Power plant fault early warning method and system based on edge calculation and 5G communication
Through edge computing and 5G communication technology, combined with machine learning algorithms, the problem of insufficient data transmission delay and early warning accuracy of power plant equipment fault warning systems is solved, real-time monitoring and efficient fault warning of power plant equipment are realized, and the safety of equipment operation is improved and reliability.
Patent Information
- Application Number
- CN202410864415.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-30
- Publication Date
- 2024-11-01
AI Technical Summary
Traditional power plant equipment fault warning systems have problems such as high data transmission delay, large bandwidth pressure, network fluctuations, and insufficient warning accuracy and timeliness.
Adopt edge computing and 5G communication technology to deploy edge computing nodes near key equipment in the power plant, collect and preprocess data in real time, and use 5G base stations and network slicing technology to build a communication network, combine machine learning algorithms to establish a predictive maintenance model, and dynamically adjust it Fault warning threshold, real-time transmission of early warning information to the central monitoring system.
Real-time monitoring and accurate fault warning of power plant equipment, reduce data transmission delay, improve early warning accuracy and timely transmission of fault information, and ensure the safety and reliability of equipment operation.
Smart Images

Figure CN118887779A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of industrial intelligent monitoring technology, and specifically to a power plant fault early warning method and system based on edge computing and 5G communication. Background Art
[0002] With the continuous development of industrial automation and intelligent monitoring technology, fault warning and maintenance of power plant equipment has become an important research field. In recent years, the progress of edge computing and 5G communication technology has provided new solutions for real-time monitoring and data transmission of industrial equipment. Edge computing technology can effectively reduce the delay of data transmission and improve the real-time performance of data processing by deploying computing nodes near the equipment to realize local processing and analysis of data. At the same time, 5G communication technology, with its high bandwidth, low latency and high reliability, provides guarantee for data transmission between equipment and between equipment and central monitoring system. On this basis, machine learning technology plays an important role in equipment failure prediction and warning. Through the analysis and modeling of historical data, the operating status and potential failures of equipment can be accurately predicted, so as to take preventive measures in advance, reduce equipment downtime and improve production efficiency.
[0003] However, the existing power plant equipment fault warning system still has some shortcomings in practical applications. First, most traditional monitoring systems rely on centralized data processing modes, and all data need to be transmitted to the central server for processing, resulting in high data transmission delays and unable to meet the needs of real-time monitoring. Secondly, the existing system is easily affected by network fluctuations and delays during data transmission, resulting in data loss or delays, affecting the accuracy and timeliness of monitoring. Thirdly, traditional fault warning methods are mostly based on simple threshold judgments, and fail to make full use of the characteristics of historical data and real-time data, resulting in insufficient accuracy and sensitivity of the warning. In addition, the existing system lacks an effective mechanism for the transmission and processing of fault warning information, and cannot promptly transmit the warning information to the central monitoring system, affecting the efficiency of fault handling. In contrast, the present invention can effectively solve the above problems by introducing edge computing and 5G communication technology, combined with machine learning algorithms, and provides an efficient and accurate power plant equipment fault warning method. Summary of the invention
[0004] In view of the above-mentioned problems, the present invention is proposed.
[0005] Therefore, the technical problem solved by the present invention is: how to solve the problems of high latency and high bandwidth pressure in traditional monitoring systems, improve the real-time performance and reliability of the monitoring system, and realize real-time monitoring and fault warning of power plant equipment.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: A power plant fault early warning method based on edge computing and 5G communication, comprising: deploying edge computing nodes near key equipment in the power plant to collect and pre-process the operation data of the equipment in real time;
[0007] A communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system;
[0008] Establish a predictive maintenance model through historical data and machine learning algorithms, and deploy it on edge computing nodes for real-time data analysis and fault warning;
[0009] Dynamically adjust the fault warning threshold based on real-time and historical data;
[0010] The warning information is transmitted to the central monitoring system in real time through the 5G network, triggering the corresponding fault handling process.
[0011] As a preferred solution of the power plant fault early warning method based on edge computing and 5G communication described in the present invention, the key equipment of the power plant includes boilers, steam turbines, generators, transformers and cooling systems;
[0012] The operating data of the acquisition equipment includes temperature, pressure, vibration, current, voltage and rotation speed;
[0013] The preprocessing includes filtering, denoising, normalization and feature extraction.
