Enterprise digital factory equipment operation and maintenance management system and method

Abstract

The invention discloses an enterprise digital factory equipment operation and maintenance management system and method, and relates to the technical field of operation and maintenance management, and the method comprises the steps: collecting equipment operation state data, inputting the data into a fault analysis engine, extracting a feature frequency, a harmonic distortion rate and a gradient change value, matching a fault feature library in a maintenance knowledge graph, and generating a fault code and a diagnosis report; generating and executing a production department plan adjustment instruction, a maintenance department maintenance instruction and a purchase department spare part scheduling instruction based on the contribution weight of each department; collecting the executed equipment parameters through the AR operation terminal, extracting the equipment efficiency index, and generating a maintenance record; and generating token excitation according to the equipment efficiency index and the contribution weight of each department, and distributing and returning a maintenance record through a block chain smart contract to update the maintenance knowledge graph. According to the method, the contribution weight is distributed by adopting the Sharpley value algorithm, so that mathematical publication quantification of the cross-department collaboration value is realized.

Description

Technical Field

[0001] This invention relates to the field of operation and maintenance management technology, and in particular to an enterprise digital factory equipment operation and maintenance management system and method. Background Technology

[0002] Currently, fault diagnosis engines and knowledge graphs are widely used in the field of industrial equipment operation and maintenance. Based on feature extraction from vibration signals combined with a fault feature database, preliminary identification of typical fault modes can be achieved; maintenance knowledge graphs map fault features to maintenance solutions through semantic association, supporting partial automation of decision-making. In recent years, blockchain has been applied to the maintenance history storage process to ensure data immutability, while augmented reality (AR) tools are gradually being applied to data collection and operation guidance in the maintenance process, improving the level of digitalization in on-site operations.

[0003] Existing technologies have significant shortcomings in quantifying the value of multi-departmental collaboration and incentivizing dynamic incentives. The lack of a fair contribution allocation mechanism for the comprehensive benefits generated by maintenance strategies leads to insufficient motivation for departmental collaboration. Traditional weighted averaging cannot satisfy the three axioms of the Shapley value algorithm, causing knowledge graph updates to rely on subjective experience and making it difficult to form a closed loop of "maintenance-incentive-knowledge evolution". Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for the operation and maintenance management of equipment in a digital factory to solve the problems of fairness in the value quantification and allocation of collaborative maintenance across multiple departments and the construction of a dynamic knowledge iteration closed loop.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for equipment operation and maintenance management in a digital factory of an enterprise, which includes: collecting equipment operating status data and inputting it into a fault analysis engine; extracting characteristic frequencies, harmonic distortion rates and gradient change values; matching the fault feature library in the maintenance knowledge graph; and generating fault codes and diagnostic reports. Based on the fault code and diagnostic report, query the maintenance knowledge graph to find the associated maintenance solutions, and output the maintenance strategy in combination with the factory resource constraints; Based on the maintenance strategy, the value of production downtime reduction, maintenance cost reduction, and procurement response improvement are calculated, and the contribution weight of each department is allocated using the Shapley value algorithm. Based on the contribution weight of each department, generate and execute production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare parts dispatch instructions; The system collects equipment parameters after execution using an AR operating terminal and extracts equipment efficiency indicators to generate maintenance records. Token incentives are generated based on equipment efficiency indicators and the contribution weights of each department. These incentives are then distributed through blockchain smart contracts, and maintenance records are sent back to update the maintenance knowledge graph.

[0007] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps for generating fault codes and diagnostic reports are as follows: Based on the intensity of vibration signal changes in the equipment operation status data, the sampling signal is generated by dynamically switching between high and low sampling rates, and quantum tunneling probability denoising is performed on the sampling signal to output an optimized signal. Feature frequencies, harmonic distortion rates, and gradient change values ​​are extracted synchronously from the optimized signal and fused to generate a feature vector through a gated attention mechanism. Calculate the fractional Mahalanobis distance between the feature vector and the fault feature database node in the maintenance knowledge graph; If the fractional Mahalanobis distance exceeds the preset fault tolerance threshold, an adversarial sample is generated to re-trigger the quantum tunneling probability denoising process; otherwise, a fault code and diagnostic report are output.

[0008] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps of outputting maintenance strategies based on factory resource constraints are as follows: Based on the fault codes and diagnostic reports, retrieve the associated spare parts list, human resources and tool configurations from the maintenance knowledge graph to generate an initial maintenance plan; Based on real-time factory resource data, select feasible maintenance solutions that meet all constraints from the initial maintenance plans; Spatiotemporal conflict detection is performed on feasible maintenance solutions, and the timing of tasks and resource allocation are adjusted through dynamic priority scheduling to generate conflict-free maintenance strategies. Input the conflict-free maintenance strategy into the blockchain smart contract, verify the availability of resources and the dependency relationship of tasks, and output the maintenance strategy.

