CN-122019996-A - Equipment fault diagnosis dynamic optimizing method and system based on intelligent power plant

CN122019996ACN 122019996 ACN122019996 ACN 122019996ACN-122019996-A

Abstract

The invention discloses a dynamic optimizing method and a system for equipment fault diagnosis based on an intelligent power plant, which relate to the technical field of intelligent operation and maintenance of the intelligent power plant and comprise the steps of constructing a three-level structure of a field acquisition layer, an edge calculation layer and a data center layer, and carrying out standardized acquisition and feature construction of multi-source operation data through unified coding, time synchronization and a semantic tag system; the method comprises the steps of constructing a mechanism model and AI model double-channel fusion diagnosis engine, utilizing residual calculation and consistency analysis to output parallel reasoning and structural diagnosis of multi-source data, constructing a dynamic optimizing model containing three elements of energy efficiency, time delay and risk based on structural diagnosis, and reconfiguring operation parameters in real time through constraint solving and control quantity mapping mechanism. The method realizes the fusion diagnosis and dynamic optimizing linkage of the multi-source data, has the capabilities of real-time analysis, self-adaptive optimization and continuous learning, and improves the intelligent level of equipment operation.

Inventors

WAN QIANG
ZHANG QIANG
ZHOU SHIQUAN
DU QIANG
GUO ZHAOYUN
ZHAO RUNLIN
ZHANG HUANHUAN
WANG SUO

Assignees

贵州织金平远清洁能源有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260128

Claims (10)

1. An equipment fault diagnosis dynamic optimizing method based on an intelligent power plant is characterized by comprising the following steps: Constructing a site acquisition layer, an edge calculation layer and a data center layer three-level structure, and carrying out standardized acquisition and feature construction of multi-source operation data through unified coding, time synchronization and a semantic tag system; After the characteristics are built, a mechanism model and AI model double-channel fusion diagnosis engine is built, and parallel reasoning and structural diagnosis of multi-source data are output by residual calculation and consistency analysis; based on structural diagnosis, constructing a dynamic optimizing model comprising three elements of energy efficiency, time delay and risk, and reconfiguring operation parameters in real time through a constraint solving and control quantity mapping mechanism; And constructing a sample management, incremental training and parameter recalibration mechanism through the dynamic optimizing model and log data output by the structural diagnosis.
2. The dynamic optimizing method for equipment fault diagnosis based on intelligent power plant as claimed in claim 1, wherein the construction of the three-level structure of the field acquisition layer, the edge calculation layer and the data center layer comprises the steps of, The field acquisition layer performs data interaction with DCS, BOP, NCS systems through OPC UA, modbus TCP or IEC 61850 protocols; the edge computing layer performs data cleaning, time sequence alignment, feature extraction and wavelet compression processing on the original signals; the data center layer adopts a hierarchical storage structure of a real-time database, a historical database and a relational database to centrally manage the characteristic data in a unified JSON Schema format.
3. The intelligent power plant-based equipment fault diagnosis dynamic optimizing method as claimed in claim 1 or 2, wherein the construction mechanism model and AI model dual-channel fusion diagnosis engine comprises, The system comprises a mechanism modeling layer, an AI modeling layer, a fusion reasoning layer and a model management layer; The mechanism modeling layer establishes a mathematical model based on a device thermodynamic equation, an energy balance equation and structural characteristics, the AI modeling layer trains characteristic data samples based on multiple algorithms, and the fusion reasoning layer calculates through residual errors.
4. The method for dynamic optimizing equipment fault diagnosis in a smart power plant as claimed in claim 3, wherein said dynamic optimizing model comprises, The multi-objective optimization model is adopted as follows: ; Wherein, the Indicating a loss of energy efficiency and, Indicating a time delay in maintenance and, Representing the risk factor(s), 、、 Representing weight coefficients, and dynamically adjusting by the running state of the system; Meaning of representation is preferably added to And the optimizing process executes constraint solving in a manner of combining a Lagrangian multiplier method and Newton iteration, and converts an optimizing result into control quantity set values of the combustion, the fan and the pump valve based on a mapping table.
5. The method for dynamic optimizing equipment fault diagnosis in a smart power plant according to claim 1, 2 or 4, wherein the dynamic optimizing model comprises, After receiving equipment running state information output by a diagnosis engine, the dynamic optimizing model acquires dynamic fluctuation values of load change, temperature deviation and energy efficiency indexes through a real-time data feedback module; constructing a multi-parameter optimizing iteration equation comprising operation energy consumption, equipment load and working condition stability according to the dynamic fluctuation value, and executing rolling update in a continuous time window; And correcting the change step length of the control variable according to the real-time data by each round of optimizing iteration, and synchronously verifying the change step length through a feedback channel and a mechanism model.
6. The method for dynamic optimizing equipment fault diagnosis in a smart power plant as claimed in claim 5, wherein said real-time reconfiguration of operating parameters by constraint solving and control quantity mapping mechanism comprises, In the execution process of the optimizing control quantity, the double-layer buffer control structure formed by the logic buffer layer and the time buffer layer is used for carrying out optimizing instruction smoothing; the logic buffer layer automatically compares the difference of the set values before and after the optimization, and pauses the instruction issuing and requests the manual confirmation when the change amplitude of the control quantity exceeds a threshold value; the time buffer layer adopts a linear recurrence updating mode in the interval of 5 seconds to 15 seconds to adjust the control parameters step by step.
7. The method for dynamic optimizing equipment fault diagnosis in a smart power plant according to claim 1, 2, 4, or 6, wherein the mechanism for constructing sample management, incremental training, and parameter recalibration comprises, Sample data extraction, label binding, sample cleaning and sample balancing; the sample label is generated by a confirmation record of the inspection system or the inspection work order system, the training sample updates model parameters in an incremental learning mode, and the parameter update is expressed as: ; Wherein, the Representation of The parameters of the time-of-day model, Representation of The parameters of the time-of-day model, Representing the gradient update amount of the new sample training, Representing the migration coefficient.
8. The intelligent power plant-based equipment fault diagnosis dynamic optimizing system adopts the intelligent power plant-based equipment fault diagnosis dynamic optimizing method according to any one of claims 1-7, and is characterized by comprising a multi-source data acquisition and feature construction module, a mechanism and AI fusion diagnosis modeling module, a dynamic optimizing decision and working condition reconfiguration module and an online self-learning and model self-adaptive updating module; the multi-source data acquisition and feature construction module is used for unifying the structures of the field acquisition layer, the edge calculation layer and the data center layer and carrying out standardized acquisition, time synchronization and feature vector generation on multi-source operation data from DCS, BOP, NCS and the sensor; The mechanism and AI fusion diagnosis modeling module is used for establishing a mechanism model and AI model double-channel diagnosis engine, and realizing equipment running state identification, abnormality detection and structural diagnosis output through residual calculation and consistency analysis; The dynamic optimizing decision and working condition reconfiguration module is used for constructing a dynamic optimizing model, carrying out constraint solving on comprehensive energy efficiency, time delay and risk factors, and mapping optimal control parameters into control system set values so as to realize real-time working condition adjustment; the online self-learning and model self-adaptive updating module is used for executing sample management, incremental training and parameter recalibration.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the intelligent power plant based device fault diagnosis dynamic optimization method of any one of claims 1-7.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the intelligent power plant based equipment fault diagnosis dynamic optimizing method according to any one of claims 1-7.