[0014] As a preferred solution of the power plant fault early warning method based on edge computing and 5G communication described in the present invention, the constructed communication network includes deploying 5G base stations around key equipment in the power plant and necessary monitoring areas, and configuring multi-band support between base stations;
[0015] Using 5G network slicing technology, the internal communication needs of the power plant are divided into multiple network slices, and each slice allocates independent bandwidth and priority for specific data flows;
[0016] Dynamically adjust network resource allocation and automatically optimize bandwidth configuration of each slice based on real-time monitored data traffic;
[0017] The specific data flows include monitoring data flows, control data flows and warning data flows.
[0018] As a preferred solution of the power plant fault early warning method based on edge computing and 5G communication described in the present invention, the automatic optimization of the bandwidth configuration of each slice includes weighted integration of the real-time data flow of each slice to evaluate its data transmission demand within a specific time period, and the formula is:
[0019]
[0020] Among them, I i (t) represents slice S i The weighted integral of data flow in time t, α i Represents slice S i The bandwidth weight coefficient, F i (τ) represents the real-time data flow at time τ; τ represents the time variable, which is used for integral calculation;
[0021] Exponential smoothing is performed on the data traffic change rate to capture the fluctuation of data traffic and consider the impact on bandwidth requirements. The formula is:
[0022]
[0023] Among them, E i (t) represents slice S i With other slices j The sum of the interaction effects of ij Represents slice S i With other slices j The interaction coefficient between them; represents the exponential smoothing function of the data flow rate change, γ j represents the exponential smoothing coefficient, Indicates the rate of change of data traffic; M indicates the total number of all slices in the network;
[0024] The data flow of each slice is normalized and multiplied by complex transformation, and the overall bandwidth requirement of each slice is comprehensively considered. The formula is:
[0025]
[0026] Among them, P i (t) represents slice S i The product of normalization and complex transformation; σ k represents the normalization coefficient; ψ k (F k (τ)) represents the transformation function of data flow;
[0027] The bandwidth allocation calculation formula for each slice is:
[0028]
[0029] Among them, B i (t) represents slice S i The bandwidth allocation at time t, B total Indicates the total network bandwidth, I k(t represents the set of weighted integrals of data traffic of all slices within time t, E k (t) represents the sum of the interaction effects of all slices, P k (t) represents the set of product results of normalization and complex transformation of all slices;
[0030] If B i When (t) = 0, the slice S i No bandwidth is currently allocated;
[0031] If B i (t) = B total When S i Currently occupying the entire network bandwidth.
[0032] As a preferred solution of the power plant fault early warning method based on edge computing and 5G communication described in the present invention, wherein: the establishment of a predictive maintenance model includes using principal component analysis to perform feature selection and dimensionality reduction to extract the most representative feature vector;
[0033] New features are constructed through feature engineering. The formula is:
[0034]
[0035] Among them, Φ represents the weighted eigenvector, α i represents the feature weight coefficient, ψ i (x i ) represents the feature transformation function, x i represents the original feature, τ represents the time variable, and m represents the number of features;
[0036] Establish a predictive maintenance model, the formula is:
[0037]
[0038] in, represents the predicted value, N represents the number of decision trees, and f i (x) represents the prediction result of the i-th decision tree;
[0039] Use historical data to train the model, evaluate the model performance through the validation set, and adjust the model parameters;
[0040] Deploy the trained model on the edge computing node to perform real-time data analysis and fault warning, monitor the equipment operation status in real time, and generate warning information based on the model prediction results. The formula is:
[0041]
[0042] in, represents the prediction result, βj represents the model parameters, γ j represents the exponential smoothing coefficient, represents the rate of change of the weighted eigenvector;
[0043] Use real-time data for analysis and fault warning. The formula is:
[0044]
[0045] Among them, Ω(t) represents the fault warning index, σ k represents the normalization coefficient, ψ k (x k (τ)) represents the real-time feature transformation function, x k (τ( represents the real-time features at time τ, and n represents the number of real-time features.
[0046] As a preferred solution of the power plant fault warning method based on edge computing and 5G communication described in the present invention, the dynamically adjusting the threshold of the fault warning includes calculating the initial fault warning threshold, and the formula is:
[0047] θ0=μ h +k h ·σ h
[0048] Among them, θ0 represents the initial warning threshold, μ h represents the mean of historical data, k h represents the adjustment coefficient, σ h Indicates the standard deviation of historical data;
[0049] Dynamically adjust the warning threshold, combined with real-time data characteristics, the formula is:
[0050]
[0051] Among them, θ(t) represents the warning threshold after dynamic adjustment, θ0 represents the initial warning threshold, α represents the adjustment coefficient, T represents the time window length, and x i (t) represents the real-time data characteristics at time t, μ r Represents the mean of the real-time data features.
