A nuclear power DCS signal distribution method, device, equipment, medium and product

Abstract

The present application discloses a nuclear power DCS signal distribution method, device, equipment, medium and product, which relates to the field of industrial automation technology. The method includes: obtaining information data; the information data includes input data obtained by the input layer; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data; based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results; based on the decision layer, a resource allocation plan is determined according to the processing results; the resource allocation plan is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system. The present application aims to realize nuclear power DCS signal distribution and improve system reliability and safety.

Description

Technical Field

[0001] The present application relates to the field of industrial automation technology, and in particular to a nuclear power DCS signal distribution method, device, equipment, medium and product. Background Art

[0002] As the core control platform of nuclear power plants, the input and output signal distribution of nuclear power plant digital control system (DCS) directly affects the performance, reliability and maintainability of the system. With the development of nuclear power technology, the input and output signal distribution technology of DCS system has gone through the following development stages:

[0003] 1. Manual allocation stage (before the 1990s). The main characteristics of this stage are: signal allocation completely depends on the experience of engineers, lack of standardized allocation rules, poor system scalability and high maintenance costs.

[0004] 2. Rule-oriented stage (1990s to early 21st century). The main features of this stage are: establishing a basic allocation rule system and developing simple automated allocation tools, but using fixed allocation strategies and priorities and lacking flexibility.

[0005] 3. Intelligent optimization stage (early 21st century to present): the main features of this stage are: introduction of artificial intelligence technology, support for dynamic optimization and adaptive adjustment, but there are still many problems.

[0006] The main problems currently exist include:

[0007] 1. Specific problems in engineering practice.

[0008] (1) Insufficient signal correlation expression: Taking the main pump bearing monitoring system of a nuclear power plant as an example, the temperature measurement points TI101-TI103 are scattered on different boards, resulting in asynchronous data collection, reduced accuracy of status assessment, and a 30% extension of fault diagnosis time.

[0009] (2) Unbalanced resource allocation: In the secondary circuit control system of a nuclear power plant, the load rate of the AI03 board reached 92% while that of the AI08 board was only 45%, resulting in a system performance bottleneck. On average, 3-4 system abnormalities due to uneven load occurred each year.

[0010] (3) Difficulty in reusing knowledge: Experience rules are solidified in the form of code, which has high modification costs and lacks a formal knowledge expression mechanism, making it difficult to reuse optimization experience between different projects.

[0011] 2. Limitations at the methodological level.

[0012] (1) There is a lack of systematic analysis of the multi-dimensional correlation between signals.

[0013] (2) No unified knowledge expression and reuse framework has been established.

[0014] (3) The optimization strategy is too static and difficult to adapt to dynamic changes in the system.

[0015] 3. Limitations of technical implementation.

[0016] (1) Signal allocation relies too much on preset rules.

[0017] (2) Lack of real-time performance feedback and dynamic optimization mechanisms.

[0018] (3) The system lacks scalability and maintainability.

[0019] 4. Limitations of engineering applications.

[0020] (1) The configuration efficiency is low, and the average project cycle is extended by 20%.

[0021] (2) The operating performance is poor and the system response delay exceeds the standard by 35%.

[0022] (3) The maintenance cost is high, and the average annual maintenance hours increase by 40%.

[0023] Therefore, it is crucial to realize nuclear power DCS signal distribution to improve system reliability and safety. Summary of the invention

[0024] The purpose of this application is to provide a nuclear power DCS signal distribution method, device, equipment, medium and product, which can realize nuclear power DCS signal distribution and improve system reliability and safety.

[0025] To achieve the above objectives, this application provides the following solutions:

[0026] In a first aspect, the present application provides a nuclear power DCS signal distribution method, which is implemented by a nuclear power DCS signal distribution system based on a three-layer architecture design; the nuclear power DCS signal distribution system comprises: an input layer, a core processing layer and a decision layer connected in sequence; wherein the core processing layer is composed of an affinity calculation module, a knowledge graph module and an intelligent allocation module; the affinity calculation module, the knowledge graph module and the intelligent allocation module are all connected to the input layer; the affinity calculation module and the knowledge graph module are also connected to the intelligent allocation module;

[0027] The nuclear power DCS signal distribution method comprises:

[0028] Acquire information data; the information data includes input data acquired by the input layer; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data;

[0029] Based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results;

[0030] Based on the decision layer, a resource allocation scheme is determined according to the processing result; the resource allocation scheme is used to characterize the mapping relationship between signals and resources in a nuclear power DCS system.

[0031] In a second aspect, the present application provides a nuclear power DCS signal distribution device, comprising:

[0032] An information data acquisition module, used to acquire information data; the information data includes acquired input data; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data;

[0033] A processing module, used to perform data association interaction and constraint optimization processing according to the information data to obtain a processing result;

[0034] The decision module is used to determine a resource allocation scheme according to the processing result; the resource allocation scheme is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system.

[0035] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned nuclear power DCS signal distribution method.

[0036] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned nuclear power DCS signal distribution method when executed by a processor.

[0037] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the above-mentioned nuclear power DCS signal distribution method when executed by a processor.

[0038] According to the specific embodiments provided in this application, this application has the following technical effects:

[0039] The present application provides a nuclear power DCS signal distribution method, device, equipment, medium and product, which relates to the field of industrial automation technology. Information data is obtained, and based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results; based on the decision layer, a resource allocation plan is determined according to the processing results; the resource allocation plan is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system. The present application realizes nuclear power DCS signal distribution through the cooperation of the input layer, the core processing layer and the decision layer. Due to the existence of the core processing layer, data association interaction and constraint optimization processing are performed, thereby improving system reliability and safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 It is a flow chart of the nuclear power DCS signal distribution method;

[0042] Figure 2 This is the overall system architecture diagram of the nuclear power DCS signal distribution system;

[0043] Figure 3 Flow chart for affinity calculation;

[0044] Figure 4 It is a knowledge graph structure diagram;

[0045] Figure 5 Execute the flow chart for the hierarchical optimization algorithm;

[0046] Figure 6 This is a physical deployment diagram of Example 1;

[0047] Figure 7 This is a functional structure diagram of Embodiment 2;

[0048] Figure 8 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0051] At present, there are three main technical routes for DCS signal distribution technology in nuclear power plants:

[0052] 1. Rule-based allocation method. Taking patent CN102393778B as an example, a hierarchical rule structure is used for signal classification, and channel mapping is performed based on system coding. In actual applications, problems such as rigid rule structure and lack of coordination mechanism are exposed. For example, in the main pump bearing monitoring system of a nuclear power plant, the relevant signals are dispersed, resulting in asynchronous sampling, and the accuracy of state assessment is only 72%.

