A system and method for identifying unsafe factors in the process of new energy infrastructure construction

Abstract

The present invention discloses a system and method for identifying unsafe factors in the process of new energy infrastructure construction. The specific technical solution is as follows: constructing a feature database of unsafe factors in different construction stages and different accident types, calculating the similarity between the unsafe factor feature vectors collected in real time and the unsafe factor feature vectors in the feature database, and when the calculated similarity exceeds a set similarity threshold, preliminarily identifying the current unsafe factor, and then analyzing the causal relationship, temporal sequence, and spatial relationship between the current unsafe factor and other unsafe factors associated with it, constructing a factor association network model, determining the unsafe factor chain, and mining potential unsafe factors. By analyzing the mutual influence between the current unsafe factor and the potential unsafe factor and deeply mining the complex relationship between the data, it can help staff fully understand the root causes and development trends of unsafe factors and formulate more effective response measures.

Description

Technical Field

[0001] The present invention relates to the field of identifying unsafe factors in infrastructure construction processes, and in particular to a system and method for identifying unsafe factors in new energy infrastructure construction processes. Background Art

[0002] Against the backdrop of the global push to develop clean energy, the scale of new energy infrastructure continues to expand, encompassing a wide range of sectors, including wind farm construction, photovoltaic power stations, and energy storage facilities. These infrastructure projects often have long construction periods, complex working environments, and involve a large number of new equipment and technologies. During construction, they face various unsafe factors, such as equipment failure, improper operation, and environmental fluctuations. These factors can not only lead to construction accidents but also affect project progress and economic benefits. Therefore, effectively identifying unsafe factors during new energy infrastructure construction has become a key requirement for ensuring the healthy development of the new energy industry.

[0003] Currently, existing technologies for identifying unsafe factors in new energy infrastructure suffer from numerous shortcomings. Traditional manual inspections rely on the experience and sense of responsibility of personnel, leading to significant subjective bias, low efficiency, and an inability to monitor in real time. This makes it difficult to fully capture complex and ever-changing unsafe factors. For example, in the construction of large wind farms, manual inspections are unable to promptly identify potential hazards during high-altitude operations, nor can they detect early-stage failures within equipment. Furthermore, some sensor-based monitoring technologies lack scientific planning for sensor layout, resulting in incomplete data collection and susceptibility to environmental interference, leading to insufficient data accuracy. Furthermore, existing identification methods are relatively simplistic in data processing and analysis, often employing simple threshold judgments or traditional statistical analysis. These methods are unable to deeply explore complex relationships between data, making it difficult to identify potential chains of unsafe factors and providing inaccurate assessments of their severity.

[0004] Furthermore, existing unsafe factor identification systems lack dynamic updating and adaptive capabilities. With the continuous development of new energy infrastructure technologies, new types of unsafe factors continue to emerge. However, the feature libraries of existing systems cannot be updated in a timely manner, and the identification models struggle to adapt to new situations. This leads to a gradual decline in the timeliness and accuracy of identification, making it unable to meet the growing safety needs of new energy infrastructure. More advanced and intelligent identification systems and methods are urgently needed to address these issues. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides a system and method for identifying unsafe factors in the process of new energy infrastructure construction.

[0006] A method for identifying unsafe factors in the process of new energy infrastructure construction builds a feature database of unsafe factors in different construction stages and different accident types; collects field data from the infrastructure site in real time and extracts the corresponding unsafe factor feature vectors; uses a matching algorithm based on similarity measurement to calculate the similarity between the collected unsafe factor feature vectors and the unsafe factor feature vectors in the feature database, and determines the current unsafe factor when the calculated similarity is greater than a set similarity threshold; analyzes the causal relationship, temporal sequence and spatial relationship between the current unsafe factor and other related factors through a factor association network model, determines the unsafe factor chain, thereby exploring potential unsafe factors and comprehensively assessing the severity of the current unsafe factor and potential unsafe factors.