[0052] As a preferred solution of the power plant fault warning method based on edge computing and 5G communication described in the present invention, the fault handling process includes setting two warning thresholds as a low warning threshold θ(t)1 and a high warning threshold θ(t)2, where θ(t)1<θ(t)2;
[0053] According to the fault warning index Ω(t) and the set threshold, the fault warning is divided into three levels:
[0054] When Ω(t)≤θ(t)1, a low-level warning is triggered, the equipment status is checked, the current equipment operating parameters are recorded, and the equipment operating parameters are adjusted;
[0055] When θ(t)1<Ω(t)≤θ(t)2, an intermediate warning is triggered and an intermediate warning message is sent to the central monitoring system to remind maintenance personnel to pay attention to the equipment status and to shut down the equipment for inspection and component replacement.
[0056] When Ω(t)>θ(t)2, an advanced warning is triggered, and the advanced warning information is immediately sent to the central monitoring system, triggering the emergency maintenance process, immediately shutting down, starting the backup equipment, and calling a professional maintenance team for comprehensive inspection and repair;
[0057] Using the 5G communication network, the packaged warning information data packets are transmitted to the central monitoring system in real time;
[0058] The warning information data packet includes a device ID, a fault warning index, a dynamically adjusted warning threshold, and a warning level.
[0059] A power plant fault early warning system based on edge computing and 5G communication, characterized in that: it includes:
[0060] Data collection and preprocessing module: By deploying edge computing nodes near key equipment in the power plant, the operating data of the equipment is collected in real time and preprocessed;
[0061] Communication network building module: A communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system;
[0062] Predictive maintenance module: Establish a predictive maintenance model through historical data and machine learning algorithms, and deploy it on edge computing nodes for real-time data analysis and fault warning;
[0063] Dynamic threshold adjustment module: dynamically adjusts the fault warning threshold based on real-time data and historical data;
[0064] Early warning information transmission and fault handling module: The early warning information is transmitted to the central monitoring system in real time through the 5G network, and the corresponding fault handling process is triggered.
[0065] The beneficial effects of the invention are as follows: The invention realizes real-time monitoring and accurate fault warning of power plant equipment through edge computing, 5G communication and machine learning technology, reduces data transmission delay, improves warning accuracy, ensures timely transmission of fault information and triggers corresponding processing procedures, and improves equipment operation safety and reliability. BRIEF DESCRIPTION OF THE DRAWINGS
[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the figures required for describing the embodiments. Obviously, the figures described below are only some embodiments of the present invention. For ordinary technicians in this field, other figures can be obtained based on these figures without creative work. Among them:
[0067] Figure 1 A flowchart of a power plant fault warning method based on edge computing and 5G communication is provided for the first embodiment of the present invention. DETAILED DESCRIPTION
[0068] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the specification drawings. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0069] Example 1
[0070] Reference Figure 1 , as an embodiment of the present invention, provides a power plant fault early warning method based on edge computing and 5G communication, including:
[0071] S1: By deploying edge computing nodes near key equipment in power plants, the operating data of the equipment is collected in real time and pre-processed.
[0072] The key equipment of a power plant includes boilers, turbines, generators, transformers and cooling systems; these equipment are the core of the power plant operation, and their operating status is directly related to the safety and efficiency of the power plant.
[0073] The operating data of the collected equipment include temperature, pressure, vibration, current, voltage and speed; these data are key parameters reflecting the operating status of the equipment. By monitoring these parameters in real time, potential faults can be discovered in time.
[0074] Preprocessing includes filtering, denoising, normalization, and feature extraction; these technical steps are used to clean and standardize the data to make it suitable for subsequent analysis and processing.
[0075] Furthermore, edge computing nodes refer to computing devices deployed near key equipment in power plants, which are responsible for real-time collection and preprocessing of equipment operating data. The use of edge computing nodes can significantly reduce data transmission delays and improve real-time performance.
[0076] It should be noted that by deploying edge computing nodes near key equipment in power plants, collecting the operating data of the equipment in real time and performing preprocessing, the present invention solves the problems of poor real-time performance, low data quality, and insufficient accuracy of fault warning in the prior art. Specifically, edge computing nodes reduce data transmission delays and improve the real-time performance of data processing; preprocessing technology ensures data quality and comparability; feature extraction technology enhances the accuracy of data analysis and lays the foundation for accurate fault warning. Through these technological innovations, the present invention has significant advantages in improving the safety and reliability of power plant equipment operation.