[0053] 2. Template-based allocation technology. Taking patent CN104732716A as an example, standard templates and parameterized configuration interfaces are used. In actual applications, there are problems such as insufficient template flexibility and low knowledge reuse efficiency. For example, in the transformation of the secondary water supply system of a nuclear power plant, the cascade control performance is unstable and the regulation accuracy is reduced by 25%.

[0054] 3. Optimization method based on artificial intelligence. Taking patent CN109902591A as an example, deep learning and genetic algorithm are used for optimization. There are problems such as large data demand, poor interpretability, and difficulty in modeling nuclear safety constraints.

[0055] In an exemplary embodiment, Figure 1 As shown, a nuclear power DCS signal distribution method is provided. The method is executed by a computer device, specifically, it can be executed by a computer device such as a terminal or a server alone, or it can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to a server as an example for explanation, and includes the following steps.

[0056] The nuclear power DCS signal distribution method is implemented by a nuclear power DCS signal distribution system based on a three-layer architecture design; the nuclear power DCS signal distribution system includes: an input layer, a core processing layer and a decision layer connected in sequence; wherein the core processing layer is composed of an affinity calculation module, a knowledge graph module and an intelligent allocation module; the affinity calculation module, the knowledge graph module and the intelligent allocation module are all connected to the input layer; the affinity calculation module and the knowledge graph module are also connected to the intelligent allocation module.

[0057] like Figure 1 As shown, the nuclear power DCS signal distribution method includes:

[0058] Step 100: Acquire information data. The information data includes input data acquired by the input layer; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data.

[0059] Step 200: Based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results.

[0060] Step 300: Based on the decision layer, determine a resource allocation scheme according to the processing result. The resource allocation scheme is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system.

[0061] In one embodiment, based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results, which specifically include:

[0062] The signal affinity quantification model is based on the affinity calculation module to determine the signal affinity according to the signal configuration information data; the signal affinity is used to characterize the degree of correlation between the signals in the nuclear power DCS system; the signal affinity quantification model is a mathematical model determined based on the four-dimensional system according to the historical operation data and the principal component analysis method.

[0063] Based on the knowledge graph module, the constructed knowledge graph is used to determine the optimization strategy according to signal affinity, DCS system configuration data and historical operation data. The optimization strategy is based on the set priority search strategy, with node affinity, resource matching and path length as search factors, and the analysis strategy corresponding to the configuration path found under the set constraints; the knowledge graph is a physical simulation model of node-relationship edges constructed under the constraint relationship constraints based on the parallel processing mechanism and cache reuse strategy. Among them, the constraint relationship constraints include: constraint identification, constraint type, constraint expression, priority and mandatory flag.

[0064] Based on the intelligent allocation module, the information data is clustered according to the signal affinity, and the resources are allocated to the clustered data based on the optimization strategy to obtain the processing result.

[0065] The four-dimensional system includes: physical location dimension, functional association dimension, security level dimension and power supply characteristic dimension; the expression of the signal affinity quantification model is:

[0066] .

[0067] in: For signal With signal The signal affinity between For the The weight coefficient of the dimension; For the Dimensional signal With signal The distance function of .

[0068] As an optional implementation method, based on the intelligent allocation module, the information data is subjected to signal clustering processing according to the signal affinity, and the clustered data is subjected to resource allocation based on the optimization strategy to obtain the processing result, which specifically includes:

[0069] Construct a signal unassigned set; the signal unassigned set contains all the information data to be assigned in the information data.

[0070] Based on the signal unassigned set, the importance value is calculated for each information data according to the function type and security level.

[0071] The information data with the highest importance value is selected as the seed signal in the signal unassigned set; with the seed signal as the center, the signal is absorbed according to the signal affinity to perform signal clustering processing to obtain clustered data; the clustered data is optimized and adjusted to obtain clustered data; resources are allocated to the clustered data based on the optimization strategy to obtain processing results.

[0072] In one embodiment, optimizing and adjusting the cluster data to obtain clustered data specifically includes:

[0073] The cluster size of the cluster data is checked to determine whether the cluster size is within a limited range, and a judgment result is obtained; if the judgment result is yes, the compatibility of the intra-cluster signal of the cluster data is verified, and the performance index is determined, and the cluster data is optimized based on the performance index to obtain the clustered data; the performance index includes: compactness; compactness is determined by calculating the average affinity of the intra-cluster signal. If the judgment result is no, the cluster data is split or merged.

[0074] The application scenarios of this application may include:

[0075] 1. Automatic distribution of input and output signals of control cabinets during the DCS engineering design phase of nuclear power plants.

[0076] 2. Optimal configuration of input and output resources during DCS system upgrade and transformation.

[0077] 3. Configuration optimization during regular system maintenance.

[0078] like Figure 2As shown, the method mentioned in this application adopts a three-layer architecture design. The input layer is responsible for receiving and preprocessing various input data, including signal configuration information data (KKS coding, technical parameters, physical characteristics, etc.), hardware resource status data (board, cage, cabinet configuration), DCS system configuration data (control strategy, constraints) and historical operation data; the core processing layer is composed of an affinity calculation module, a knowledge graph module and an intelligent allocation module. The three modules form a closed-loop optimization mechanism through data interaction. The affinity calculation module provides associated data for the knowledge graph, and the knowledge graph module provides constraints for intelligent allocation. The intelligent allocation module adjusts the affinity weight according to the optimization result feedback; the decision-making layer outputs resource allocation plans, performance evaluation reports and configuration optimization suggestions based on the calculation results of the core processing layer to provide guidance for project implementation.

[0079] In one embodiment, the input layer: input and output signal information (signal configuration information data) includes: technical parameters, physical characteristics, safety requirements, etc.; hardware resource information includes: configuration and status of boards, cages, and cabinets; DCS system configuration information includes: control strategies, system constraints, and operating parameters. Historical configuration data: operation records, performance data, and optimization experience.

[0080] Core processing layer: For the affinity calculation module, its function is to calculate the correlation between signals and maintain the weight system. The input is the signal feature data (signal configuration information data) from the input layer, and the output is the affinity matrix in the standardized [-1,1] interval; the affinity calculation module can provide association strength for the knowledge graph and provide clustering basis for intelligent allocation.

[0081] The knowledge graph module is used to manage system knowledge and provide decision support. The input is the affinity calculation results, system configuration information, and historical data, and the output is constraints, optimization suggestions, and reasoning results. The knowledge graph module receives affinity data to build a graph and provides constraints and optimization strategies for intelligent allocation.