[0007] Preferably, the identification method comprises the following steps:

[0008] Step S1: Analyze unsafe factors in different construction stages and different accident types, construct unsafe factor feature vectors in different construction stages and different accident types, and thus construct a feature database of unsafe factors;

[0009] Step S2: With the goal of maximizing coverage of the infrastructure area and minimizing the number of sensors, various sensors are deployed at the infrastructure site to collect real-time data and extract unsafe factor feature vectors.

[0010] Step S3: Match the unsafe factor feature vectors collected in real time with the unsafe factor feature vectors in the feature database, and use a matching algorithm based on similarity measurement to calculate the similarity between the two. When the calculated similarity is greater than a set similarity threshold, it is preliminarily determined that the corresponding unsafe factor exists and is recorded as the current unsafe factor.

[0011] Step S4: Analyze other factors associated with the current unsafe factor and construct a factor association network model. Consider the causal relationship, temporal sequence, and spatial position relationship between different physical quantities to determine the unsafe factor chain and explore potential unsafe factors.

[0012] Step S5: Use the fusion model of fuzzy hierarchical analysis and grey relational analysis to comprehensively evaluate the severity of current unsafe factors and potential unsafe factors.

[0013] Preferably, the step S1 includes:

[0014] Step S11: Analyze unsafe factors in different construction stages and accident types based on historical accident cases, industry standards, and expert experience, and determine the corresponding relationship between characteristics and unsafe factors;

[0015] Step S12: Use the improved hierarchical clustering algorithm to preliminarily cluster the feature vectors of the infrastructure site. In the clustering process, the spatial correlation weights between features are introduced. The improved hierarchical clustering distance calculation formula is:

[0016] ,

[0017] in, and are two eigenvectors, and are vectors and No. The first dimension eigenvalues, For the The spatial correlation weight of the dimension, is the dimension of the feature vector, is the number of features under each dimension;

[0018] Step S13: Input the clustering results in step S12 into the decision tree model for further classification. The decision tree model splits the nodes according to the information gain ratio of the features, classifies and organizes the features, and improves the feature database of unsafe factors.

[0019] Step S14: Use the incremental learning algorithm to update the feature database of unsafe factors, and promptly incorporate the newly emerged unsafe factor features into the feature database. The weight update formula based on gradient descent is used in the incremental learning process:

[0020] ,

[0021] in, for The weight vector of the moment feature library model, is the learning rate, is the loss function About the weight vector gradient.

[0022] Preferably, the step S2 includes:

[0023] Step S21: Arrange various sensors at the infrastructure construction site based on the spatial optimization layout model. The layout model formula is as follows:

[0024] ,

[0025] Constraints: ,

[0026] in, Indicates whether it is at the location Arrange sensors, Indicates the location The cost of deploying sensors, Indicates location Sensor pair area coverage capabilities, Indicates location The set of areas that can be covered by the sensor, Represents the set of all areas where the infrastructure site is divided;

[0027] Step S22: Determine the sampling frequency of each sensor type based on the sensor type and the dynamic characteristics of the infrastructure construction site, and collect data on the physical parameters and real-time scenes at the infrastructure construction site at the specific sampling frequency;

[0028] Step S23: extracting unsafe factor feature vectors in the infrastructure construction site in real time, wherein the unsafe factor feature vectors include time domain features, frequency domain features, and abstract features.

[0029] Preferably, the similarity calculation formula in step S3 is:

[0030] ,

[0031] in, and are the real-time feature vector and the feature vector in the feature database respectively, and are vectors and No. and subvectors at each time point, is the kernel function, is the time alignment factor determined by the dynamic time warping algorithm, and are vectors and The number of time points.