[0009] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps for allocating the contribution weights of each department using the Shapley value algorithm are as follows: Based on the maintenance strategy, the value of reduced production downtime, reduced maintenance costs, and improved procurement response are calculated and mapped to a three-dimensional quantum state amplitude vector. Perform quantum entanglement operations on the three-dimensional quantum state amplitude vector to generate a departmental synergy value tensor; Based on the value tensor of departmental synergy effect, under the dynamic constraints of Lorentz attractor, the tensor Shapley integral equation is solved to generate the initial solution of the contribution weight of each department. The initial solution of the contribution weights of each department is normalized, and the Shapley value algorithm is used to allocate the contribution weights of each department.

[0010] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps for generating and executing production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare parts scheduling instructions are as follows. Based on the contribution weights of each department and the location of the equipment, a spatiotemporal weighted topology rule is constructed. Based on the spatiotemporal weighted topology rules, quantum superposition command states are generated for production department plan adjustments, maintenance department repairs, and procurement department spare parts scheduling. The Pareto optimality of the quantum superposition instruction state is solved by quantum annealing, and an executable instruction set is output. The executable instruction set is pushed to the terminals of each department in real time according to the Industrial Internet of Things (IIoT) protocol and triggered for execution.

[0011] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps for generating maintenance records are as follows: The point cloud spatial coordinates, vibration spectrum, and infrared thermogram of the device were collected after execution via an AR operation terminal. The spatial coordinates of the point cloud are encoded as the real part, the vibration spectrum is encoded as the imaginary part, and the phase is modulated by the thermogram, and then fused to generate a quantum spatial field. Perform multidimensional tensor differential operations on the quantum space field to extract equipment efficiency indicators and generate maintenance records.

[0012] As a preferred embodiment of the enterprise digital factory equipment operation and maintenance management method described in this invention, the specific steps for updating the maintenance knowledge graph by transmitting maintenance records are as follows: Based on equipment efficiency indicators and the contribution weight of each department, the token incentive quantity for each department is dynamically calculated using the sigmoid activation function. Based on the number of token incentives for each department, adaptive compression is performed on the maintenance records to generate compressed records; Update the maintenance knowledge graph using compressed records and deep learning networks; The updated maintenance knowledge graph and the number of token incentives for each department will be stored and distributed through a dual-chain blockchain.

[0013] Secondly, the present invention provides an enterprise digital factory equipment operation and maintenance management system, including a fault diagnosis module, a strategy generation module, a contribution quantification module, an instruction coordination module, an execution feedback module, and a knowledge iteration module; The fault diagnosis module is used to collect equipment operating status data and input it into the fault analysis engine to extract characteristic frequencies, harmonic distortion rates and gradient change values, match them with the fault feature library in the maintenance knowledge graph, and generate fault codes and diagnostic reports. The strategy generation module is used to query the maintenance solutions associated with the maintenance knowledge graph based on the fault code and diagnostic report, and output the maintenance strategy in combination with the factory resource constraints. The contribution quantification module is used to calculate the value of production downtime reduction, maintenance cost reduction, and procurement response improvement based on the maintenance strategy, and uses the Shapley value algorithm to allocate the contribution weight of each department. The instruction coordination module is used to generate and execute production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare parts scheduling instructions based on the contribution weight of each department. The execution feedback module is used to collect equipment parameters after execution, generate equipment efficiency indicators, compare them with the standard operation trajectory of the maintenance knowledge graph, update the maintenance history and store the evidence. The knowledge iteration module is used to generate token incentives based on equipment efficiency indicators and contribution weights, and automatically allocates and transmits the updated maintenance knowledge graph through blockchain smart contracts.

[0014] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the enterprise digital factory equipment operation and maintenance management method as described in the first aspect of the present invention.

[0015] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the enterprise digital factory equipment operation and maintenance management method as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By employing the Shapley value algorithm to allocate contribution weights, the value of cross-departmental collaboration is mathematically axiomatically quantified. By strictly satisfying three axioms, the value reduction from production downtime, the value reduction from maintenance costs, and the value enhancement from procurement response are transformed into marginal contribution weights for each department. This accurately isolates the independent contribution of each department, and the output weight vector can directly drive instruction generation, ensuring fairness in incentives among departments. Simultaneously, the weight results are strongly coupled with blockchain token incentives, forming a positive feedback loop where contributions are quantifiable, incentives are traceable, and knowledge is evolvable, thereby improving the dynamic optimization efficiency of the maintenance knowledge graph. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart for the operation and maintenance management methods of equipment in a digital factory; Figure 2 A module diagram for an enterprise's digital factory equipment operation and maintenance management system; Figure 3 This is a flowchart for fault diagnosis; Figure 4 A flowchart for allocating contribution weights. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0022] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for the operation and maintenance management of equipment in an enterprise digital factory, including the following steps: S1. Collect equipment operating status data and input it into the fault analysis engine to extract characteristic frequencies, harmonic distortion rates and gradient change values, match them with the fault feature library in the maintenance knowledge graph, and generate fault codes and diagnostic reports.

[0023] Furthermore, based on the intensity of vibration signal changes in the equipment's operating status data, the sampling signal is dynamically generated by switching between high and low sampling rates, and quantum tunneling probability denoising is applied to the sampling signal to output an optimized signal. It should be noted that the equipment operating status data includes vibration signals, current waveforms, and temperature data.