[0082] For the intelligent allocation module, the function is to perform signal clustering and resource allocation. Its input is the affinity matrix, constraints and suggestions of the knowledge graph; the output is the allocation plan, performance evaluation, and optimization suggestions. Clustering can be performed based on affinity, and optimization can be performed based on the knowledge graph.

[0083] The resource allocation plan of the decision-making layer is a specific signal-resource mapping relationship; performance evaluation report: analysis of system performance, resource utilization, etc.; configuration optimization suggestions: short-term adjustments, long-term planning, etc.

[0084] Input layer → core processing layer corresponding signal data → affinity calculation: provide the original data required for calculation. System configuration → knowledge graph: provide the basis for knowledge construction; historical data → intelligent allocation: provide reference for optimization.

[0085] Within the core processing layer, affinity calculation → knowledge graph: provides association strength data; knowledge graph → intelligent allocation: provides constraints and optimization strategies; intelligent allocation → affinity calculation: provides weight adjustment feedback.

[0086] Core processing layer → decision-making layer: generation and output of allocation plans, performance evaluation and optimization suggestions.

[0087] The signal affinity quantification model defines a standardized affinity index in the interval [-1,1], where [0.7, 1.0]: highly correlated signals, such as the temperature-vibration-speed signal group in the main pump bearing monitoring system. [0.3, 0.7): moderately correlated signals, such as the general monitoring signal group in the water supply system. [0, 0.3): low-correlated signals, such as independent signals between different systems. [-1.0, 0): isolation-required signals, such as Class 1E and non-Class 1E equipment signals.

[0088] The four dimensions of the four-dimensional evaluation system include: physical location dimension (w1=0.25): on-site box distance, wiring path, and regional distribution; functional association dimension (w2=0.35): control loop, monitoring relationship, and data flow; safety level dimension (w3=0.25): equipment level, redundancy requirements, and isolation level; power supply characteristics dimension (w4=0.15): power supply type, load characteristics, and power supply partitions.

[0089] Adaptive weight adjustment based on historical data: Based on two years of operating data of three nuclear power plants (576 signal points), the principal component analysis method is used to determine the initial weights, achieving dynamic adjustment within the range of ±10%.

[0090] A three-layer entity model is designed for affinity and knowledge graph structure. Signal layer: including technical parameters, physical properties, safety features, etc. Resource layer: three-level resource structure of board, cage, and cabinet; Control layer: loop characteristics, performance indicators, constraints, etc.

[0091] Four types of relationship mappings are implemented: affinity relationship: the strength of correlation between signals; allocation relationship: signal-resource binding relationship; constraint relationship: hard constraints and soft constraints; performance relationship: system operation indicator correlation.

[0092] Regarding the intelligent allocation algorithm system in the intelligent allocation module, the input of its clustering algorithm features: signal set S, knowledge graph G, parameter set P; output: signal cluster set C; complexity: time O(n×logn), space O(n); parallelization: support for multi-threaded processing.

[0093] Its resource allocation algorithm features multi-objective optimization: performance, reliability, maintainability; constraint processing: security isolation, resource capacity, performance indicators; load balancing: ±5% dynamic balance control; real-time adjustment: support dynamic optimization in a 5-minute cycle.

[0094] This application can solve signal correlation identification problems, resource balancing problems and configuration efficiency problems.

[0095] Regarding the signal correlation identification problem, the problem is that in the main pump bearing monitoring system, the temperature-vibration-speed signals are dispersed, resulting in a sampling delay of 12ms and a state assessment accuracy of only 72%. This application realizes the intelligent identification and centralized configuration of correlation signals, controls the sampling delay within 3ms, and improves the state assessment accuracy to more than 90%.

[0096] Regarding the resource balancing problem, the problem is as follows: the maximum load of the AI ​​board is 92%, the minimum load is 45%, and the difference is 47%; the power supply load imbalance is 50%. This application controls the board load within the range of 75%±5%, and the power supply load imbalance is reduced to less than 15%.

[0097] Regarding the configuration efficiency issue, the problems are as follows: the manual configuration error rate is 15%, the average project cycle is 8 weeks, and the maintenance hours are 280 man-days / year; this application will reduce the configuration error rate to less than 5%, shorten the project cycle to 5 weeks, and reduce the maintenance hours to 150 man-days / year.

[0098] Figure 3 This is the affinity calculation flow chart. The affinity calculation function, that is, the expression of the signal affinity quantification model is:

[0099] .

[0100] in, For signal With signal The signal affinity between For the The weight coefficient of the dimension; For the Dimensional signal With signal The distance function of , representing four evaluation dimensions.

[0101] (2) Constraint C1: ; C2: 0;C3: ; C4: .

[0102] About distance function: The expression of physical location distance D1 is as follows.

[0103] .

[0104] in, is the in-place box position distance function; is the wiring path overlap function; is the control area correlation function; , and All are coefficients; + + =1, >0. .

[0105] Functional association distance D2: .

[0106] in, is the control loop correlation function; It is a data flow dependent function; is the timing requirement function; + + = 1, >0. . , and All are coefficients.

[0107] Safety level distance D3: .

[0108] in, is the security level matching function; To isolate the required function; is the redundant relation function; + + = 1, >0. . , and All are coefficients.

[0109] Power supply characteristic distance D4: .

[0110] in, is the power type matching function; is the voltage level matching function; is the load characteristic function; + + = 1, >0. . , and All are coefficients.

[0111] Regarding the method for determining the weight coefficient, two years of operating data from three nuclear power plants were collected as the basis for weight calculation. The data scale included 576 signal points and 121 control loops; the time span was two years (17,520 hours) of continuous operating data.

[0112] Preprocessing of collected data: elimination of invalid data (signal-to-noise ratio < 20dB); normalization (0-1 standardization); outlier processing (3σ principle); missing value supplementation (linear interpolation).

[0113] The analytic hierarchy process (AHP) is used to determine the initial weights. First, the judgment matrix A is established, which is constructed based on expert evaluation and historical data. The data corresponding to the judgment matrix are shown in Table 1.

[0114] Table 1 Judgment matrix data table

[0115]

[0116] Calculate the maximum eigenvalue λmax=4.0486 of matrix A, and the corresponding eigenvector is:

[0117] V=[0.2428, 0.4524, 0.2428, 0.1310].

[0118] A consistency check is performed based on the consistency index CI = (λmax-n) / (n-1) = 0.0162 and the consistency ratio CR = CI / RI = 0.0182 < 0.1 to meet the consistency requirements.

[0119] To verify the principal component analysis, correlation analysis was first performed.

