Running inspection method and system based on nuclear power field

Abstract

The invention discloses an operation inspection method and system based on the nuclear power field, and the method comprises the steps: collecting first object data, and carrying out the first preprocessing; and extracting a first key feature from the preprocessed data, and fusing the features. And inputting the fused features into a calculation model, outputting a risk level and giving out an early warning. According to the operation inspection method and system based on the nuclear power field, automatic monitoring of the internal and external environments and the equipment operation state of the nuclear power station is achieved, the requirement for manual inspection is reduced, and the inspection efficiency and safety are improved. An advanced machine learning algorithm and a big data analysis technology are adopted to carry out trend analysis and fault prediction, and the early warning capability for potential problems is improved. And an alarm signal can be automatically triggered, so that the safety monitoring capability is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of operation inspection technology, and in particular to an operation inspection method and system based on the nuclear power field. Background Art

[0002] Energy facility management is a traditional industry. The quality of inspection work is not guaranteed, meter reading is not clear, errors are large, subjective factors have a great influence, the cost of repeated work is high, records are taken with pen and paper, which is not easy to archive, retrieve, or perform statistical analysis. The work time management efficiency is low. In short, the management quality does not meet the requirements of increasingly advanced factory equipment.

[0003] The operation is unreasonable, the energy consumption is high, the consumption of raw materials is large, the equipment runs at full load at all times, the excess energy consumption is difficult to separate and count, leakage is difficult to find, manual adjustment relies on experience, and the equipment cannot accurately supply energy dynamically according to demand.

[0004] In order to manage these devices systematically, domestic and foreign companies and institutions have made various efforts and developed a variety of information technology management systems that use information technology to apply computers and the Internet to equipment management, including inspection management. These management systems have achieved database systems for the purpose of management registration, management platforms with information collection and statistics and command transmission as functions, and inspection records and inspection reports are realized on the management platform. However, the functions realized by the system are only to combine the SCADA information monitoring system and the automated office system. It cannot form an effective operation management plan, nor can it be pushed to the work site for execution in real time. There is no substantial improvement for the fundamental defects such as low operation and maintenance quality, only being able to discover major problems in the equipment, being unable to comprehensively judge the working status of the system, and short equipment life. In particular, decades of traditional problems such as scattered station buildings, difficult management, and difficult personnel on-duty quality management have not been effectively solved. Summary of the invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is that the existing technology cannot form an effective operation management plan, nor can it be pushed to the work site for execution in real time. There is no substantial improvement on the fundamental defects such as low operation and maintenance quality, only being discovered when the equipment has a major problem, being unable to comprehensively judge the working status of the system, and short equipment life. In particular, the traditional problems of decades such as the scattered station buildings, difficult management, and difficult personnel on-duty quality management have not been effectively solved.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for operation inspection based on the nuclear power field, comprising: collecting first object data and performing first preprocessing.

[0008] The first key feature is extracted from the preprocessed data and the features are fused.

[0009] The fused features are input into the computational model, which outputs the risk level and issues a warning.

[0010] As a preferred solution of the nuclear power field operation inspection method described in the present invention, wherein: the first object data is the key data collected by the inspection robot.

[0011] As a preferred solution of the nuclear power field operation inspection method described in the present invention, the first preprocessing includes performing preprocessing to obtain high-quality and high-availability data.

[0012] As a preferred solution of the nuclear power field operation inspection method described in the present invention, wherein: the extraction of the first key feature includes extracting the key feature of the preprocessed data, and the key feature is the feature of the data that affects the first object.

[0013] As a preferred solution of the nuclear power field operation inspection method described in the present invention, the feature fusion includes fusing the key features through weighted averaging to improve data reliability.

[0014] As a preferred solution of the nuclear power field operation inspection method described in the present invention, the first object includes but is not limited to a nuclear power plant, and the first object data includes but is not limited to the operation data of the internal and external environment and equipment of the nuclear power plant.

[0015] The first preprocessing includes but is not limited to data cleaning, denoising and normalization processing.

[0016] The first key characteristics include radiation level, temperature, humidity and vibration key parameters.

[0017] As a preferred solution of the nuclear power operation inspection method described in the present invention, the inputting of the fused features into the calculation model, outputting the risk level and issuing an early warning includes training the extracted features with a machine learning algorithm, establishing a prediction model for equipment status and environmental parameters, outputting the risk level and triggering the corresponding level of alarm action according to different risk levels.

[0018] A nuclear power field operation inspection system, characterized in that: it includes:

[0019] The preprocessing module collects first object data and performs first preprocessing.