[0032] Preferably, step S4 includes:

[0033] Step S41: Analyze other factors associated with the current unsafe factor and map the other factors as variables to the Bayesian network nodes;

[0034] Step S42: Determine directed edges between nodes based on the causal relationships, temporal sequences, and spatial position relationships between different physical quantities, draw a factor association network, and adjust the positions of nodes and edge layouts based on the actual logical relationships in the infrastructure construction process;

[0035] Step S43: For each node, estimate its conditional probability table based on historical data and expert experience, and analyze the probability of occurrence of each factor in different situations;

[0036] Step S44: Sort the occurrence probability of each factor in different situations from large to small to mine potential unsafe factors.

[0037] Preferably, in step S5, the weights of different unsafe factors and their associated factors are determined by fuzzy hierarchical analysis, and the fuzzy judgment matrix elements are The calculation is based on expert evaluation of factors and factors The fuzzy evaluation of relative importance is then performed, and then the correlation between the real-time unsafe factors and the standard risk pattern is calculated through grey correlation analysis. The grey correlation calculation formula is:

[0038] ,

[0039] in, Reference sequence Compare with sequence The grey relational degree of and The reference sequence and the ith comparison sequence are The value under the indicator, is the resolution coefficient.

[0040] Correspondingly, a system for identifying unsafe factors in the process of new energy infrastructure construction includes a multi-source heterogeneous data aggregation and fusion unit, an adaptive data cleaning and feature extraction unit, a dynamic feature library construction and maintenance unit, a similarity matching and preliminary judgment unit, a multi-factor association mining and analysis unit, and a comprehensive evaluation and visualization presentation unit; the multi-source heterogeneous data aggregation and fusion unit is used to receive and integrate physical parameters and image multi-source data from different sensors; the adaptive data cleaning and feature extraction unit is used to clean the aggregated data and extract features based on a specific algorithm; the dynamic feature library construction and maintenance unit is responsible for constructing and dynamically updating the unsafe factor feature library based on historical cases; the similarity matching and preliminary judgment unit makes a preliminary judgment of the unsafe factors through a similarity algorithm; the multi-factor association mining and analysis unit uses a constructed factor association network model to analyze the associated factors of the unsafe factors; the comprehensive evaluation and visualization presentation unit uses a comprehensive evaluation algorithm to evaluate the severity of the unsafe factors.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. By building a feature database of unsafe factors in different construction stages and accident types, the similarity between the unsafe factor feature vectors collected in real time and the unsafe factor feature vectors in the feature database is calculated. When the calculated similarity exceeds the set similarity threshold, the current unsafe factor is preliminarily identified. The causal relationship, temporal sequence, and spatial relationship between the current unsafe factor and other related unsafe factors are then analyzed. A factor association network model is constructed to determine the unsafe factor chain, thereby mining potential unsafe factors. By analyzing the mutual influence between current unsafe factors and potential unsafe factors and deeply exploring the complex relationships between data, the severity of current and potential unsafe factors can be accurately assessed, which can help staff fully understand the root causes and development trends of unsafe factors and formulate more effective response measures.

[0043] 2. The entire identification system realizes efficient processing of the whole process of unsafe factors of new energy infrastructure from data collection, processing and analysis to comprehensive evaluation output through the coordinated operation of six major units, including multi-source heterogeneous data aggregation and fusion unit, adaptive data cleaning and feature extraction unit. It can identify potential risks in a timely and accurate manner, provide strong guarantees for the safe, stable and efficient construction of new energy infrastructure, effectively reduce the probability of accidents, reduce economic losses and casualties, and promote the healthy development of the new energy infrastructure industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of the method steps of the present invention;

[0045] Figure 2 It is a diagram of the system unit composition of the present invention. DETAILED DESCRIPTION

[0046] It should be noted that, unless there is a conflict, the embodiments in this application and the features described in the embodiments can be combined with each other. The application is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] In this embodiment, the wind farm infrastructure and operation and maintenance scenario is used as an example to illustrate the specific implementation process of the identification method in this application. Figure 1 As shown, the present invention discloses a method for identifying unsafe factors in the process of new energy infrastructure construction, the method comprising:

[0048] Step S1: Constructing a feature database of unsafe factors: Based on historical accident cases, industry standards, and expert experience in the new energy infrastructure sector, we analyze unsafe factors at different construction stages and accident types, construct feature vectors for these factors, and thus construct a feature database of unsafe factors. A clear mapping relationship is established between the features in the feature database and the unsafe factors, and a dynamic update mechanism is provided to adapt to emerging unsafe factor characteristics.