[0120] The correlation coefficient matrix R of the four dimensions is calculated, and the data of the correlation coefficient matrix is ​​shown in Table 2.

[0121] Table 2 Correlation coefficient matrix data table

[0122]

[0123] Then the eigenvalue decomposition was performed, λ1=2.3847 (explanation rate 42.36%); λ2=1.4526 (explanation rate 25.78%); λ3=0.9847 (explanation rate 17.48%); λ4=0.8124 (explanation rate 14.38%).

[0124] Based on the eigenvalue, the principal component contribution rate W'=[0.2612, 0.3452, 0.2518, 0.1418] is solved.

[0125] Comprehensive determination of weights: First, a fusion calculation is performed to weightedly fuse the AHP results and PCA results W=α×V+(1-α)×W'; where α=0.6 (based on historical data verification), and the final weight values ​​are obtained by substitution: w1=0.25 (physical location dimension); w2=0.35 (functional association dimension); w3=0.25 (security level dimension); w4=0.15 (power supply characteristic dimension).

[0126] Verification indicators, consistency ratio: 0.0182<0.1; cumulative variance contribution rate: 85.62%; goodness of fit R²: 0.92.

[0127] Dynamic adjustment mechanism, whose adjustment trigger conditions are: performance deviation > 20%; load imbalance > 30%; failure rate increase > 15%. Adjustment parameters: adjustment step: ±2%; adjustment cycle: 24 hours; cumulative limit: ±10%. Adjustment range: w1: 0.225~0.275; w2: 0.315~0.385; w3: 0.225~0.275; w4: 0.135~0.165.

[0128] The knowledge graph structure is as follows Figure 4 The input data interface includes the DCS configuration data interface: system configuration, hardware information, network topology reading; real-time operation data interface: performance indicators, load status, alarm information collection; historical library data interface: trend records, event logs, maintenance record access.

[0129] Preprocess the input data, including: data standardization: unit conversion, precision unification, time alignment; validity verification: data integrity, value range, time series continuity; exception handling: missing value supplementation, outlier filtering, data correction.

[0130] Signal entities contain five basic attributes: identification attributes, technical attributes, physical attributes, safety attributes and power supply attributes. Among them, identification attributes include globally unique identifiers, KKS codes and system codes; technical attributes include signal type, range, sampling frequency and accuracy requirements; physical attributes include local box number, cable routing and control area; safety attributes include safety level, redundant grouping and isolation requirements; power supply attributes include power supply type, voltage level and power consumption indicators.

[0131] Resource entities adopt a three-level structure: board level, cage level and cabinet level. Board-level attributes include board identification, model, number of channels, number of used channels, power supply capacity and communication bandwidth; cage-level attributes include cage identification, number of slots, power module configuration and backplane bandwidth; cabinet-level attributes include cabinet identification, installation location, environmental requirements and power distribution plan.

[0132] The control loop entity describes the complete control function unit, including loop identification, control type (such as PID, cascade or selection control), component signal list, performance requirement indicators and control parameter configuration.

[0133] The affinity relationship describes the degree of association between signals, including the source node identifier, target node identifier, comprehensive affinity weight, scores of each dimension, and timestamp information. For example, in the main pump bearing monitoring system, the affinity relationship between the temperature signal TI101 and the vibration signal VI101 records their degree of association in four dimensions: physical location, functional association, safety level, and power supply characteristics.

[0134] The allocation relationship describes the mapping relationship between signals and hardware resources, including signal identifier, resource identifier, specific channel number, allocation time, performance indicators, etc. For example, the temperature signal TI101 is allocated to the second channel of the AI03 board, and the allocation time, sampling performance, and communication performance are recorded.

[0135] Constraint relationships describe various types of restrictions in the system, including constraint identifiers, constraint types, constraint expressions, priorities, and mandatory flags. For example, the isolation constraint between Class 1E and non-Class 1E device signals specifies the allocation restrictions for signals of different security levels.

[0136] Node generation adopts a four-step method: first, data normalization is performed, including unit conversion, parameter verification and data completion; second, attribute extraction is performed to parse identification information, technical parameters and configuration data; then node construction is performed to generate a unique identifier and set node attributes; finally, an integrity check is performed to verify necessary attributes, data types and value ranges.

[0137] To improve construction efficiency, a parallel processing mechanism is adopted to process multiple independent nodes at the same time; a cache reuse strategy is used to cache repeated attribute values; batch write optimization is implemented to reduce the number of database accesses; and the frequency of index updates is controlled to reduce system overhead.

[0138] The edge generation process includes four steps: relationship verification, attribute calculation, edge construction, and consistency check. Relationship verification ensures the legitimacy of the connection node type; attribute calculation determines the edge weight and attribute value; edge construction establishes the actual connection relationship; and consistency check verifies the bidirectional relationship and constraint satisfaction.

[0139] Adopt an incremental update strategy to only update the relationship edges that have changed; implement batch update optimization to merge related edge update operations; ensure transactional guarantees and maintain data consistency; establish a version control mechanism to support update rollback.

[0140] Regarding the reasoning mechanism of the knowledge graph, the first is the module interaction interface, which interacts with the affinity calculation module. Input interface: receiving the affinity matrix calculated in real time; data synchronization: affinity update triggers graph update; feedback mechanism: graph constraints affect affinity calculation.

[0141] Its output for the intelligent allocation module includes allocation constraints: output hard constraints and soft constraints; reasoning results: provide optimal allocation path suggestions; optimization strategy: give dynamic adjustment suggestions.

[0142] Path reasoning algorithm: The optimal path search adopts a priority search strategy to find the optimal configuration path under the constraints. The search process includes three stages: initialization, iterative search, and path generation. The iterative search stage is based on a weighted path scoring model, which comprehensively considers three factors: node affinity, resource matching, and path length.

[0143] The path weight calculation adopts a weighted sum model, which comprehensively considers the affinity between each node pair on the path, the resource matching degree of the node, and the complexity of the path. Each weight coefficient is determined by analyzing the actual operation data and supports dynamic adjustment.

[0144] Constraint reasoning mechanism, constraint verification is divided into two stages: hard constraint checking and soft constraint evaluation. Hard constraints include security level constraints, resource capacity constraints, and isolation requirement constraints; soft constraints include performance indicator constraints, load balancing constraints, and optimization target constraints. For found constraint conflicts, they are processed based on preset priority rules.

[0145] The inference results are directly used to support allocation decisions, generate optimization suggestions, provide abnormal warnings, and guide maintenance strategies. For example, in the main pump bearing monitoring system, based on the inference results, the measurement points with high correlation are centrally configured to improve the accuracy of condition assessment.