[0024] Specifically, the Reynolds number is obtained based on the intensity of the vibration signal change. When the Reynolds number exceeds the preset turbulence threshold, a high sampling rate is used to generate the sampling signal. When the Reynolds number is lower than the preset turbulence threshold, a low sampling rate is used to generate the sampling signal. The quantum tunneling probability of the sampling signal is obtained, signal components with a quantum tunneling probability lower than the preset probability threshold are filtered out, and an optimized signal is output.

[0025] It should be noted that the preset turbulence threshold is based on the Reynolds number critical criterion for the transition from laminar to turbulent flow in fluid mechanics, with an example value of 2000; the preset probability threshold is based on the Heisenberg uncertainty principle and the probability distribution of the sensor's thermal noise background, with an example value of 0.005.

[0026] Feature frequencies, harmonic distortion rates, and gradient change values ​​are extracted synchronously from the optimized signal and fused to generate a feature vector through a gated attention mechanism. Specifically, the optimized signal is simultaneously input into the feature frequency extraction unit to obtain the main frequency component, the harmonic distortion rate extraction unit to obtain the total harmonic distortion rate, and the gradient change value extraction unit to obtain the signal change rate. The outputs of each unit form a three-dimensional feature data array. The gated attention mechanism uses a query vector to act on the three-dimensional feature data array to generate a weight vector. The weight vector includes feature frequency dimension weight, harmonic distortion rate dimension weight, and gradient change value dimension weight. The three-dimensional feature data array is weighted by dot product to generate weighted feature data, which is then integrated into a feature vector along the channel dimension.

[0027] Specifically, the fractional Mahalanobis distance between the feature vector and the fault feature database node is calculated in the maintenance knowledge graph, and the expression is: ; In the formula, Representing the eigenvector With fault feature library node vector The fractional Mahalanobis distance, Denotes the fractional order of the differential. Indicates the attention regulation index, This indicates that the summation is performed by iterating through and summing the eigencomponents of all eigenvectors. Indicates the index of the feature component of the feature vector. Representing the eigenvector The Each feature component The feature components representing the node vectors in the fault feature library, Represents the node vector of the fault feature library The Each feature component Representing the eigenvector The Fractional derivatives of each characteristic component Represents the node vector of the fault feature library The Fractional derivatives of each characteristic component The exponentially decaying function representing the probability of quantum tunneling. Indicates the sensitivity adjustment factor. The eigenvector represents the first eigenvector. The quantum tunneling probability of each characteristic component Represents the quantum tunneling probability. This represents the threshold for the quantum tunneling probability. This represents the Sigmoid gate function. Indicates the cross-item type index, This indicates that the summation is performed by iterating through all intersection item types. Indicates the type of cross item Attention weights The eigenvector represents the first eigenvector. Interaction term type of each feature component The cross value.

[0028] It should be noted that the sensitivity adjustment factor is optimized through quantum Monte Carlo simulation to maximize the noise suppression effect, and the example value is 3.5; the quantum tunneling probability threshold is determined based on the Heisenberg uncertainty principle and the statistical distribution of the equipment's thermal noise background value, and the example value is 0.005.

[0029] It should be noted that, For feature vectors The Each feature component For characteristic frequencies, Harmonic distortion rate, The gradient change value; For fault feature library node vectors The Each feature component For standard characteristic frequencies, Standard harmonic distortion rate, This represents the standard gradient change value.

[0030] If the fractional Mahalanobis distance exceeds the preset fault tolerance threshold, an adversarial sample is generated to re-trigger the quantum tunneling probability denoising process; otherwise, a fault code and diagnostic report are output.

[0031] It should be noted that the preset fault tolerance threshold is determined based on the statistical distribution standard deviation of the equipment vibration characteristics under normal operating conditions and the optimization analysis of historical fault false alarm rate. The example value is 0.15.

[0032] Specifically, when the fractional Mahalanobis distance exceeds the preset fault tolerance threshold, Gaussian perturbation is added to the vibration signal to generate adversarial samples, and the adversarial samples are used as new input to re-trigger the quantum tunneling probability denoising process. When the fractional Mahalanobis distance does not exceed the preset fault tolerance threshold, the fault code corresponding to the fault feature library node is retrieved and associated with historical diagnostic records to generate a fault code and diagnostic report.

[0033] S2. Based on the fault code and diagnostic report, query the maintenance knowledge graph for associated maintenance solutions, and output maintenance strategies in combination with factory resource constraints.

[0034] Furthermore, based on the fault codes and diagnostic reports, the system retrieves relevant spare parts lists, human resources, and tool configurations from the maintenance knowledge graph to generate an initial maintenance plan. Specifically, based on the fault code and diagnostic report, the corresponding fault feature library node is queried in the maintenance knowledge graph. The associated spare parts list node, human resources node, and tool configuration node are retrieved along the graph relationship edges. The spare parts list node stores the spare parts type and quantity information, the human resources node stores the skill level and working hour information, and the tool configuration node stores the equipment model and quantity information. The spare parts list information, human resources information, and tool configuration information are combined into an initial maintenance plan according to the maintenance task requirements.

