dynamic assignment of the topological directed edge state attributes.

[0117] Furthermore, combining the topology network construction module 310 and the state evolution control module 340 within the logic evolution and control layer 300, the predictive maintenance of triggering and lifecycle flow records includes the following steps:

[0118] Step S901: The state evolution control module 340 reads the dynamically updated weight values ​​in the directed edges of the topology. The state evolution control module 340 extracts the preset failure critical weight threshold from the attributes of the corresponding matched static functional nodes. The state evolution control module 340 compares the dynamically updated weight values ​​with the failure critical weight threshold. Based on the comparison result, the state evolution control module 340 generates maintenance warning state parameters, which satisfy the formula:

[0119]

[0120] In the formula, To maintain the early warning status parameters, To dynamically update the weight values ​​of directed edges in the topology. The preset failure threshold weight;

[0121] In step S902, when the value of the maintenance warning state parameter is 1, the state evolution control module 340 extracts the spatial location index of the static functional node connected by the topological directed edge and the entity identification parameter of the entity node. The state evolution control module 340 splices the spatial location index, entity identification parameter and dynamic update weight value according to the preset data frame format to generate the equipment maintenance warning command. The state evolution control module 340 transmits the equipment maintenance warning command to the raw coal bunker field terminal.

[0122] In step S903, the topology network construction module 310 receives the equipment dismantling confirmation command sent by the raw coal bunker field terminal. The topology network construction module 310 parses the equipment dismantling confirmation command, obtains the dismantling timestamp attribute, updates the Boolean value of the connection between the corresponding source entity node and the target functional node to zero in the adjacency matrix of the directed acyclic graph data structure, and removes the current topological directed edge connection. The topology network construction module 310 then solidifies the removed topological directed edge as a static historical snapshot edge, which satisfies the formula:

[0123]

[0124] In the formula, For static historical snapshot edges containing time intervals, For the installation timestamp attribute, To remove the timestamp attribute, For the source entity node, For the target functional node, A closed-interval parameter characterizing the lifetime of node connections;

[0125] In step S904, the topology network construction module 310 reads the final dynamic update weight value corresponding to the disassembly timestamp attribute, sets the final dynamic update weight value as the current residual fatigue weight parameter of the source entity node, and writes the current residual fatigue weight parameter as a historical connection edge record parameter into the node attribute data structure of the source entity node.

[0126] In step S905, the topology network construction module 310 restores the connection state parameters of the source entity node to their initial independent state. The static historical snapshot edges, together with the flow records of each flow of the source entity node, are stored in the global state topology network model. In terms of data structure, this forms a set of historical connection subgraphs starting from the source entity node. When the source entity node receives a new device installation confirmation instruction and establishes a new topology directed edge connection, the topology network construction module 310 extracts the updated current residual fatigue weight parameter and assigns the current residual fatigue weight parameter to the initial weight attribute field of the new topology directed edge, maintaining the continuous mapping between physical characteristics and fatigue data in the cross-node flow state.