[0146] Establish a three-level update mechanism: real-time updates triggered by system configuration changes (response time less than 1 second), low-frequency updates triggered by performance anomalies (intervals of 1 minute), and regular maintenance updates (intervals of 1 day). The update scope is determined by the triggering conditions and may be local updates, associated updates, or global updates.

[0147] Use version control and transaction management to ensure data consistency, implement concurrency control to avoid update conflicts, and establish a rollback mechanism to deal with update failures. For example, during system upgrades, use version control mechanisms to ensure the safe transition of configuration data.

[0148] Adopt a hierarchical storage strategy to store frequently accessed data in a fast storage layer; optimize index design to improve query efficiency; implement cache management to reduce database access; use data compression to reduce storage overhead.

[0149] Implement query plan optimization and select the optimal query path; use parallel query processing to improve response speed; establish a result caching mechanism to reuse query results; implement incremental calculation strategies to reduce repeated operations.

[0150] The input interface of the intelligent allocation module includes the affinity calculation module interface and the knowledge graph module interface.

[0151] The affinity calculation module interface is used to obtain the affinity matrix. Its data format is an n×n matrix in the standardized [-1,1] interval. The update cycle is the real-time calculation result, and the maximum delay is <50ms. The data items are the signal pair affinity value, calculation timestamp, and credibility tag.

[0152] Regarding weight update response, dimension weights: physical location (w1), functional association (w2), security level (w3), power supply characteristics (w4). Adjustment range: support ±10% dynamic adjustment. Update trigger: performance abnormality, uneven load, manual intervention. Regarding calculation result verification, integrity check: matrix data integrity, time series continuity; validity verification: value range, logical consistency; exception handling: data compensation, smooth transition, alarm trigger.

[0153] The knowledge graph module interface is used to obtain constraint conditions. Hard constraint rules: security isolation, resource capacity, signal compatibility; soft constraint parameters: load balancing, performance indicators, optimization goals; dynamic constraints: operating status, system load, environmental changes.

[0154] Application of reasoning results: Optimization path: optimal allocation path based on graph reasoning; Resource matching: resource allocation suggestions based on historical experience; Conflict handling: rule-based constraint conflict resolution method;

[0155] Optimization strategy supplement, experience rules: historical best practices, typical case references; expert knowledge: key scenario processing, special situation response; dynamic adjustment: strategies based on real-time status.

[0156] The output interface outputs the allocation plan and performance evaluation data.

[0157] The allocation plan is also called signal allocation mapping, and its format definition is: {signal ID, resource ID, channel number, parameter configuration}; batch processing: supports single-point allocation and batch allocation; version control: plan version number, change record, rollback support; execution instruction sequence, operation steps: detailed execution steps and timing requirements; risk control: key point confirmation, emergency response plan; rollback mechanism: failure recovery, state restoration, log recording.

[0158] Performance evaluation data, evaluation indicators include response performance: signal response time, sampling synchronization, control accuracy; resource utilization: load distribution, bandwidth occupancy, power consumption; system reliability: stability score, fault prediction, maintenance difficulty.

[0159] Comparative analysis, historical comparison: comparison with historical optimal state; standard comparison: comparison with design index requirements; similar comparison: comparison with similar system performance.

[0160] This application adopts a two-stage hierarchical optimization strategy: in the first stage, the affinity clustering algorithm is used to organize highly correlated signals into signal clusters to reduce the scale of the problem; in the second stage, hardware resources are allocated to each signal cluster based on the global resource optimization algorithm. This strategy can effectively reduce the computational complexity and improve the quality of the allocation scheme. Figure 5 shown.

[0161] A signal cluster refers to a collection of signals with strong affinity. For example, in the main pump bearing monitoring system, a typical signal cluster includes: temperature signals (TI101-TI103), vibration signals (VI101) and speed signals (SI101). The average affinity of these signals is greater than 0.85. Each signal cluster has a minimum size (2 signals) and a maximum size (8 signals).

[0162] The resource pool adopts a three-level structure: board-level resource pool, cage-level resource pool, and cabinet-level resource pool. Each level of resource pool maintains a list of currently available resources, resource status information, and performance indicator data. For example, the AI ​​board resource pool records information such as channel usage, load level, and response time of each board.

[0163] Hard constraints include safety level isolation requirements (such as isolation of level 1E and non-level 1E signals), channel capacity restrictions (such as a maximum of 16 channels per AI board), power supply compatibility requirements (such as UPS power supply signal concentration), and signal type matching (such as separation of analog and digital quantities). Allocation plans that violate hard constraints are directly rejected.

[0164] Soft constraints include load balancing requirements (board load deviation does not exceed ±5%), resource utilization requirements (board utilization is not less than 70%), performance optimization goals (response time is less than 3ms), etc. Soft constraints affect solution evaluation through weighted scoring.

[0165] Put all the signals to be assigned into the unassigned set U and create an empty signal cluster set C. For each signal in U, calculate its importance score, and the score calculation takes into account the signal's function type (such as control signal 1.0, monitoring signal 0.8) and security level (such as 1E level 1.0, non-1E level 0.8).

[0166] The signal with the highest importance score is selected from the unassigned set U as the seed signal of the new cluster. For example, in the water supply control system, the flow control signal (FIC101) has a higher importance score and is preferentially used as the seed signal.

[0167] With the seed signal as the center, other signals are absorbed in turn according to the affinity. The specific process is: calculate the affinity of the unassigned signal and the seed signal, select the signal with the highest affinity and satisfying the constraint conditions to join the cluster. For example, for the seed signal FIC101, the flow measurement signal FI101 (affinity 0.92) is added to the cluster first.

[0168] Optimize and adjust the formed signal clusters, including: checking whether the cluster size is within the specified range, verifying the compatibility of the signals within the cluster, evaluating the performance indicators of the cluster, etc. If necessary, split or merge the clusters that do not meet the requirements.

[0169] Evaluate whether the size of the signal cluster is appropriate. A cluster that is too small (less than 2 signals) may lead to a waste of resources, and a cluster that is too large (greater than 8 signals) may affect system flexibility. For example, in the main pump bearing monitoring system, a cluster consisting of 3 temperature measurement points and 1 vibration measurement point has an appropriate size.

[0170] Calculate the average affinity of the signals within the cluster to evaluate the compactness of the cluster. For example, the average affinity of a signal cluster is 0.88, indicating that the correlation degree of the signals within the cluster is high and it is suitable for centralized configuration.