[0020] The feature extraction module extracts the first key feature from the preprocessed data and fuses the features.

[0021] The calculation module inputs the fused features into the calculation model, outputs the risk level and issues an early warning.

[0022] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0023] A computer-readable storage medium stores a computer program, which implements the steps of the method described above when executed by a processor.

[0024] The beneficial effects of the present invention are: to realize automatic monitoring of the internal and external environment of nuclear power plants and the operating status of equipment, reduce the need for manual inspections, and improve inspection efficiency and safety. Advanced machine learning algorithms and big data analysis technologies are used for trend analysis and fault prediction to improve the early warning capability of potential problems. Alarm signals are automatically triggered to enhance safety monitoring capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0026] Figure 1 An overall flow chart of an operation inspection method and system based on the nuclear power field is provided for the first embodiment of the present invention.

[0027] Figure 2 An architecture diagram of an operation inspection method and system based on the nuclear power field provided for the first embodiment of the present invention. DETAILED DESCRIPTION

[0028] 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 drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. 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.

[0029] Example 1, reference Figure 1-2 , is an embodiment of the present invention, and provides an operation inspection method based on the nuclear power field, comprising:

[0030] S1: Collect first object data and perform first preprocessing.

[0031] In the present invention, a patrol robot is used to collect operating data of the internal and external environment and equipment of a nuclear power plant. The first object is the nuclear power plant, the first object data is the operating data of the internal and external environment and equipment of the nuclear power plant, and the first preprocessing is data cleaning, denoising and normalization.

[0032] The first object includes but is not limited to nuclear power plants, and may also be chemical plants and urban smart transportation systems. The corresponding first object data includes but is not limited to the operating data of the internal and external environment and equipment of the nuclear power plant, and may also be chemical plant operation monitoring data and traffic operation and environmental monitoring data.

[0033] In an optional embodiment of the present invention, the first object is a chemical plant, and the first object data is chemical plant operation monitoring data. The chemical plant operation monitoring data includes:

[0034] Internal equipment operating parameters: including reactor temperature and pressure, pipeline flow, valve status, tank liquid level, etc.

[0035] External environment monitoring data: such as surrounding air quality, gas leakage concentration, and toxic gas diffusion range.

[0036] Energy usage: such as electricity and gas consumption monitoring.

[0037] In an optional embodiment of the present invention, the first object is an urban intelligent transportation system, and the first object data is traffic operation and environmental monitoring data. The traffic operation and environmental monitoring data include:

[0038] Traffic flow data: such as real-time traffic volume, road congestion level, and traffic light configuration information.

[0039] Environmental monitoring data: such as road dust concentration, noise decibel value, weather conditions (rain, snow, temperature and humidity).

[0040] Equipment status data: including traffic light control systems, camera working status, and road monitoring sensor operating status.

[0041] It should be noted that in the present invention, when the first object is a nuclear power plant, the first preprocessing includes but is not limited to data cleaning, denoising and normalization processing, and can also be feature enhancement processing, anomaly detection and filtering, data difference and reconstruction, and multimodal data synchronization and fusion.

[0042] Feature enhancement is a means of deep mining and expansion of original data. For example:

[0043] Time series feature enhancement: Extract change trend features through the sliding window method to analyze abnormal fluctuations in temperature or radiation levels.

[0044] Spatial feature enhancement: Based on the distribution of monitoring points, the distance characteristics between devices or the correlation of environmental data in the area are calculated.

[0045] This method is suitable for detecting hidden faults in equipment that runs for a long time or for monitoring potential threats from environmental changes.

[0046] In the inspection data of nuclear power plants, abnormal data may be caused by sensor failure or external interference. Anomaly detection includes:

[0047] Statistical methods: Calculate the mean and standard deviation, and mark the data points that exceed three times the standard deviation.

[0048] Machine learning methods: Use algorithms such as isolation forests or autoencoders to mark outliers with high confidence.

[0049] Detected abnormal data can be selected to be eliminated or recorded as special events for subsequent analysis.

[0050] In the inspection data, there may be missing data due to equipment or network interruption. Data interpolation methods can fill these missing values:

[0051] Linear interpolation: suitable for parameters with stable trend changes (such as temperature).

[0052] Time Series Forecast Interpolation: Predict the values ​​of missing points by using time series models (such as ARIMA, LSTM).

[0053] This technology ensures the data integrity of subsequent processing links and improves the computational reliability of the model.