[0077] S2: A communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system.
[0078] 5G base stations are deployed around key equipment in power plants and in necessary monitoring areas, and multi-frequency band support is configured between base stations.
[0079] Using 5G network slicing technology, the internal communication needs of the power plant are divided into multiple network slices, and each slice allocates independent bandwidth and priority to specific data flows.
[0080] Dynamically adjust network resource allocation and automatically optimize the bandwidth configuration of each slice based on real-time monitored data traffic.
[0081] Specific data flows include monitoring data flows, control data flows, and warning data flows.
[0082] The real-time data traffic of each slice is weighted and integrated to evaluate its data transmission demand in a specific time period. The formula is:
[0083]
[0084] Among them, I i (t) represents slice S i The weighted integral of data flow in time t, α i Represents slice S i The bandwidth weight coefficient, F i (τ) represents the real-time data flow at time τ; τ represents the time variable, which is used for integral calculation.
[0085] Exponential smoothing is performed on the data traffic change rate to capture the fluctuation of data traffic and consider the impact on bandwidth requirements. The formula is:
[0086]
[0087] Among them, E i (t) represents slice S i With other slices jThe sum of the interaction effects of ij Represents slice S i With other slices j The interaction coefficient between them; represents the exponential smoothing function of the data flow rate change, γ j represents the exponential smoothing coefficient, Indicates the rate of change of data traffic; M represents the total number of all slices in the network.
[0088] The data flow of each slice is normalized and multiplied by complex transformation, and the overall bandwidth requirement of each slice is comprehensively considered. The formula is:
[0089]
[0090] Among them, P i (t) represents slice S i The product of normalization and complex transformation; σ k represents the normalization coefficient; ψ k (F k (τ)) represents the transformation function of data flow.
[0091] The bandwidth allocation calculation formula for each slice is:
[0092]
[0093] Among them, B i (t) represents slice S i The bandwidth allocation at time t, B total Indicates the total network bandwidth, I k (t represents the set of weighted integrals of data traffic of all slices within time t, E k (t) represents the sum of the interaction effects of all slices, P k (t) represents the set of product results of normalization and complex transformation of all slices.
[0094] When B i (t) approaches 0, then the slice S i The current bandwidth allocation is low. If B i When (t) = 0, the slice S i No bandwidth is currently allocated; when B i (t) approaches B total , then slice S i The current bandwidth allocation is high. If B i (t) = B total When S i Currently occupying the entire network bandwidth.
[0095] It should be noted that 5G base stations refer to base stations deployed around key equipment in power plants and necessary monitoring areas, which support multi-band communications and ensure high-bandwidth, low-latency data transmission.
[0096] Dedicated network slicing technology: The communication network is divided into multiple virtual slices, each of which is independently allocated bandwidth and priority to meet the transmission requirements of different data streams.
[0097] Monitoring data flow: real-time collection and transmission of equipment operating data, such as temperature, pressure, vibration, etc.
[0098] Control data stream: transmits device control instructions and parameter adjustment information.
[0099] Early warning data stream: Transmits fault early warning information to ensure timely response.
[0100] Furthermore, by using a communication network constructed using 5G base stations and dedicated network slicing technology, the present invention realizes real-time data transmission between edge computing nodes and the central monitoring system, effectively solving the problems of poor data transmission reliability and real-time performance, inflexible network resource allocation, and insufficient bandwidth allocation optimization in the prior art. Through the deployment of 5G base stations supported by multiple frequency bands, the application of network slicing technology, and a dynamic bandwidth allocation method based on weighted integration, exponential smoothing, and normalization, the present invention has significant advantages in improving the accuracy and timeliness of power plant equipment monitoring and fault warning.
[0101] S3: Establish a predictive maintenance model through historical data and machine learning algorithms, and deploy it on edge computing nodes for real-time data analysis and fault warning.
[0102] Principal component analysis is used for feature selection and dimensionality reduction to extract the most representative feature vectors.
[0103] New features are constructed through feature engineering. The formula is:
[0104]
[0105] Among them, Φ represents the weighted eigenvector, α i represents the feature weight coefficient, ψ i (x i ) represents the feature transformation function, x i represents the original features, τ represents the time variable, and m represents the number of features.
[0106] Establish a predictive maintenance model, the formula is:
[0107]
[0108] in, represents the predicted value, N represents the number of decision trees, and fi (x) represents the prediction result of the i-th decision tree.
[0109] Use historical data to train the model, evaluate the model performance through the validation set, and adjust the model parameters.