[0035] Based on real-time factory resource data, select feasible maintenance solutions that meet all constraints from the initial maintenance plans; It should be noted that meeting all constraints means that the number of spare parts in the spare parts list is less than or equal to the spare parts inventory in the factory's real-time resource data, the man-hours are less than or equal to the upper limit of man-hours in the factory's real-time resource data, and the number of tools configured is less than or equal to the tool availability status in the factory's real-time resource data.

[0036] Specifically, based on the spare parts inventory, manpower hours limit, and tool availability status in the factory's real-time resource data, check whether the number of spare parts in the initial maintenance plan is less than or equal to the spare parts inventory, whether the manpower hours are less than or equal to the manpower hours limit, and whether the number of tools is less than or equal to the tool availability status. Only when the number of spare parts, manpower hours, and tool configurations all meet the constraints can the initial maintenance plan be retained as a feasible maintenance plan.

[0037] Spatiotemporal conflict detection is performed on feasible maintenance solutions, and the timing of tasks and resource allocation are adjusted through dynamic priority scheduling to generate conflict-free maintenance strategies. Specifically, spatiotemporal conflict detection is performed on feasible maintenance plans to identify combinations of overlapping task time windows and conflicting resource locations. Dynamic priority scores are obtained based on the urgency of maintenance tasks, resource cost coefficients, and contribution weights of each department. The task execution sequence is rearranged in descending order of priority scores. The starting point of the time window or the resource allocation position of conflicting tasks is iteratively adjusted until all task time windows are non-overlapping and the distance between resource locations is greater than a preset safety threshold, thus generating a conflict-free maintenance strategy.

[0038] It should be noted that the preset safety threshold is set based on the safety radius of equipment maintenance operations and the electromagnetic compatibility specifications of industrial sites; the example value is 10.

[0039] Input the conflict-free maintenance strategy into the blockchain smart contract, verify the availability of resources and the dependency relationship of tasks, and output the maintenance strategy.

[0040] Specifically, after the conflict-free maintenance strategy is input into the blockchain smart contract, the resource status database is called to verify whether the real-time value of the spare parts inventory is greater than or equal to the number of spare parts in the conflict-free maintenance strategy, whether the real-time available value of manpower hours is greater than or equal to the manpower hours in the conflict-free maintenance strategy, and whether the tool occupancy status meets the tool configuration quantity in the conflict-free maintenance strategy. At the same time, it checks whether the completion status of the preceding tasks in the task dependency graph is marked as completed. Only when all resources are available and the task dependencies are conflict-free, resource locking instructions and task distribution instructions are triggered, and the maintenance strategy is output.

[0041] S3. Calculate the value of production downtime reduction, maintenance cost reduction, and procurement response improvement based on the maintenance strategy, and assign contribution weights to each department using the Shapley value algorithm.

[0042] Based on the maintenance strategy, the value of reduced production downtime, reduced maintenance costs, and improved procurement response are calculated and mapped to a three-dimensional quantum state amplitude vector. Specifically, the square root of the value reduction due to production downtime is taken as the first real part amplitude value of the three-dimensional quantum state amplitude vector. The square root of the value reduction due to maintenance costs is combined with the maintenance efficiency phase factor as the second complex amplitude value. The square root of the value increase due to the procurement department's response is taken as the third real part amplitude value. The first real part amplitude value, the second complex part amplitude value, and the third real part amplitude value are combined to form the three-dimensional quantum state amplitude vector.

[0043] Based on the maintenance strategy, the value of reduced production downtime, reduced maintenance costs, and improved procurement response are calculated using the following expression: ; In the formula, This indicates that production shutdowns reduce value. This indicates that the repair department's costs have reduced its value. This indicates that the procurement department is responding to enhance value. Indicates the interval Integrating within, Indicates the strategy execution cycle. Indicates the time decay factor. This indicates a decrease in the rate of shutdown. Indicates the value decay coefficient. Indicates the current time. This indicates the cumulative reduction in downtime over time. Indicates the cost decay factor. Indicates the cost attenuation coefficient. Indicates the cost reduction rate. Indicates the baseline maintenance cost. Indicates the actual repair cost. This represents a function that maps response time differences to value enhancement. Represents the response sensitivity coefficient. Indicates the actual response time. Indicates the baseline response time. This represents the tensor product operator. Indicates the current time The production quantum state at that time, Indicates the current time Maintenance quantum state at time, Indicates the current time The quantum state of procurement at that time This represents a time infinitesimal element.

[0044] It should be noted that the time decay factor is derived from the exponential relaxation model of equipment operating status recovery, with an example value of 0.05; the value decay coefficient is calculated by normalizing the maximum Lyapunov exponent of Lorentz chaos, with an example value of 0.05; the cost decay factor is designed based on the economic depreciation model of maintenance resource consumption, with an example value of 0.1; the cost decay coefficient is derived from the intersection analysis of the fitted asset depreciation curve and the maintenance benefit curve, with an example value of 0.1; and the response sensitivity coefficient is obtained through the Sigmoid curvature analysis of the procurement response time-value conversion, with an example value of 10.