[0171] Evaluate whether the distribution of signals in the cluster is balanced in various dimensions, including physical location distribution, function type distribution, security level distribution, etc. For example, check whether there is an overly concentrated security level distribution in the cluster.

[0172] Establish a three-level priority system: safety level priority (to ensure safety isolation requirements), control loop priority (to ensure control performance), and load balancing priority (to improve resource utilization). For example, the 1E-level signal cluster is given priority in resource allocation to ensure nuclear safety requirements. The allocation sequence is determined according to the cluster size and priority. The general order is: large-scale high-priority cluster → small-scale high-priority cluster → large-scale low-priority cluster → small-scale low-priority cluster. In the allocation process, the association between clusters is considered.

[0173] The load status of the board is evaluated from four aspects: channel utilization, processing load, communication load, and power consumption load. For example, the channel utilization of an AI board is 75%, the processing load is 68%, the communication load is 72%, the power consumption load is 70%, and the overall load level is 71%.

[0174] Set the load balancing target: the board load is maintained within the range of 75% ± 5%. When load imbalance is detected, three levels of adjustment are triggered: fine adjustment (signal migration), medium adjustment (cluster redistribution), and major adjustment (resource reorganization). For example, when the AI03 board load reaches 82% and the AI08 is only 65%, the signal migration mechanism is activated.

[0175] Define the signal cluster set to be allocated C={c1,c2,...,cn}; available resource pool R={r1,r2,...,rm}; constraint condition set CON={security level constraint, capacity constraint...}; optimization target set OBJ={response time, load balancing, resource utilization}.

[0176] Signal cluster priority sorting: Calculate the priority score of each cluster: P(ci)=w1×S(ci)+w2×C(ci)+w3×U(ci); where S(ci) is the safety level score, C(ci) is the control function score, and U(ci) is the urgency score. Sort the signal clusters in descending order according to the scores.

[0177] Resource matching calculation: For each signal cluster ci, calculate the matching degree with each resource rj:

[0178] M(ci,rj)=w1×T(ci,rj)+w2×L(ci,rj)+w3×E(ci,rj).

[0179] Where T is the performance matching, L is the load matching, and E is the position matching. Generate a matching matrix M[n×m].

[0180] Iterative allocation process: loop through each signal cluster ci: obtain the resource subset Ri⊆R that meets the constraints; select the resource rj with the highest matching degree in Ri; verify the load balance after allocation; execute allocation and update resource status.

[0181] Load balancing optimization: Calculate the current load distribution standard deviation σ. If σ> threshold threshold: Identify high-load and low-load resources, find migratable signal clusters and perform load redistribution.

[0182] Allocation mapping table Map<signal cluster, resource>; resource load distribution status; performance evaluation indicators.

[0183] Complexity analysis: Time complexity: O(n×m), n is the number of signal clusters, m is the number of resources; Space complexity: O(n×m), mainly used to store the matching matrix.

[0184] Key parameter settings: priority weight: w1=0.5, w2=0.3, w3=0.2; matching weight: w1=0.4, w2=0.4, w3=0.2; load balancing threshold: threshold=10%.

[0185] Embodiment 1: Intelligent signal distribution of a nuclear power plant main pump bearing monitoring system.

[0186] Application background: This embodiment is aimed at upgrading and reconstructing the main pump bearing monitoring system of a nuclear power plant. The system includes 48 signal points:

[0187] Temperature measurement signal: TI101-TI112 (12 units); vibration monitoring signal: VI101-VI108 (8 units); speed measurement signal: SI101-SI104 (4 units); temperature alarm signal: TAH101-TAH112 (12 units); protection trip signal: TAHH101-TAHH112 (12 units).

[0188] Problem diagnosis: The original system has the following problems: Distributed configuration of related signals: temperature, vibration, and speed signals are distributed on 4 different boards; asynchronous sampling: the maximum sampling delay is 12ms; unbalanced load: the AI03 board load is 92%, while the AI08 board load is only 45%; difficult fault diagnosis: the average fault diagnosis time is 4 hours.

[0189] Demand analysis and transformation target requirements: realize centralized configuration of related signals; control sampling delay within 3ms; control board load balancing within 75%±5%; shorten fault diagnosis time to less than 1 hour.

[0190] Solution implementation process, affinity analysis: taking temperature signal TI101 and vibration signal VI101 as examples.

[0191] Physical location dimension (weight 0.25): same local box (1.0) + shared bridge (0.8) = 0.93; functional association dimension (weight 0.35): monitoring loop association (1.0) + data flow dependency (0.8) = 0.90; security level dimension (weight 0.25): both non-1E level = 0.90; power supply characteristic dimension (weight 0.15): same UPS power supply = 0.90. The comprehensive affinity calculation result is 0.91, indicating that the two signals have a strong configuration correlation.

[0192] The knowledge graph was constructed, creating 48 signal nodes and 12 board nodes; 186 affinity relationship edges between signals were established; 48 allocation relationship edges between signals and resources were established; and 24 constraint relationship edges were established.

[0193] Intelligent allocation execution, the solution generation process includes: signal clustering, forming 6 signal clusters, each cluster contains 6-8 highly correlated signals; resource evaluation, identifying AI05-AI08 boards that are most suitable for carrying these signal clusters; allocation mapping, allocating signal clusters to corresponding boards and optimizing channel allocation; load balancing, adjusting the load of each board to the range of 70%-80%.

[0194] The implementation effect is as follows Figure 6 As shown in the figure, before optimization, the data was scattered on 4 boards, and after optimization, it was concentrated on 2 boards.

[0195] Performance improvement: Sampling synchronization delay is reduced from 12ms to 3ms; state assessment accuracy is improved from 72% to 92%; response time is reduced from 180ms to 95ms; system stability jitter is reduced from ±8ms to ±2ms.

[0196] Resource optimization: Number of boards: reduced from 8 to 6; load balancing: difference reduced from 47% to 12%; channel utilization: increased from 65% to 85%; power supply balance: increased from 35% to 90%; for operation and maintenance improvement, fault diagnosis time was shortened from 4 hours to 1 hour; annual average maintenance hours were reduced from 280 man-days to 140 man-days; system availability increased from 99.95% to 99.999%; maintenance efficiency increased from 65% to 85%.

[0197] Embodiment 2: Optimal configuration of DCS control resources in a nuclear power plant water supply system.

[0198] This embodiment is aimed at the control transformation of the secondary water supply system of a nuclear power plant. The transformation content includes: upgrading 8 single-loop PID controls to 4 cascade PID controls; completing the quantitative analysis upgrade of the equipment monitoring function; a total of 104 signal changes: 24 PID control related signals, 32 monitoring function upgrade signals, and 48 new measurement points. Signal type adjustment: 32 DI signals upgraded to AI signals; 36 new AI signals; 12 AO signal adjustments.