Description

Equipment fault diagnosis dynamic optimizing method and system based on intelligent power plant Technical Field The invention relates to the technical field of intelligent operation and maintenance of intelligent power plants, in particular to an equipment fault diagnosis dynamic optimizing method and system based on an intelligent power plant. Background Along with the continuous improvement of energy structure transformation and power system digitization level, intelligent power plants become an important direction for modern power generation enterprises to realize intensive management and efficient operation and maintenance. In recent years, technologies such as the Internet of things, big data, cloud computing and artificial intelligence are gradually applied to the fields of operation monitoring and fault diagnosis of generator sets, and an intelligent system with multi-source data acquisition, online monitoring analysis and remote operation and maintenance as cores is formed. However, most of the existing systems focus on static feature analysis and single model judgment, and real-time multi-source data fusion and dynamic optimization decision making are difficult to achieve. In this context, how to implement device state diagnosis and dynamic optimization of operation parameters based on an intelligent power plant architecture has become an important research direction in the field of intelligent operation and maintenance. The existing intelligent power plant equipment operation and maintenance method generally depends on a single monitoring system or a static algorithm model, and lacks the unified modeling capability for multi-source heterogeneous data. The data structure, time stamp and semantics are different among different systems, which results in inconsistency between feature extraction and state evaluation. In addition, although the traditional mechanism model has physical constraint, the traditional mechanism model is difficult to cope with multidimensional nonlinear change under complex operation conditions, while the pure AI model has self-learning capability, is easily influenced by data noise and sample deviation, and cannot form interpretable fault mechanism reasoning. In the aspect of operation optimization, the existing research generally adopts an offline calculation mode to analyze energy efficiency and control parameters, lacks a dynamic feedback mechanism based on real-time diagnosis results, and cannot automatically adjust working conditions or optimize control quantity in the operation process, so that the problems of high energy consumption, adjustment lag and the like are caused. Meanwhile, the mainstream system generally lacks a continuous learning mechanism of log data, and model parameters are not updated for a long time, so that diagnosis accuracy and optimization performance are reduced with time. Disclosure of Invention The present invention has been made in view of the above-described problems. The intelligent power plant equipment diagnosis and operation optimization method solves the technical problems that the existing intelligent power plant equipment diagnosis and operation optimization method has inaccurate feature extraction, low model fusion degree, difficulty in realizing real-time diagnosis, lack of a dynamic optimization mechanism for operation parameter optimization lag and how to realize equipment fault diagnosis and self-adaptive dynamic optimization of operation working conditions under a multi-model fusion framework due to the fact that a multi-source data structure is not uniform. The technical scheme includes that an equipment fault diagnosis dynamic optimizing method based on an intelligent power plant comprises the steps of constructing a site acquisition layer, an edge calculation layer and a data center layer three-level structure, carrying out standardized acquisition and feature construction of multi-source operation data through a unified coding, time synchronization and semantic tag system, constructing a mechanism model and an AI model dual-channel fusion diagnosis engine after feature construction, utilizing residual calculation and consistency analysis to output parallel reasoning and structural diagnosis of the multi-source data, constructing a dynamic optimizing model containing three elements of energy efficiency, time delay and risk based on the structural diagnosis, carrying out real-time reconfiguration of operation parameters through a constraint solving and control quantity mapping mechanism, and constructing a sample management, incremental training and parameter recalibration mechanism through the dynamic optimizing model and log data output by the structural diagnosis. The intelligent power plant-based equipment fault diagnosis dynamic optimizing method is characterized in that the construction of a three-level structure of a field acquisition layer, an edge calculation layer and a data center layer comprises the steps that the fi