[0045] Perform quantum entanglement operations on the three-dimensional quantum state amplitude vector to generate a departmental synergy value tensor; Specifically, a quantum Hadamard gate transformation is applied to the three-dimensional quantum state amplitude vector to achieve amplitude superposition. Then, a CNOT quantum gate is applied to generate an entangled state between the production department and the maintenance department. A Tofoli quantum gate is added to obtain the phase control of the procurement department. The measurement results of the entangled state qubits are projected onto the Pauli operator basis, and the third-order tensor form of the Pauli operator eigenvalue correlation function is calculated to output the departmental synergy effect value tensor.

[0046] Specifically, based on the value tensor of departmental synergy effect, and under the dynamic constraints of the Lorentz attractor, the tensor Shapley integral equation is solved to generate an initial solution for the contribution weights of each department, expressed as: ; In the formula, Indicates department type Preliminary solution to contribution weights Indicates a department type index. This represents the reciprocal of the period of the Lorentz attractor integral. Indicates the interval Integrating within, Indicates the period of the Lorentz attractor integral. This indicates that there are no department types. Departmental subset Perform a traversal and summation. Indicates no department type Departmental subset, Indicates all department types. Indicates the type of department to be removed. The set, Indicates no department type Departmental subset factorial, This represents the factorial of the number of remaining department types. This represents the Shapley weight kernel. Indicates adding department type No department type Departmental subset At the current time value, Indicates the current time Lorentz state vector at time, Indicates no department type Departmental subset At the current time The value of time This represents the instantaneous gradient magnitude of the trajectory of the Lorentz state vector.

[0047] The initial solution of the contribution weights of each department is normalized, and the Shapley value algorithm is used to allocate the contribution weights of each department.

[0048] Specifically, the sum of the absolute values ​​of the initial solutions of the contribution weights of each department is calculated. The initial solutions of the contribution weights of each department are then divided by the sum of the absolute values ​​of the initial solutions of the contribution weights of each department to obtain the normalized weights. The names of the normalized weights maintain the correspondence between the department indexes, ensuring that the sum of the weight output values ​​of the production department, maintenance department, and purchasing department is 1. The normalized weights are then output as the contribution weights of each department assigned by the Shapley value algorithm.

[0049] S4. Generate and execute production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare parts dispatch instructions based on the contribution weight of each department.

[0050] Furthermore, based on the contribution weights of each department and the location of the equipment, a spatiotemporal weighted topology rule is constructed; Specifically, based on the contribution weights of each department, a weight field is generated by combining the spatial decay function of the equipment location. The spatial decay function adopts the negative exponential form of the equipment location coordinate distance. The decay function coefficient of the weight field of the production department is doubled, the decay function coefficient of the weight field of the maintenance department is the benchmark value, and the decay function coefficient of the weight field of the procurement department is halved, forming a spatiotemporal weighted topology rule.

[0051] Based on the spatiotemporal weighted topology rules, quantum superposition command states are generated for production department plan adjustments, maintenance department repairs, and procurement department spare parts scheduling. Specifically, the spatiotemporal weighted topology rules are input into the quantum amplitude distributor. The amplitude value of the production department's planning adjustment command is taken as the square root of the spatiotemporal weighted topology rule at the equipment location as the real part. The amplitude value of the maintenance department's repair command is taken as the square root of the spatiotemporal weighted topology rule at the fault location, combined with the maintenance priority phase factor as the imaginary part. The amplitude value of the procurement department's spare parts scheduling command is taken as the square root of the spatiotemporal weighted topology rule at the warehouse location as the real part amplitude. The amplitude values ​​of the three departments are combined to form a complex vector. Gram-Schmidt orthogonalization is performed on the complex vector to ensure the orthogonality of the command states. The normalized probability amplitude is applied using the Haar measure, and the quantum superposition command states of the production department's planning adjustment, the maintenance department's repair, and the procurement department's spare parts scheduling are output.

[0052] The Pareto optimality of the quantum superposition instruction state is solved by quantum annealing, and an executable instruction set is output. Specifically, the quantum superposition instruction states of production department plan adjustment, maintenance department overhaul, and procurement department spare parts scheduling are mapped to a Hamiltonian matrix of a QUBO problem. The objective function is defined as a multi-objective weighted sum of maximizing production planning efficiency, minimizing maintenance costs, and minimizing procurement response. The constraints are encoded as Lagrange penalty terms appended to the diagonal elements of the Hamiltonian. The quantum annealing machine performs Boltzmann evolution to search for the ground state solution. The binary vector of the ground state solution is decoded into a production department plan adjustment instruction switch flag, a maintenance department overhaul instruction priority flag, and a procurement department spare parts scheduling instruction trigger flag. The instruction items with set flags are combined to output an executable instruction set.

[0053] The executable instruction set is pushed to the terminals of each department in real time according to the Industrial Internet of Things (IIoT) protocol and triggered for execution.