[0199] The original system had the following problems: control response time: cascade control loop delay 180ms; regulation accuracy: overshoot up to 25%; system stability: control jitter occurs 12 times per year on average; resource utilization: average board utilization is only 65%.

[0200] The following specific indicators are set: control response time is less than 100ms; overshoot is controlled within 10%; control jitter is eliminated; board utilization is increased to more than 85%.

[0201] Taking flow-pressure cascade control as an example, affinity analysis of related signals is performed: the affinity between flow measurement FI101 and flow control FC101 is 0.92; the affinity between pressure measurement PI102 and pressure control PC102 is 0.90; and the overall affinity of the flow-pressure cascade loop is 0.85.

[0202] A comprehensive assessment of system resources: AI resources: 286 channels are used and 52 channels are available; AO resources: 124 channels are used and 45 channels are available.

[0203] Renovation requirements: 68 new AI channels; 32 AI channels renovated; 12 new AO channels; 24 AO channels renovated;

[0204] Develop a two-stage implementation plan: Phase I (control transformation): Time: First week; Content: Complete cascade transformation; Steps: Single-loop shutdown → Parameter migration → cascade operation → Performance verification; Phase II (monitoring upgrade): Time: 2nd-3rd week; Content: Complete monitoring function upgrade; Method: Gradually switch to zoning isolation mode.

[0205] The implementation effect is as follows Figure 7 As shown, before optimization: single-loop PID, after optimization: cascade PID control.

[0206] Response time: reduced from 180ms to 95ms; overshoot: reduced from 25% to 8%; stability: the average annual jitter number reduced from 12 times to 0 times; control accuracy: improved from ±2.5% to ±0.8%; board utilization: increased from 65% to 85%; load balancing: deviation reduced from 28% to 12%; channel reservation rate: optimized from 20% to 12%; system scalability: reserve 30% upgrade space; system reliability: system availability: reached 99.999%; mean time between failures: reached 8760 hours; mean repair time: reduced to 0.5 hours; maintenance efficiency: improved by 52%.

[0207] Example 3: Optimization of operating performance of DCS system in nuclear power plant.

[0208] This embodiment is aimed at the operation optimization of the DCS of the primary circuit main system of a nuclear power plant. After two years of operation, the system had the following problems: the response time was extended from the design value of 100ms to 180ms; the data quality was degraded: the sampling loss rate reached 2.5%, and the sampling jitter was 15ms; the resource load was uneven: the maximum load of the AI ​​board was 92%, and the minimum load was 45%; the maximum load of the AO board was 88%, and the minimum load was 40%.

[0209] Affinity analysis of 576 signal points revealed that: the physical location dimension scored 0.65 (cross-regional distribution); the functional association dimension scored 0.72 (dispersed control loops); the safety dimension scored 0.88 (insufficient redundant configuration); and the power supply dimension scored 0.70 (unbalanced load). Knowledge graph analysis revealed that: 12 key signal clusters were improperly configured; 8 control paths needed to be optimized; AI03, AI08, and AO02 were system hotspots.

[0210] Signal reorganization optimization: Identify high affinity (>0.85) signal groups: FI101 (flow measurement); FC101 (flow control); PI101 (pressure measurement); PC101 (pressure control); reallocate the signal groups to Rack01-Card03. Resource reallocation: Board load targets: AI boards: 75% ± 5%; AO boards: 70% ± 5%; AI boards use gradual migration; AO boards use batch migration.

[0211] A three-level trigger mechanism is set up, and load deviation >20% and lasts for 5 minutes to trigger balancing; the single adjustment range does not exceed 5%; the interval between two adjustments is no less than 30 minutes.

[0212] System performance: response time reduced from 180ms to 90ms; data collection: packet loss rate reduced from 2.5% to 0.1%; jitter reduced from 15ms to 5ms; data quality: integrity improved from 97.5% to 99.9%; accuracy improved from 98% to 99.95%; resource utilization, board load: load deviation reduced from 47% to 8%; average efficiency improved from 65% to 85%; power supply load: imbalance reduced from 50% to 10%; utilization efficiency improved by 38%; operation and maintenance indicators: failure rate reduced from 24 times / year to 3 times / year; repair time reduced from 4 hours to 1 hour; maintenance efficiency: improved by 52%; availability: reached 99.999%.

[0213] Through Examples 1 to 3, the practicability and effectiveness of the method in different application scenarios of the DCS system of a nuclear power plant are verified. The method is applicable to both system transformation and operation optimization, and has good engineering practice value. The implementation effect comparison is shown in Table 3.

[0214] Table 3 Optimization effect comparison table

[0215]

[0216] This application was actually applied in the digital control systems of three nuclear power plants with an application period of 24 months. The application scenarios include: the transformation of the main pump bearing monitoring system of Nuclear Power Plant A (576 signal points), the upgrade of the secondary circuit water supply system of Nuclear Power Plant B (468 signal points), and the maintenance and optimization of the DCS system of Nuclear Power Plant C (824 signal points).

[0217] Compared with traditional methods, this application significantly improves the efficiency of engineering configuration: signal configuration error rate: reduced from 15.2% to 3.8%; average engineering cycle: shortened from 8 weeks to 5 weeks; manual workload: reduced from 280 man-days / year to 140 man-days / year; rework rate: reduced from 22% to 6%.

[0218] The automation level of system configuration has been improved: the automatic plan generation rate has increased from 48% to 85%; the verification automation rate has increased from 35% to 82%; the accuracy rate of optimization suggestions has increased from 62% to 93%.

[0219] The data sampling synchronization deviation was reduced from 12ms to 3ms; the state assessment accuracy was increased from 72% to 92%; the data integrity was increased from 97.5% to 99.9%; the acquisition cycle jitter was reduced from ±8ms to ±2ms.

[0220] Control performance: loop response time is reduced from 180ms to 95ms; control accuracy is improved from ±2.5% to ±0.8%; overshoot is reduced from 25% to 8%; stabilization time is reduced from 8s to 3s.

[0221] Hardware resource utilization: AI board utilization increased from 65% to 85%; load balancing differences decreased from 47% to 8%; channel reservation rate was optimized from a unified 20% to 8%-15% dynamic reservation; wiring workload was reduced by 28%. System resource allocation: processing load balancing increased by 62%; communication bandwidth utilization increased by 45%; power supply load balancing increased by 55%; heat dissipation efficiency: increased by 35%.