[0054] In actual inspections, data of different modalities may be collected, such as image data and sensor data. These data need to be synchronized and fused to achieve unified analysis:

[0055] Time synchronization: Align multimodal data based on timestamps.

[0056] Feature fusion: Fusion of texture features extracted from image data with sensor data to generate richer input features.

[0057] This pre-processing technology can improve the comprehensiveness and accuracy of risk assessment.

[0058] S2: Extract the first key feature from the preprocessed data and fuse the features.

[0059] Extract key features from the preprocessed data, including radiation level, temperature, humidity, and vibration key parameters.

[0060] Different types of sensor data are fused to improve the accuracy and reliability of detection. The weighted average method is used to improve the reliability of data. The expression is:

[0061]

[0062] In the formula, x fused represents the weighted average, w i represents the i-th weight, x i represents the i-th data point.

[0063] In an optional embodiment of the present invention, the first object is a chemical plant. In the operation monitoring of the chemical plant, the key features extracted include:

[0064] Temperature and pressure characteristics: Monitor the temperature and pressure changes of the reactor to ensure that the equipment operates within a safe range.

[0065] Gas concentration characteristics: Detect the concentration of flammable or toxic gases (such as H2S, CO) in key areas.

[0066] Equipment vibration characteristics: Evaluate the health status of mechanical equipment such as pumps and compressors through vibration sensors.

[0067] Level and flow characteristics: Monitor tank level and pipeline flow changes to avoid exceeding design limits.

[0068] In an optional embodiment of the present invention, the first object is an urban intelligent transportation system. In the intelligent transportation system, the extracted key features include:

[0069] Traffic flow characteristics: record vehicle flow and road congestion index, and analyze traffic flow dynamics.

[0070] Environmental characteristics: Monitor road dust concentration and noise decibel value, and evaluate the impact of traffic on the environment.

[0071] Equipment operation characteristics: Monitor the working status of traffic lights to ensure the normal operation of the traffic control system.

[0072] Meteorological characteristics: Real-time recording of meteorological parameters such as rainfall, humidity, wind speed, etc., and prediction of the impact of severe weather on traffic.

[0073] In the present invention, feature fusion can select weighted average method, or the following method:

[0074] Multimodal data fusion:

[0075] A deep learning model is used to embed different types of features into a unified high-dimensional space to generate more expressive comprehensive features.

[0076] Rules Engine Fusion:

[0077] Use predefined rules to conditionally filter and combine features, such as dynamically adjusting the weight of vibration or radiation data based on environmental parameters such as humidity or temperature.

[0078] Bayesian Ensemble:

[0079] The Bayesian inference model is used to weight different features through probability distribution, which is particularly suitable for scenarios where feature data is uncertain.

[0080] Principal Component Analysis (PCA):

[0081] Perform dimensionality reduction on multi-dimensional feature data to retain the main change characteristics while reducing the interference of redundant data on the analysis.

[0082] S3: Input the fused features into the computing model, output the risk level and issue an early warning.

[0083] The extracted features are trained using a machine learning algorithm to establish a prediction model for equipment status and environmental parameters. The expression is:

[0084]

[0085] Where f is the model function and θ is the model parameter.

[0086] Big data technology is used to store, query, analyze and mine the large amount of data collected by the inspection robot to discover potential rules and trends. According to the changing trend of equipment operation, the parameters and structure of the analysis model are adjusted in real time to improve the accuracy of fault prediction. The expression is:

[0087]

[0088] Where η is the learning rate and L is the loss function.

[0089] Output risk level, and trigger corresponding level of alarm response according to different risk levels.

[0090] The computer device may be a server. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via 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 a 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 via a network connection. When the computer program is executed by the processor, a method is implemented.

[0091] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method 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.

[0092] Example 2 is an embodiment of the present invention, which provides an operation inspection method and system based on the nuclear power field. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0093] In order to verify the effectiveness of the method of the present invention in nuclear power operation inspection, a comparative experiment was conducted with the prior art. The test subjects include two methods: the existing manual data processing method (hereinafter referred to as "the prior art") and the method of the present invention. The test was carried out in a simulated nuclear power plant environment, and the inspection robot was responsible for collecting the operating data of the equipment and environment, including radiation levels, temperature, humidity and vibration. The data comes from sensors and environmental monitors in 10 key areas of the nuclear power plant.

[0094] Existing technology implementation steps:

[0095] Data collection: Manually record the raw data provided by the inspection robot and enter it manually for analysis using EXCEL.

[0096] Data processing: Data cleaning and outlier removal are done manually.