[0054] Specifically, the executable instruction set is divided into production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare parts scheduling instructions according to department type. Each instruction is encoded as an OPC UA protocol data packet, an MQTT protocol message, and an HTTP-API request. It is written to the production execution system database in real time through the OPC UA server, published to the maintenance terminal message queue through the MQTT agent, and submits a scheduling request by calling the procurement RESTful interface. After receiving the request, the terminal device parses the instruction content and triggers the local execution program.

[0055] S5. Collect the equipment parameters after execution through the AR operation terminal and extract the equipment efficiency indicators to generate maintenance records.

[0056] Furthermore, the point cloud spatial coordinates, vibration spectrum, and infrared thermogram of the device are collected after execution via an AR operation terminal; Specifically, the AR operating terminal uses a depth camera to scan the surface of the device to generate point cloud spatial coordinates, simultaneously activates a vibration sensor to record time-domain vibration signals, outputs vibration spectrum through fast Fourier transform, and activates an infrared thermal imager to capture the temperature field distribution of the device and map it into an infrared thermal map.

[0057] The spatial coordinates of the point cloud are encoded as the real part, the vibration spectrum is encoded as the imaginary part, and the phase is modulated by the thermogram, and then fused to generate a quantum spatial field. Specifically, the device surface height value of the point cloud spatial coordinates is extracted as the real part, the main frequency amplitude value of the vibration spectrum is mapped as the imaginary part, the relative temperature difference of the infrared thermogram data is converted into the phase angle, the real part is assigned to the real part of the complex number, the imaginary part is assigned to the imaginary part of the complex number, the phase angle controls the complex exponential rotation factor, and the complex numbers at each point in space are merged into quantum state amplitude through the complex exponential function to form a quantum spatial field.

[0058] Perform multidimensional tensor differential operations on the quantum space field to extract equipment efficiency indicators and generate maintenance records.

[0059] Specifically, the quantum spatial field is input into a three-dimensional complex-valued tensor grid. The first-order partial derivatives of the real component are calculated along the equipment height direction, the second-order partial derivatives of the imaginary component are calculated along the vibration propagation direction, and the curl of the phase component is calculated along the thermal gradient direction. The Kronecker product is performed on the real derivative tensor, the imaginary derivative tensor, and the phase curl tensor. The tensor Frobenius norm is extracted as the energy efficiency ratio index, the variance of the imaginary derivative tensor is used as the stability index, and the phase curl divergence is used as the response rate index. The energy efficiency ratio index, stability index, and response rate index are combined into an equipment efficiency index. Maintenance records are generated by associating the equipment number, timestamp, and maintenance strategy code.

[0060] S6. Generate token incentives based on equipment efficiency indicators and the contribution weights of each department, distribute them through blockchain smart contracts, and update the maintenance knowledge graph by transmitting maintenance records back.

[0061] Specifically, based on equipment efficiency indicators and the contribution weights of each department, the token incentive quantity for each department is dynamically calculated using a sigmoid activation function. The expression is as follows: ; In the formula, Indicates department type The number of token incentives, This indicates the benchmark amount of tokens issued as incentives. This indicates the Sigmoid activation kernel function. Indicates department type The actual efficiency, Indicates the baseline efficiency. Indicates department type The ratio of actual efficiency to baseline efficiency. Represents the efficiency sensitivity coefficient. Represents the hyperbolic tangent function. Indicates department type The weighting is increased. Indicates the weighted index, Indicates department type The difference between the actual efficiency and the benchmark efficiency.

[0062] It should be noted that the efficiency sensitivity coefficient is fitted to the optimal stimulus response curve through Sigmoid function curvature analysis, with an example value of 8; the weight enhancement index is determined based on the principle of increasing marginal utility in behavioral economics and optimized through a Nash equilibrium game model, with an example value of 2.5.

[0063] Based on the number of token incentives for each department, adaptive compression is performed on the maintenance records to generate compressed records; Specifically, the compression strength coefficient is obtained based on the token incentive quantity of each department. The token incentive quantity value of the production department corresponds to the wavelet transform compression level, the token incentive quantity value of the maintenance department corresponds to the discrete cosine transform quantization step size, and the token incentive quantity value of the procurement department corresponds to the sparse coding dictionary size. The maintenance records are divided into three subsets: production equipment log, maintenance operation sequence, and procurement history. The production equipment log subset is compressed in the time domain using the Daubechies wavelet basis, the maintenance operation sequence subset is compressed in the spatial domain using the JPEG2000 standard, and the procurement history subset is compressed using K-SVD dictionary learning. The compressed subsets are then recombined into compressed records by adding department tag headers.

[0064] Update the maintenance knowledge graph using compressed records and deep learning networks; Specifically, the compressed record is decompressed and restored to a maintenance record. The maintenance record is parsed into an equipment efficiency index vector and a maintenance operation feature vector. The equipment efficiency index vector is updated into the equipment status node embedding vector in the maintenance knowledge graph through a graph convolutional network. The maintenance operation feature vector is used to predict the fault mode association weights through a long short-term memory network. The updated equipment status node embedding vector and fault mode association weights are written into the graph database storage layer of the maintenance knowledge graph, triggering the maintenance knowledge graph update.