[0110] Deploy the trained model on the edge computing node to perform real-time data analysis and fault warning, monitor the equipment operation status in real time, and generate warning information based on the model prediction results. The formula is:
[0111]
[0112] in, represents the prediction result, β j represents the model parameters, γ j represents the exponential smoothing coefficient, Represents the rate of change of the weighted eigenvector.
[0113] Use real-time data for analysis and fault warning. The formula is:
[0114]
[0115] Among them, Ω(t) represents the fault warning index, σ k represents the normalization coefficient, ψ k (x k (τ)) represents the real-time feature transformation function, x k (τ) represents the real-time features at time τ, and n represents the number of real-time features.
[0116] It should be noted that by establishing a predictive maintenance model through historical data and machine learning algorithms, the present invention effectively solves the problems of feature selection and data dimensionality reduction, feature engineering and new feature construction, predictive maintenance model establishment and optimization, and real-time data analysis and fault warning in the prior art. The most representative feature vectors are extracted through principal component analysis (PCA), and new features are constructed through feature engineering, which improves the predictive performance of the model; the predictive maintenance model is trained and optimized through historical data, and accurate prediction of equipment failures is achieved; the trained model is deployed on the edge computing node for real-time data analysis and fault warning, which improves the accuracy and timeliness of the warning. Through these technical innovations, the present invention has significant advantages in improving the safety and reliability of power plant equipment operation.
[0117] S4: Dynamically adjust the fault warning threshold based on real-time data and historical data.
[0118] Calculate the initial fault warning threshold using the formula:
[0119] θ0=μ h +k h·σ h
[0120] Among them, θ0 represents the initial warning threshold, μ h represents the mean of historical data, k h represents the adjustment coefficient, σ h Represents the standard deviation of historical data; this method takes into account the fluctuation of historical data and makes the initial warning threshold more adaptable.
[0121] Dynamically adjust the warning threshold, combined with real-time data characteristics, the formula is:
[0122]
[0123] Among them, θ(t) represents the warning threshold after dynamic adjustment, θ0 represents the initial warning threshold, α represents the adjustment coefficient, T represents the time window length, and x i (t) represents the real-time data characteristics at time t, μ r Represents the mean of real-time data features. This method captures data fluctuations and anomalies by performing secondary differential transformation on real-time data features, dynamically adjusts the warning threshold, and improves the accuracy and sensitivity of the warning.
[0124] It should be noted that by dynamically adjusting the threshold of fault warning according to real-time data and historical data, the present invention effectively solves the problem that the warning threshold in the traditional warning system is fixed and cannot adapt to the dynamic changes of data. By calculating the initial warning threshold and dynamically adjusting it in combination with real-time data characteristics, the warning system can more accurately reflect the operating status of the equipment, detect potential faults in a timely manner, and improve the accuracy and sensitivity of the warning. Through these technical innovations, the present invention has significant advantages in improving the safety and reliability of power plant equipment operation.
[0125] S5: The warning information is transmitted to the central monitoring system in real time through the 5G network, and the corresponding fault handling process is triggered.
[0126] Two warning thresholds are set as a low warning threshold θ(t)1 and a high warning threshold θ(t)2, where θ(t)1<θ(t)2.
[0127] According to the fault warning index Ω(t) and the set threshold, the fault warning is divided into three levels:
[0128] When Ω(t)≤θ(t)1, a low-level warning is triggered, the equipment status is checked, the current equipment operating parameters are recorded, and the equipment operating parameters are adjusted; when θ(t)1<Ω(t)≤θ(t)2, an intermediate warning is triggered, and an intermediate warning message is sent to the central monitoring system to remind maintenance personnel to pay attention to the equipment status, and to shut down the equipment for inspection and replacement of components; when Ω(t)>θ(t)2, an advanced warning is triggered, and an advanced warning message is immediately sent to the central monitoring system, triggering the emergency maintenance process, immediately shutting down, starting backup equipment, and convening a professional maintenance team for comprehensive inspection and repair.
[0129] Using the 5G communication network, the packaged warning information data packets are transmitted to the central monitoring system in real time.
[0130] The warning information data package includes the device ID, fault warning index, dynamically adjusted warning threshold and warning level.
[0131] It should be noted that the 5G communication network is used to transmit early warning information to the central monitoring system in real time to ensure the timely transmission of early warning information. For example, when a fault warning occurs in the equipment, the early warning information can be transmitted to the central monitoring system within a few milliseconds, and maintenance personnel can quickly take corresponding measures to prevent the fault from further expanding. By setting low warning thresholds and high warning thresholds, the fault warning is divided into low-level warnings, medium-level warnings and high-level warnings according to the fault warning index, and different maintenance measures are taken respectively. For example, for low-level warnings, simple parameter adjustments can be made; for medium-level warnings, shutdown inspections and component replacements can be carried out; for high-level warnings, emergency shutdowns and comprehensive inspections can be carried out to ensure the safe operation of the equipment.