[0049] Specifically, building a feature library of unsafe factors is an important foundation for the identification system. Extracted feature vectors are classified and organized based on historical accident cases, industry standards, and expert experience in the field of new energy infrastructure. For example, in the construction of energy storage power stations, reference is made to previous battery thermal runaway accident cases to categorize features such as abnormally high battery temperature, excessive voltage fluctuations, and changes in internal gas concentrations. In conjunction with industry standards, relevant features such as construction equipment operating specifications and safe distances are incorporated into the feature library. Furthermore, through expert experience, additional features that are difficult to quantify but have important indicative significance are supplemented, such as the fatigue status of construction workers.

[0050] With the development of new energy infrastructure technology and the emergence of new unsafe factors, such as the new risks brought about by the application of new energy storage materials, the feature library can be updated in a timely manner to incorporate new features and unsafe factors, ensuring that the feature library always adapts to the actual needs of new energy infrastructure and provides a reliable basis for accurately identifying unsafe factors.

[0051] Step S2: Real-time collection of on-site data and extraction of unsafe factor feature vectors: Utilizing various sensors distributed at the new energy infrastructure site, data is collected on the physical parameters and real-time scenarios in the infrastructure environment at a specific sampling frequency, and unsafe factor features are extracted. The sampling frequency is adaptively adjusted according to different sensor types and the dynamic characteristics of the infrastructure scenario. The collected data includes time, spatial dimensions, and multi-physical quantity characteristics.

[0052] Specifically, in wind farm infrastructure construction and operation and maintenance projects, the construction environment is complex and ever-changing, encompassing various scenarios such as high altitude, underground, high temperature, and high humidity, posing numerous potential safety risks. To fully capture this risk information, data collection requires the use of various sensors distributed throughout the construction site, including temperature sensors, pressure sensors, vibration sensors, and image acquisition equipment. For example, temperature sensors can monitor the temperature of welds in real time to prevent material degradation due to overheating; pressure sensors installed on lifting equipment monitor the forces applied during the lifting process to prevent accidents caused by overloading; vibration sensors are deployed in the foundation construction area to detect foundation vibration and prevent problems such as foundation subsidence; and image acquisition equipment is used to monitor construction personnel's operating procedures and changes in the on-site environment.

[0053] It should be noted that in the specific implementation process, the image data collected by the image acquisition device can be preprocessed using an object detection model based on a fusion of a multi-scale convolutional neural network and an attention mechanism. This model performs convolution operations on images at different scales to extract multi-scale features, while using the attention mechanism to enhance focus on key areas. The attention mechanism calculation formula is:

[0054] ;

[0055] in, For the input image The attention weight matrix, is the activation function, and are the weight matrix and bias vector, is the number of network layers. This model can more effectively extract target information related to unsafe factors from image data, assisting in the subsequent identification of unsafe factors.

[0056] Step S3, preliminarily determine the unsafe factors: match the unsafe factor feature vectors collected in real time with the unsafe factor feature vectors in the feature database, and use a matching algorithm based on similarity measurement to calculate the similarity between the two. When the calculated similarity is greater than the set similarity threshold, it is preliminarily determined that the corresponding unsafe factor exists and is recorded as the current unsafe factor.