[0065] The updated maintenance knowledge graph and the number of token incentives for each department will be stored and distributed through a dual-chain blockchain.

[0066] Specifically, the updated maintenance knowledge graph is stored as an IPFS hash value through the data blockchain. The token incentive amount for each department is issued through a smart contract triggered by the token blockchain. When storing the maintenance knowledge graph, the data blockchain records the timestamp and Merkle root hash. The token blockchain calls the ERC-20 contract to mint tokens to the wallet addresses of each department. The two blockchains ensure data consistency through a cross-chain atomic transaction protocol. After storage, structured evidence data containing the graph IPFS address, token transaction ID, and cross-chain proof is generated.

[0067] This embodiment also provides an enterprise digital factory equipment operation and maintenance management system, including: a fault diagnosis module, a strategy generation module, a contribution quantification module, an instruction coordination module, an execution feedback module, and a knowledge iteration module; the fault diagnosis module is used to collect equipment operating status data and input it into the fault analysis engine, extract characteristic frequencies, harmonic distortion rates, and gradient change values, match them with the fault feature library in the maintenance knowledge graph, and generate fault codes and diagnostic reports; the strategy generation module is used to query the maintenance solutions associated with the maintenance knowledge graph based on the fault codes and diagnostic reports, and output maintenance strategies in combination with factory resource constraints; the contribution quantification module is used to calculate production based on the maintenance strategies. The system calculates the value reduction from departmental downtime, cost reduction from maintenance, and value enhancement from procurement response, using a Shapley value algorithm to assign contribution weights to each department. The instruction coordination module generates and executes production plan adjustment instructions, maintenance instructions, and procurement spare parts dispatch instructions based on each department's contribution weight. The execution feedback module collects equipment parameters after execution to generate equipment efficiency indicators, compares them with standard operating procedures in the maintenance knowledge graph, updates maintenance records, and stores the data. The knowledge iteration module generates token incentives based on equipment efficiency indicators and contribution weights, automatically allocating and transmitting these incentives via blockchain smart contracts to update the maintenance knowledge graph.

[0068] This embodiment also provides a computer device applicable to the enterprise digital factory equipment operation and maintenance management method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the enterprise digital factory equipment operation and maintenance management method proposed in the above embodiment.

[0069] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0070] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the enterprise digital factory equipment operation and maintenance management method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0071] In summary, this invention achieves the mathematical axiomatic quantification of cross-departmental collaboration value by employing the Shapley value algorithm to allocate contribution weights. By strictly satisfying three axioms, the value reduced by production downtime, the value reduced by maintenance costs, and the value increased by procurement response are transformed into marginal contribution weights for each department. This accurately separates the independent contribution of each department, and the output weight vector can directly drive instruction generation, ensuring the fairness of incentives among departments. At the same time, the weight results are strongly coupled with blockchain token incentives, forming a positive feedback loop where contributions are quantifiable, incentives are traceable, and knowledge is evolvable, thereby improving the dynamic optimization efficiency of the maintenance knowledge graph.

[0072] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An enterprise digital plant equipment operation and maintenance management method, characterized in that: comprising, The acquisition device runs state data input fault analysis engine, extracts characteristic frequency, harmonic distortion rate and gradient change value, matches the fault feature library in the maintenance knowledge graph, generates fault code and diagnosis report; According to the fault code and the diagnosis report, the maintenance scheme associated with the maintenance knowledge graph is queried, and the maintenance strategy is output combined with the factory resource constraint condition; According to the maintenance strategy, the production department shutdown reduction value, the maintenance department cost reduction value and the procurement department response improvement value are calculated, and the contribution weight of each department is distributed by using the Shapley value algorithm; Based on the contribution weight of each department, the production department plan adjustment instruction, the maintenance department maintenance instruction and the procurement department spare part scheduling instruction are generated and executed; Through the AR operation terminal, the equipment parameters after execution are collected and the equipment efficiency index is extracted, and the maintenance record is generated; According to the equipment efficiency index and the contribution weight of each department, a token incentive is generated, and the maintenance record is distributed and returned through a blockchain smart contract to update the maintenance knowledge graph.

2. The enterprise digital plant equipment operation and maintenance management method according to claim 1, characterized by: The specific steps of generating the fault code and the diagnosis report are as follows, According to the vibration signal change intensity in the equipment running state data, the high and low sampling rates are dynamically switched to generate the sampling signal, the quantum tunneling probability noise reduction is performed on the sampling signal, and the optimized signal is output; The characteristic frequency, harmonic distortion rate and gradient change value are synchronously extracted from the optimized signal, and the feature vector is generated by fusion through the gated attention mechanism; The fractional order Mahalanobis distance between the feature vector and the fault feature library node in the maintenance knowledge graph is calculated; If the fractional order Mahalanobis distance exceeds the preset fault tolerance threshold, an adversarial sample is generated, the quantum tunneling probability noise reduction process is retriggered, otherwise the fault code and the diagnosis report are output.