[0222] The system's annual failure rate was reduced from 24 times to 3 times; the average time between failures was increased from 2160 hours to 8760 hours; the average repair time was reduced from 4 hours to 1 hour; and the system availability was increased from 99.95% to 99.999%.

[0223] Maintenance efficiency: Fault location time was reduced from 240 minutes to 60 minutes; daily maintenance hours were reduced from 8 hours / week to 3 hours / week; preventive maintenance coverage was increased from 75% to 95%. Maintenance efficiency was increased from 65% to 85%.

[0224] Security assurance: Regarding isolation effects, the effectiveness of security isolation has been increased from 95% to 99.9%; the scope of fault propagation has been reduced by an average of 65%; the fault impact time has been shortened by an average of 72%; and the security protection response has been increased from 50ms to 20ms. Regarding redundancy effects, the redundancy switching success rate has been increased from 98% to 99.99%; the switching time has been reduced from 100ms to 30ms; data consistency has been increased from 99% to 99.99%; and the system recovery rate has been increased from 99.5% to 99.99%.

[0225] Based on the same inventive concept, the embodiment of the present application also provides a nuclear power DCS signal distribution device for implementing the nuclear power DCS signal distribution method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more nuclear power DCS signal distribution device embodiments provided below can refer to the limitations of the nuclear power DCS signal distribution method above, and will not be repeated here.

[0226] In an exemplary embodiment, a nuclear power DCS signal distribution device is provided, comprising:

[0227] The information data acquisition module is used to acquire information data; the information data includes acquired input data; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data.

[0228] The processing module is used to perform data association interaction and constraint optimization processing according to the information data to obtain the processing result.

[0229] The decision module is used to determine the resource allocation plan according to the processing results; the resource allocation plan is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system.

[0230] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store nuclear power DCS signal distribution data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a nuclear power DCS signal distribution method is implemented.

[0231] Those skilled in the art will understand that Figure 8The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0232] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0233] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0234] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization of the corresponding device owner. It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0235] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0236] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0237] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0238] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A nuclear power DCS signal distribution method, characterized in that: The nuclear power DCS signal distribution method is implemented by a nuclear power DCS signal distribution system based on a three-layer architecture design; the nuclear power DCS signal distribution system includes: an input layer, a core processing layer and a decision layer connected in sequence; wherein the core processing layer is composed of an affinity calculation module, a knowledge graph module and an intelligent allocation module; the affinity calculation module, the knowledge graph module and the intelligent allocation module are all connected to the input layer; the affinity calculation module and the knowledge graph module are also connected to the intelligent allocation module; The nuclear power DCS signal distribution method comprises: Acquire information data; the information data includes input data acquired by the input layer; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data; Based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results; Based on the decision layer, a resource allocation scheme is determined according to the processing result; the resource allocation scheme is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system; Based on the core processing layer, data association interaction and constraint optimization processing are performed according to the information data to obtain processing results, which specifically include: The signal affinity quantification model is used to determine the signal affinity based on the affinity calculation module and the signal configuration information data; the signal affinity is used to characterize the degree of association between the signals in the nuclear power DCS system; the signal affinity quantification model is a mathematical model determined based on the four-dimensional system and the historical operation data by the principal component analysis method; Based on the knowledge graph module, the constructed knowledge graph is used to determine the optimization strategy according to the signal affinity, the DCS system configuration data and the historical operation data; the optimization strategy is based on the set priority search strategy, with node affinity, resource matching degree and path length as search factors, and the analysis strategy corresponding to the configuration path found under the set constraint conditions; the knowledge graph is a physical simulation model of node-relationship edges constructed under the constraint relationship restriction conditions based on the cache reuse strategy using a parallel processing mechanism; Based on the intelligent allocation module, signal clustering processing is performed on the information data according to the signal affinity, and resources are allocated to the clustered data based on the optimization strategy to obtain a processing result.

2. The nuclear power DCS signal distribution method according to claim 1, characterized in that: The four-dimensional system includes: physical location dimension, functional association dimension, security level dimension and power supply characteristic dimension; The expression of the signal affinity quantification model is: ； in: For signal With signal The signal affinity between For the The weight coefficient of the dimension; For the Dimensional signal With signal The distance function of .

3. The nuclear power DCS signal distribution method according to claim 1, characterized in that: Based on the intelligent allocation module, the information data is subjected to signal clustering processing according to the signal affinity, and resources are allocated to the clustered data based on the optimization strategy to obtain a processing result, which specifically includes: Constructing a signal unallocated set; the signal unallocated set contains all the information data to be allocated in the information data; Based on the signal unassigned set, calculating the importance value for each of the information data according to the function type and the security level; Selecting the information data with the highest importance value from the signal unassigned set as a seed signal; Taking the seed signal as the center, absorbing signals according to the signal affinity to perform signal clustering processing to obtain clustering data; Optimizing and adjusting the clustering data to obtain clustered data; Based on the optimization strategy, resources are allocated to the data after clustering processing to obtain processing results.

4. The nuclear power DCS signal distribution method according to claim 3, characterized in that: Optimizing and adjusting the clustering data to obtain clustered data specifically includes: Performing a cluster size check on the cluster data to determine whether the cluster size is within a limited range, and obtaining a judgment result; If the judgment result is yes, the compatibility of the intra-cluster signals of the cluster data is verified, and a performance index is determined. The cluster data is optimized based on the performance index to obtain clustered data; the performance index includes: compactness; the compactness is determined by calculating the average affinity of the intra-cluster signals; If the judgment result is no, the cluster data is split or merged.

5. The nuclear power DCS signal distribution method according to claim 1, characterized in that: The constraint relationship restriction conditions include: constraint identification, constraint type, constraint expression, priority and mandatory flag.

6. A nuclear power DCS signal distribution device, characterized in that: The nuclear power DCS signal distribution device is used to execute the nuclear power DCS signal distribution method described in any one of claims 1 to 5; The nuclear power DCS signal distribution device comprises: An information data acquisition module, used to acquire information data; the information data includes acquired input data; the input data includes: signal configuration information data, hardware resource status data, DCS system configuration data and historical operation data; A processing module, used to perform data association interaction and constraint optimization processing according to the information data to obtain a processing result; The decision module is used to determine a resource allocation scheme according to the processing result; the resource allocation scheme is used to characterize the mapping relationship between signals and resources in the nuclear power DCS system.

7. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the nuclear power DCS signal distribution method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the nuclear power DCS signal distribution method described in any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the nuclear power DCS signal distribution method described in any one of claims 1 to 5 is implemented.

Images