[0057] Specifically, a dynamic time warping similarity model based on kernel functions is used. This model takes into account the differences in the scalability of different types of unsafe factors on a time scale and uses kernel functions to map feature vectors to a high-dimensional space to enhance the distinguishability of features. The similarity calculation formula is:

[0058] ;

[0059] in, and are the real-time feature vector and the feature vector in the feature library respectively, and are vectors and No. and subvectors at each time point, is the kernel function, is the time alignment factor determined by the dynamic time warping algorithm, and are vectors and This model can more accurately calculate the similarity between feature vectors and improve the accuracy of preliminary identification of unsafe factors.

[0060] Specifically, in practice, the similarity threshold isn't fixed; instead, it's optimized based on the risk level and misjudgment cost of different types of unsafe factors. For high-risk unsafe factors, the threshold is set relatively low to ensure timely detection of potential risks. For low-risk factors prone to misjudgment, the threshold is raised appropriately to reduce false positives.

[0061] Step S4: Identify potential unsafe factors: Analyze other factors associated with the current unsafe factor and construct a factor association network model. This model considers the causal relationships, temporal sequence, and spatial relationships between different physical quantities to identify unsafe factor chains and identify potential unsafe factors. This factor association network model is constructed based on Bayesian network theory, and its parameters are adjusted based on the actual logical relationships in the infrastructure construction process.

[0062] Specifically, for the unsafe factors that have been initially identified, further analysis is conducted on other factors associated with them. For example, in the operation and maintenance infrastructure of a new energy wind farm, if abnormal vibration of the wind turbine blades is initially identified, a factor association network model is constructed to consider the causal relationship, time sequence, and spatial position relationship between different physical quantities to explore the potential chain of unsafe factors. It may be found that the abnormal blade vibration is due to unstable speed caused by a gearbox failure, which in turn causes uneven force on the blades. At the same time, the drastic changes in the ambient wind speed also exacerbate this vibration. This factor association network model is constructed based on Bayesian network theory and is combined with the actual logical relationship in the infrastructure process for parameter adjustment. It can accurately analyze the mutual influence between various factors, help staff fully understand the root causes and development trends of unsafe factors, and formulate more effective response measures.

[0063] Step S5: Assess the severity of unsafe factors: A model combining fuzzy hierarchical analysis and grey relational analysis is used to comprehensively assess the severity of current and potential unsafe factors. The assessment results are output in an intuitive, visual format, including risk level identification, unsafe factor location, and a schematic representation of the potential impact range. A detailed assessment report is also generated, including key data from the identification process and the analytical basis.

[0064] The model based on the fusion of fuzzy hierarchical analysis and grey relational analysis is as follows: first, fuzzy hierarchical analysis is used to determine the weights of different unsafe factors and their associated factors, and the fuzzy judgment matrix elements are The calculation is based on expert evaluation of factors and factors Fuzzy evaluation of relative importance. Then, the correlation between real-time unsafe factors and standard risk patterns is calculated through grey correlation analysis. The grey correlation calculation formula is:

[0065] ,

[0066] in, Reference sequence Compare with sequence The grey relational degree of and The reference sequence and the ith comparison sequence are The value under the indicator, is the resolution coefficient. This fusion model can more comprehensively and accurately assess the severity of unsafe factors.

[0067] Specifically, a comprehensive assessment algorithm is used to evaluate the severity of unsafe factors during infrastructure construction, taking into account preliminary identification results and multi-factor correlation analysis. For example, in the construction of a large-scale solar photovoltaic power station, if a deviation in the installation angle of photovoltaic panels is identified, and correlation analysis reveals that this can lead to reduced power generation efficiency, local overheating, and accelerated equipment aging, the comprehensive assessment algorithm will combine these factors to assess the severity of the unsafe factor from multiple perspectives, including safety risk, economic loss, and impact on project schedule.