Claims

1. A full lifecycle traceability system for power plant raw coal bunker equipment based on KKS coding, characterized in that, include: The system comprises a physical sensing layer, an edge computing layer, and a logic evolution and control layer. The physical sensing layer is deployed at the raw coal bunker equipment site to perform excitation and signal acquisition of the microscopic physical characteristics of the material. The physical sensing layer includes a portable magnetic field excitation and detection device and the raw coal bunker equipment entity. During periodic inspections or equipment transfer verifications, the portable magnetic field excitation and detection device applies an alternating excitation magnetic field to a preset reference surface of the raw coal bunker equipment entity to acquire the original voltage signal of magnetic Barkhausen noise generated by the movement of magnetic domain walls inside the raw coal bunker equipment entity, and converts the original voltage signal of magnetic Barkhausen noise into a digital signal. The edge computing layer establishes a communication connection with the physical sensing layer to perform signal processing and entity identification calculation on the acquired digital signals. The edge computing layer includes a signal preprocessing module, a feature vector construction module, and a hash calculation module. The signal preprocessing module receives digital signals and performs bandpass filtering. The feature vector construction module extracts time-domain and frequency-domain feature parameters based on the filtered signal to generate an initial physical feature vector. The hash calculation module performs a hash operation on the initial physical feature vector using a one-way hash algorithm to generate the corresponding intrinsic entity identifier. The logical evolution and control layer establishes a data connection with the edge computing layer to construct the device flow topology and perform node state updates. The logical evolution and control layer includes a topology network construction module, a stress environment configuration module, an adaptive addressing matching module, and a state evolution control module. The topology network construction module constructs a directed acyclic graph structure, and sets the global state topology network model as follows: ,in For a set of nodes, Let be a set of directed edges. The topology network construction module, used to record the temporal and spatial binding relationships of nodes, extracts the spatial location identifiers of the raw coal bunker process flow according to the power plant's KKS coding rules, and uses these identifiers as node identifiers to aggregate nodes. Divided into a subset of static functional nodes representing spatial locations A subset of entity nodes representing the raw coal bunker equipment entities static functional node subset subset of entity nodes Satisfying the relation as well as ; The stress environment configuration module is a subset of static functional nodes. Each node in the system is configured with a spatial environment stress spectrum matrix, which includes the vibration stress amplitude, dominant frequency vibration characteristics, and environmental temperature parameters of the corresponding node's spatial location. The adaptive addressing and matching module receives the current physical feature vector uploaded by the edge computing layer. The adaptive addressing and matching module retrieves the spatial environment stress spectrum matrix and entity service time parameters preset by the corresponding node, and calculates the feature expected drift compensation tensor in combination with the material degradation model. The adaptive addressing and matching module updates the initial physical feature vector using the feature expected drift compensation tensor and generates the expected comparison feature vector. The adaptive addressing and matching module calculates the Mahalanobis distance between the current physical feature vector and the expected comparison feature vector, and determines whether the Mahalanobis distance is less than the preset addressing and matching threshold. When the Mahalanobis distance is less than the addressing and matching threshold, the adaptive addressing and matching module determines that the addressing is successful, and establishes a mapping relationship between the raw coal bunker equipment entity and the corresponding node in the static functional node subset in the directed acyclic graph structure constructed by the topology network construction module. The state evolution control module calculates the feature drift residual vector between the current physical feature vector and the initial physical feature vector based on the matching result of the adaptive addressing matching module. The state evolution control module converts the feature drift residual vector into a material feature degradation index through a multidimensional fatigue mapping function, and uses the material feature degradation index to update the weight value of the corresponding edge in the directed edge set E. When the weight value of the corresponding edge reaches a preset critical threshold, the state evolution control module triggers and outputs a device maintenance early warning command. When a device disassembly command is received and the current topology directed edge connection is released, the topology network construction module solidifies the released topology directed edge into a static historical snapshot edge, and retains the flow record of the entity node in the global state topology network model to form a historical edge subgraph set, realizing the full life cycle traceability of the device's cross-node flow.

2. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The topology network construction module extracts the equipment hierarchical association relationship of the raw coal bunker process flow in the power plant and maps the equipment hierarchical association relationship into the global state topology network model. The topology network construction module initializes each spatial location index in the static functional node subset according to the power plant facility configuration ledger and the KKS coding rules. Each spatial location index corresponds to an independent running process node. The attributes of the running process node include spatial coordinate parameters and system classification parameters to which the process belongs. After the raw coal bunker equipment entity completes the registration of intrinsic entity identifier, the topology network construction module generates equipment object nodes that map the raw coal bunker equipment entity in the entity node subset.

3. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The stress environment configuration module extracts historical operating condition data for each spatial physical location according to the spatial physical location index of each element in the static functional node subset; the stress environment configuration module quantifies the extracted historical operating condition data, calculates and obtains the average vibration stress amplitude, main frequency vibration characteristic parameter, and extreme environmental temperature parameter for the corresponding spatial physical location, and retrieves the rated mechanical load parameter, coal powder friction coefficient measure, and environmental corrosion index factor for the corresponding spatial physical location from the preset physical property mapping database through a lookup matching command based on the rated operating parameters and coal quality characteristics of the raw coal bunker equipment. The stress environment configuration module constructs the quantified parameters into a two-dimensional feature matrix according to a preset dimension, generating a spatial environment stress spectrum matrix corresponding to the spatial physical location.

4. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The signal preprocessing module within the edge computing layer performs envelope detection processing on the filtered signal sequence, extracts the outer contour features of the filtered signal sequence, and generates an envelope signal sequence. The feature vector construction module calculates the root mean square value of the filtered signal sequence within one magnetization period as the root mean square value parameter, retrieves the maximum amplitude point in the envelope signal sequence as the Barkhausen jump peak parameter, and performs discrete accumulation integration on the envelope signal sequence in the time domain to obtain the envelope integral area parameter. The feature vector construction module performs discrete Fourier transform calculation on the filtered signal sequence to obtain the peak energy parameter of the power spectral density, and then concatenates and combines the extracted root mean square value parameter, Barkhausen jump peak parameter, envelope integral area parameter, and power spectral density peak energy parameter to construct the initial physical feature vector.

5. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The hash calculation module extracts the feature parameters from the initial physical feature vector and performs fixed-point numerical truncation on each feature parameter to generate a fixed-point numerical sequence. The hash calculation module converts the fixed-point numerical sequence into a binary data stream and retrieves the secure hash algorithm model from the system memory. It then uses the secure hash algorithm model to perform hash mapping calculation on the binary data stream to generate intrinsic entity identifiers. The topology network construction module allocates node storage units in the entity node subset of the global state topology network model, writes the intrinsic entity identifiers as index primary keys into the node storage units to generate entity nodes, and writes the initial physical feature vector and the device entry time parameter as node attribute data.

6. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The topology network construction module receives the device installation confirmation instruction, extracts the intrinsic entity identifier and the target installation location index carried by it, and retrieves and determines the matching source entity node and the matching target functional node respectively; the topology network construction module generates a topology directed edge with the source entity node as the starting point and the target functional node as the ending point, and writes the current time parameter as the installation timestamp attribute into the data structure of the topology directed edge; The topology network construction module extracts the historical connection edge record parameters recorded in the attribute data of the source entity node to determine the service attributes, obtains the residual fatigue weight parameter, and assigns the obtained residual fatigue weight parameter to the weight attribute field of the topology directed edge, thus completing the initial binding of action connection edge and state attribute.

7. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The adaptive addressing matching module calls a preset material degradation model and calculates the characteristic expected drift compensation tensor by combining the space environment stress spectrum matrix and service time parameters. The characteristic expected drift compensation tensor satisfies the formula: In the formula, For the characteristic expected drift compensation tensor, The stress spectrum matrix of the space environment. For service time parameters, This is the drift sensitivity coefficient matrix. This is the time drift compensation coefficient; The adaptive addressing matching module performs a vector addition operation on the initial physical feature vector and the feature expected drift compensation tensor to obtain the expected comparison feature vector; The adaptive addressing matching module extracts the drift tolerance covariance matrix contained in the attributes of the matched static functional nodes, and calculates the Mahalanobis distance based on the current physical feature vector, the expected alignment feature vector, and the drift tolerance covariance matrix. The Mahalanobis distance satisfies the formula: In the formula, The Mahalanobis distance, This is the current physical feature vector. To compare the feature vectors, This is the drift tolerance covariance matrix.

8. The power plant raw coal bunker equipment full life cycle traceability system based on KKS coding according to claim 1, characterized in that, The state evolution control module calculates a feature drift residual vector by subtracting the initial physical feature vector from the current physical feature vector. The state evolution control module then uses the feature drift residual vector as an input parameter to import it into a multidimensional fatigue mapping function, calculating the L2 norm squared value and the multidimensional gradient feature parameter of the feature drift residual vector. The state evolution control module multiplies the L2 norm squared value by a preset first transformation coefficient constant to obtain a first weighted parameter, multiplies the multidimensional gradient feature parameter by a preset second transformation coefficient constant to obtain a second weighted parameter, and adds the first weighted parameter and the second weighted parameter to obtain the material feature degradation index.

9. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The state evolution control module retrieves and determines the topological directed edge connecting the current entity node and the matching static functional node from the set of directed edges in the global state topology network model, and extracts the state evolution weight parameter recorded at the initial binding time of the topological directed edge. The state evolution control module performs numerical accumulation operation on the state evolution weight parameter and the calculated material characteristic degradation index to obtain the dynamically updated weight value. The state evolution control module writes the dynamically updated weight value into the weight attribute field of the topological directed edge, and updates the current system time parameter into the observation timestamp attribute of the topological directed edge.

10. The power plant raw coal bunker equipment lifecycle traceability system based on KKS coding according to claim 1, characterized in that, The state evolution control module compares the dynamically updated weight values ​​with the preset failure critical weight threshold, and generates maintenance early warning state parameters based on the comparison results. The topology network construction module receives the equipment dismantling confirmation command, obtains the dismantling timestamp attribute, updates the Boolean value of the connection between the corresponding source entity node and the target functional node to zero in the adjacency matrix of the directed acyclic graph data structure to remove the topology directed edge connection, and solidifies the disconnected topology directed edge as a static historical snapshot edge. The topology network construction module reads the final dynamically updated weight value corresponding to the dismantling timestamp attribute, sets the final dynamically updated weight value as the current residual fatigue weight parameter of the source entity node, and writes it as a historical connection edge record parameter into the node attribute data structure of the source entity node.

Images