[0132] The above embodiments also include a power plant fault warning system based on edge computing and 5G communication, specifically:
[0133] Data collection and preprocessing module: By deploying edge computing nodes near key equipment in power plants, the operating data of the equipment can be collected in real time and preprocessed.
[0134] Communication network building module: The communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system.
[0135] Predictive maintenance module: A predictive maintenance model is established through historical data and machine learning algorithms, and deployed on edge computing nodes for real-time data analysis and fault warning.
[0136] Dynamic threshold adjustment module: dynamically adjusts the fault warning threshold based on real-time data and historical data;
[0137] Early warning information transmission and fault handling module: The early warning information is transmitted to the central monitoring system in real time through the 5G network, and the corresponding fault handling process is triggered.
[0138] The computer device may be a server. 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 data cluster data of the power monitoring system. 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 power plant fault early warning method based on edge computing and 5G communication is implemented.
[0139] A person of ordinary skill 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 the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. 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 are not limited to this.
[0140] Example 2
[0141] As an embodiment of the present invention, a power plant fault warning method based on edge computing and 5G communication is provided. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation / comparative experiments.
[0142] In order to verify the effectiveness of the power plant fault warning method based on edge computing and 5G communication of the present invention, a series of experiments were carried out. This experiment aims to verify the effectiveness and advantages of the power plant fault warning system based on edge computing and 5G communication. The test objects are key equipment in the power plant, including boilers, steam turbines, generators, transformers and cooling systems. Each device is equipped with sensors to collect operating data in real time, including temperature (℃), pressure (MPa), vibration (mm / s), current (A), voltage (V) and speed (rpm). The experiment deployed edge computing nodes near these devices, used 5G communication networks to achieve real-time data transmission, and performed fault warnings through the predictive maintenance model established by machine learning algorithms.
[0143] The experiment is divided into two parts: the traditional method and the method of the present invention. In the traditional method, data is transmitted to the central monitoring system through a wired network for processing, and a fixed threshold is used for fault warning. In the method of the present invention, data is transmitted to the central monitoring system in real time through a 5G network, and dynamic threshold adjustment and machine learning algorithms are used for fault warning. The specific experimental data content can be referred to Table 1.
[0144] Data collection and preprocessing:
[0145] Sensors are used to collect temperature, pressure, vibration, current, voltage and speed data of boilers, turbines, generators, transformers and cooling systems in real time.
[0146] The preprocessing steps include data filtering, denoising and normalization.
[0147] 5G communication network construction:
[0148] Deploy 5G base stations near key equipment in power plants to support multi-band communications and ensure high bandwidth and low latency for data transmission.
[0149] Using 5G network slicing technology, the internal communication needs of the power plant are divided into multiple network slices, and each slice is independently allocated bandwidth and priority.
[0150] Predictive maintenance model establishment:
[0151] Collect and analyze historical operation data, use principal component analysis (PCA) to perform feature selection and dimensionality reduction, and extract the most representative feature vectors.
[0152] New features are constructed through feature engineering, and a predictive maintenance model based on random forest is established to perform real-time data analysis and fault warning.
[0153] Dynamic Threshold Adjustment:
[0154] Calculate the initial warning threshold and make dynamic adjustments based on real-time data characteristics to improve the accuracy and sensitivity of the warning.
[0155] Early warning information transmission and fault handling:
[0156] Set low warning thresholds and high warning thresholds, and classify fault warnings into low, medium and high warnings according to the fault warning index, and take different maintenance measures respectively.
[0157] Utilizing the 5G communication network, early warning information data packets are transmitted to the central monitoring system in real time to ensure timely delivery of early warning information.
[0158] Table 1 Experimental data comparison table
[0159]
[0160] By comparing the data of the traditional method and the method of the present invention, it can be clearly seen that the method of the present invention has significant advantages in early warning accuracy and response time.
[0161] The warning accuracy of the traditional method is 74% (boiler) and 79% (turbine). This method relies on fixed thresholds and fails to fully utilize the characteristics of historical data and real-time data, so the warning accuracy is low.