3. The enterprise digital plant equipment operation and maintenance management method according to claim 2, characterized by: The specific steps of outputting the maintenance strategy combined with the factory resource constraint condition are as follows, According to the fault code and the diagnosis report, the associated spare parts list, human resources and tool configuration in the maintenance knowledge graph are searched, and an initial maintenance scheme is generated; According to the factory real-time resource data, the feasible maintenance scheme that meets all the constraint conditions is selected from the initial maintenance scheme; The feasible maintenance scheme is subjected to time and space conflict detection, the task timing and resource allocation are adjusted through dynamic priority scheduling, and a conflict-free maintenance strategy is generated; The conflict-free maintenance strategy is input into the blockchain smart contract, the resource callability and task dependency relationship are verified, and the maintenance strategy is output.

4. The enterprise digital plant equipment operation and maintenance management method according to claim 3, characterized by: The specific steps of distributing the contribution weight of each department by using the Shapley value algorithm are as follows, According to the maintenance strategy, the production department shutdown reduction value, the maintenance department cost reduction value and the procurement department response improvement value are calculated, and the three-dimensional quantum state amplitude vector is mapped; The three-dimensional quantum state amplitude vector is subjected to quantum entanglement operation to generate the department synergy effect value tensor; Based on the department synergy effect value tensor, the tensor Shapley integral equation is solved under the constraint of the Lorenz attractor dynamics to generate the initial solution of the contribution weight of each department; The initial solution of the contribution weight of each department is subjected to normalization processing, and the contribution weight of each department is output by using the Shapley value algorithm.

5. The enterprise digital plant equipment operation and maintenance management method according to claim 4, characterized by: The specific steps of generating the production department plan adjustment instruction, the maintenance department maintenance instruction and the procurement department spare part scheduling instruction and executing them are as follows, Based on the contribution weight of each department, the time and space weighted topology rules are constructed combined with the equipment position; According to the space-time weighted topological rule, a quantum superposition instruction state of production department plan adjustment, maintenance department repair, and procurement department spare part scheduling is generated; A Pareto optimal is solved for the quantum superposition instruction state through quantum annealing, and an executable instruction set is output; The executable instruction set is pushed to each department terminal in real time according to an industrial Internet of Things protocol and execution is triggered.

6. The enterprise digital plant equipment operation and maintenance management method according to claim 5, characterized by: The maintenance record is generated, and the specific steps are as follows, Point cloud space coordinates, vibration spectrum, and infrared thermal map of the equipment after execution are collected through an AR operation terminal; The point cloud space coordinates are encoded as real parts, the vibration spectrum is encoded as imaginary parts, and the thermal map modulation phase is fused to generate a quantum space field; A multi-dimensional tensor differential operation is performed on the quantum space field, an equipment efficiency index is extracted, and a maintenance record is generated.

7. The enterprise digital plant equipment operation and maintenance management method according to claim 6, characterized by: The maintenance record is fed back to update the maintenance knowledge graph, and the specific steps are as follows, According to the equipment efficiency index and the contribution weight of each department, the number of department token incentives is dynamically calculated through a sigmoid activation function; Based on the number of department token incentives, a compressed record is generated by performing adaptive compression on the maintenance record; The compressed record is used to update the maintenance knowledge graph through a deep learning network; The updated maintenance knowledge graph and the number of department token incentives are stored and distributed through a double-chain blockchain.

8. An enterprise digital plant equipment operation and maintenance management system based on the enterprise digital plant equipment operation and maintenance management method of any one of claims 1-7, characterized in that: The system comprises a fault diagnosis module, a strategy generation module, a contribution quantification module, an instruction coordination module, an execution feedback module, and a knowledge iteration module; The fault diagnosis module is configured to collect equipment operation state data, input the data into a fault analysis engine, extract characteristic frequencies, harmonic distortion rates, and gradient change values, match fault characteristics in a maintenance knowledge graph, generate fault codes and diagnosis reports, and output the fault codes and diagnosis reports. The strategy generation module is configured to query a maintenance scheme associated with the maintenance knowledge graph according to the fault codes and diagnosis reports, and output a maintenance strategy in combination with factory resource constraint conditions. The contribution quantification module is configured to calculate production department downtime reduction value, maintenance department cost reduction value, and procurement department response improvement value according to the maintenance strategy, and distribute contribution weights of each department by using a Shapley value algorithm. The instruction coordination module is configured to generate production department plan adjustment instructions, maintenance department repair instructions, and procurement department spare part scheduling instructions based on the contribution weights of each department, and execute the instructions. The execution feedback module is configured to collect equipment parameters after execution to generate equipment efficiency indexes, compare the equipment efficiency indexes with standard operation trajectories of the maintenance knowledge graph, update maintenance records, and store the maintenance records. The knowledge iteration module is configured to generate token incentives according to the equipment efficiency indexes and the contribution weights, automatically distribute the token incentives through a blockchain smart contract, and update the maintenance knowledge graph with the equipment efficiency indexes. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to implement the steps of the enterprise digital factory equipment operation and maintenance management method of any one of claims 1-7.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the enterprise digital factory equipment operation and maintenance management method of any one of claims 1-7.

Images