[0097] Feature extraction: Only simple means and extreme values ​​are recorded, lacking complex feature extraction and fusion.

[0098] Risk assessment: The threshold is set by fixed experience value and cannot be adjusted dynamically.

[0099] The method of the present invention is implemented in the following steps:

[0100] Data collection and preprocessing: The inspection robot collects radiation, temperature, humidity and vibration data and automatically performs preprocessing, including denoising, normalization and anomaly detection.

[0101] Anomaly detection uses the isolation forest algorithm to remove noise data.

[0102] Missing data are interpolated through the ARIMA model to ensure completeness.

[0103] Key feature extraction and fusion: Extract key features such as temperature fluctuation, humidity change rate, and vibration root mean square value from the preprocessed data and fuse them using the weighted average method.

[0104] Weights are dynamically adjusted based on historical operating data to enhance data reliability.

[0105] Risk level prediction and early warning: The fused features are input into the risk assessment model based on random forests, and the risk level (low, medium, high) is output. Different levels trigger corresponding alarm actions, such as visual signal alarm or automatic shutdown signal.

[0106] The experimental results are shown in Table 1.

[0107] Table 1 Experimental results

[0108]

[0109]

[0110] The experimental results show that the method of the present invention is significantly better than the existing technology in multiple key performance indicators:

[0111] The efficiency of automated preprocessing in the method of the present invention is significantly improved, and the cleaning time is shortened to 30 seconds, which is only 1 / 6 of the existing technology. At the same time, the isolation forest algorithm increases the accuracy of anomaly detection to 96%, avoiding the problems of manual misjudgment and missed judgment.

[0112] Through automated feature extraction and weighted fusion technology, feature extraction time is shortened to 60 seconds and accuracy is improved. Existing technologies can only extract simple means, ignore complex features, and cannot accurately predict risk levels.

[0113] The risk assessment model based on machine learning has achieved a prediction accuracy of 92%, 22 percentage points higher than the existing technology. At the same time, the warning response time has been reduced from 500ms to 120ms, greatly improving the system's response speed to potential problems.

[0114] The present invention improves data integrity to 99% through abnormal data filling and dynamic feature enhancement technology, ensuring the accuracy of subsequent analysis. In addition, the underreporting rate of risk events is reduced to 5%, which is 75% less than the existing technology, further enhancing security.

[0115] Embodiment 3 is an embodiment of the present invention, including an operation inspection system based on the nuclear power field, specifically:

[0116] A preprocessing module, collecting first object data and performing first preprocessing;

[0117] The feature extraction module extracts the first key feature from the preprocessed data and fuses the features;

[0118] The calculation module inputs the fused features into the calculation model, outputs the risk level and issues an early warning.

[0119] 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 nuclear power operation inspection method, characterized in that: include: Collecting first object data and performing first preprocessing; Extract the first key feature from the preprocessed data and fuse the features; The fused features are input into the computational model, which outputs the risk level and issues a warning.

2. The nuclear power field operation inspection method according to claim 1, characterized in that: The first object data is key data collected by the inspection robot.

3. The nuclear power field operation inspection method according to claim 2, characterized in that: The first preprocessing includes performing preprocessing to obtain data with high quality and high availability.

4. The nuclear power field operation inspection method according to claim 3, characterized in that: The extracting the first key feature includes extracting the key feature from the preprocessed data, where the key feature is a feature of the data that affects the first object.

5. The nuclear power field operation inspection method according to claim 4, characterized in that: The feature fusion includes fusing the key features by weighted averaging to improve data reliability.

6. The nuclear power field operation inspection method according to claim 5, characterized in that: The first object includes but is not limited to a nuclear power plant, and the first object data includes but is not limited to operating data of the internal and external environment and equipment of the nuclear power plant; The first preprocessing includes but is not limited to data cleaning, denoising and normalization processing; The first key characteristics include radiation level, temperature, humidity and vibration key parameters.

7. The nuclear power field operation inspection method according to claim 6, characterized in that: The step of inputting the fused features into a computing model, outputting risk levels and issuing warnings includes training the extracted features using a machine learning algorithm, establishing a prediction model for equipment status and environmental parameters, outputting risk levels and triggering alarm actions of corresponding levels according to different risk levels.

8. A nuclear power field operation inspection system using the method according to any one of claims 1 to 7, characterized in that: A preprocessing module, collecting first object data and performing first preprocessing; The feature extraction module extracts the first key feature from the preprocessed data and fuses the features; The calculation module inputs the fused features into the calculation model, outputs the risk level and issues an early warning.

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.

Images