[0162] The warning accuracy of the method of the present invention is 95% (boiler) and 96% (turbine). By dynamically adjusting the warning threshold and applying the machine learning algorithm, the method of the present invention can predict faults more accurately and significantly improve the accuracy of the warning.
[0163] The response time of the traditional method is 498ms (boiler) and 453ms (turbine), respectively. The data needs to be transmitted to the central monitoring system through the wired network for processing, resulting in a long response time.
[0164] The response time of the method of the present invention is 49ms (boiler) and 47ms (steam turbine). By deploying edge computing nodes near the equipment and combining 5G communication networks to achieve real-time data transmission and processing, the method of the present invention significantly shortens the response time and improves the timeliness of fault handling.
[0165] Through the above analysis, it can be seen that the method of the present invention has obvious advantages in early warning accuracy and response time. The use of edge computing and 5G communication technology makes data processing more real-time and accurate. Dynamic adjustment of early warning thresholds and application of machine learning algorithms improve the accuracy and sensitivity of early warnings, thereby realizing efficient real-time monitoring and fault early warning of power plant equipment.
[0166] The method of the present invention effectively solves the problems of insufficient early warning accuracy and long response time in traditional methods. By comparing experimental data, it can be verified that the method of the present invention has significant advantages in practical applications and improves the operating safety and reliability of power plant equipment. The method not only has high early warning accuracy and rapid response capability, but also can dynamically adjust the early warning threshold according to real-time data, making the early warning system more flexible and intelligent, and has high application value and promotion prospects.
[0167] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
Claims
1. A power plant fault early warning method based on edge computing and 5G communication, characterized in that: include: By deploying edge computing nodes near key equipment in power plants, the operating data of the equipment can be collected and pre-processed in real time; A communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system; Establish a predictive maintenance model through historical data and machine learning algorithms, and deploy it on edge computing nodes for real-time data analysis and fault warning; Dynamically adjust the fault warning threshold based on real-time and historical data; The warning information is transmitted to the central monitoring system in real time through the 5G network, triggering the corresponding fault handling process.
2. The power plant fault early warning method based on edge computing and 5G communication according to claim 1, characterized in that: The key equipment of the power plant includes boilers, steam turbines, generators, transformers and cooling systems; The operating data of the acquisition equipment includes temperature, pressure, vibration, current, voltage and rotation speed; The preprocessing includes filtering, denoising, normalization and feature extraction.
3. The power plant fault early warning method based on edge computing and 5G communication as claimed in claim 2, characterized in that: The communication network constructed includes deploying 5G base stations around key equipment in the power plant and in necessary monitoring areas, and configuring multi-band support between base stations; Using 5G network slicing technology, the internal communication needs of the power plant are divided into multiple network slices, and each slice allocates independent bandwidth and priority for specific data flows; Dynamically adjust network resource allocation and automatically optimize bandwidth configuration of each slice based on real-time monitored data traffic; The specific data flows include monitoring data flows, control data flows and warning data flows.
4. The power plant fault early warning method based on edge computing and 5G communication as claimed in claim 3, characterized in that: The automatic optimization of the bandwidth configuration of each slice includes weighted integration of the real-time data traffic of each slice to evaluate its data transmission demand in a specific time period. The formula is: Among them, I i (t) represents slice S i The weighted integral of data flow in time t, α i Represents slice S i The bandwidth weight coefficient, F i (τ) represents the real-time data flow at time τ; τ represents the time variable, which is used for integral calculation; Exponential smoothing is performed on the data traffic change rate to capture the fluctuation of data traffic and consider the impact on bandwidth requirements. The formula is: Among them, E i (t) represents slice S i With other slices j The sum of the interaction effects of ij Represents slice S i With other slices j The interaction coefficient between them; represents the exponential smoothing function of the data flow rate change, γ j represents the exponential smoothing coefficient, Indicates the rate of change of data traffic; M indicates the total number of all slices in the network; The data flow of each slice is normalized and multiplied by complex transformation, and the overall bandwidth requirement of each slice is comprehensively considered. The formula is: Among them, P i (t) represents slice S i The product of normalization and complex transformation; σ k represents the normalization coefficient; ψ k (F k (τ)) represents the transformation function of data flow; The bandwidth allocation calculation formula for each slice is: Among them, B i (t) represents slice S i The bandwidth allocation at time t, B total Indicates the total network bandwidth, I k (t represents the set of weighted integrals of data traffic of all slices within time t, E k (t) represents the sum of the interaction effects of all slices, P k (t) represents the set of product results of normalization and complex transformation of all slices; If B i When (t) = 0, the slice S i No bandwidth is currently allocated; If B i (t) = B total When S i Currently occupying the entire network bandwidth.
5. The power plant fault early warning method based on edge computing and 5G communication as claimed in claim 4, characterized in that: The establishment of the predictive maintenance model includes using principal component analysis to perform feature selection and dimensionality reduction to extract the most representative feature vector; New features are constructed through feature engineering. The formula is: Among them, Φ represents the weighted eigenvector, α i represents the feature weight coefficient, ψ i (x i ) represents the feature transformation function, x i represents the original feature, τ represents the time variable, and m represents the number of features; Establish a predictive maintenance model, the formula is: in, represents the predicted value, N represents the number of decision trees, and f i (x) represents the prediction result of the i-th decision tree; Use historical data to train the model, evaluate the model performance through the validation set, and adjust the model parameters; Deploy the trained model on the edge computing node to perform real-time data analysis and fault warning, monitor the equipment operation status in real time, and generate warning information based on the model prediction results. The formula is: in, represents the prediction result, β j represents the model parameters, γ j represents the exponential smoothing coefficient, represents the rate of change of the weighted eigenvector; Use real-time data for analysis and fault warning. The formula is: Among them, Ω(t) represents the fault warning index, σ k represents the normalization coefficient, ψ k (x k (τ)) represents the real-time feature transformation function, x k (τ) represents the real-time features at time τ, and n represents the number of real-time features.
6. The power plant fault early warning method based on edge computing and 5G communication according to claim 5, characterized in that: The dynamically adjusting the fault warning threshold includes calculating the initial fault warning threshold, the formula is: θ0=μ h +k h ·s h Among them, θ0 represents the initial warning threshold, μ h represents the mean of historical data, k h represents the adjustment coefficient, σ h Indicates the standard deviation of historical data; Dynamically adjust the warning threshold, combined with real-time data characteristics, the formula is: Among them, θ(t) represents the warning threshold after dynamic adjustment, θ0 represents the initial warning threshold, α represents the adjustment coefficient, T represents the time window length, and x i (t) represents the real-time data characteristics at time t, μ r Represents the mean of the real-time data features.
7. The power plant fault early warning method based on edge computing and 5G communication according to claim 6, characterized in that: The fault handling process includes setting two warning thresholds as a low warning threshold θ(t)1 and a high warning threshold θ(t)2, where θ(t)1<θ(t)2; According to the fault warning index Ω(t) and the set threshold, the fault warning is divided into three levels: When Ω(t)≤θ(t)1, a low-level warning is triggered, the equipment status is checked, the current equipment operating parameters are recorded, and the equipment operating parameters are adjusted; When θ(t)1<Ω(t)≤θ(t)2, an intermediate warning is triggered and an intermediate warning message is sent to the central monitoring system to remind maintenance personnel to pay attention to the equipment status and to shut down the equipment for inspection and component replacement. When Ω(t)>θ(t)2, an advanced warning is triggered, and the advanced warning information is immediately sent to the central monitoring system, triggering the emergency maintenance process, immediately shutting down, starting the backup equipment, and calling a professional maintenance team for comprehensive inspection and repair; Using the 5G communication network, the packaged warning information data packets are transmitted to the central monitoring system in real time; The warning information data packet includes a device ID, a fault warning index, a dynamically adjusted warning threshold, and a warning level.
8. A power plant fault early warning system based on edge computing and 5G communication using the method according to any one of claims 1 to 7, characterized in that: Data collection and preprocessing module: By deploying edge computing nodes near key equipment in the power plant, the operating data of the equipment is collected in real time and preprocessed; Communication network building module: A communication network built using 5G base stations and dedicated network slicing technology enables real-time data transmission between edge computing nodes and the central monitoring system; Predictive maintenance module: Establish a predictive maintenance model through historical data and machine learning algorithms, and deploy it on edge computing nodes for real-time data analysis and fault warning; Dynamic threshold adjustment module: dynamically adjusts the fault warning threshold based on real-time data and historical data; Early warning information transmission and fault handling module: The early warning information is transmitted to the central monitoring system in real time through the 5G network, and the corresponding fault handling process is triggered.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.
Citation Information
Cited By
Power distribution network operation and maintenance management method and system, electronic equipment and product
CN119417057A
Fiber core on-line rapid test method and system
CN119865238A
Power distribution network system based on 5G short slice private network and AI and fault diagnosis method
CN119944958A
Fault real-time detection method and device, equipment and medium
CN120044935A
Intelligent monitoring and early warning method and system for urban rail vehicle door based on PHM technology
CN120445300A