[0068] Assessment results are output in an intuitive, visual format. For example, on the monitoring system interface, risk levels are indicated by color codes (red for high risk, yellow for medium risk, and blue for low risk). The specific locations of unsafe factors are marked on infrastructure maps, and a graphical representation of the potential impact range is provided. A detailed assessment report is also generated, containing key data from the identification process (such as collected raw data and processed feature vectors) and analysis basis (such as similarity calculation results and factor correlation analysis). This helps staff understand the specific circumstances of unsafe factors and take timely measures to ensure the safe and smooth progress of new energy infrastructure projects.

[0069] In a further embodiment, the feature classification in step S1 is based on an improved hierarchical clustering and decision tree fusion model, including the following process:

[0070] Step S11: Analyze unsafe factors in different construction stages and accident types based on historical accident cases, industry standards, and expert experience, and determine the corresponding relationship between characteristics and unsafe factors;

[0071] Step S12: Use the improved hierarchical clustering algorithm to preliminarily cluster the feature vectors of the infrastructure site. In the clustering process, the spatial correlation weights between features are introduced. The improved hierarchical clustering distance calculation formula is:

[0072] ,

[0073] in, and are two eigenvectors, and are vectors and No. The first dimension eigenvalues, For the The spatial correlation weight of the dimension, is the dimension of the feature vector, is the number of features under each dimension;

[0074] Step S13: Input the clustering results in step S12 into the decision tree model for further classification. The decision tree model splits the nodes according to the information gain ratio of the features and classifies and organizes the features, thereby classifying and organizing the features more accurately and improving the feature database of unsafe factors.

[0075] Step S14: Use the incremental learning algorithm to update the feature database of unsafe factors, and promptly incorporate the newly emerged unsafe factor features into the feature database. The weight update formula based on gradient descent is used in the incremental learning process:

[0076] ,

[0077] in, for The weight vector of the moment feature library model, is the learning rate, is the loss function About the weight vector gradient.

[0078] In a further embodiment, step S2 includes the following process:

[0079] Step S21: Arrange various sensors at the infrastructure construction site based on a spatial optimization layout model. This model aims to maximize coverage of the infrastructure area while minimizing the number of sensors. It considers factors such as the topography of the infrastructure site, equipment distribution, and areas with high incidence of potential unsafe factors. The sensor layout determined by this model can more efficiently collect comprehensive and targeted data. The layout model formula is as follows:

[0080] ,

[0081] Constraints: ,

[0082] in, Indicates whether it is at the location Arrange sensors, Indicates the location The cost of deploying sensors, Indicates location Sensor pair area coverage capabilities, Indicates location The set of areas that can be covered by the sensor, Represents the set of all areas where the infrastructure site is divided;

[0083] Step S22: Determine the sampling frequency of each sensor based on the sensor type and the dynamic characteristics of the infrastructure construction site, and collect data on the physical parameters and real-time scenes at the infrastructure construction site at a specific sampling frequency. For example, in areas where temperature changes slowly, the sampling frequency of the temperature sensor can be set lower. In key areas where equipment is operating, the sampling frequency of the vibration sensor needs to be significantly increased to capture subtle changes in abnormal equipment vibration, ensuring that the collected data can accurately reflect the actual conditions of the infrastructure construction site and provide rich and effective raw data for subsequent analysis.

[0084] Step S23: The collected raw data is passed through a data cleaning algorithm to remove outliers caused by sensor failure, interference, and other factors. Then, a feature extraction algorithm based on the combination of time-frequency analysis and deep learning is used to extract feature vectors that can characterize potential unsafe factors in the infrastructure construction process from the cleaned data. The feature vectors include time domain features, frequency domain features, and abstract features learned through deep neural networks.

[0085] In the specific implementation process, raw data cleaning can adopt a data anomaly detection model based on dynamic threshold correction, which dynamically adjusts the anomaly detection threshold according to the time series characteristics of the data and the change trend of adjacent data points. The model formula is:

[0086] ,

[0087] in, For the The anomaly detection threshold for data points, For the front The mean of the data points, For the front The standard deviation of the data points, This coefficient is adaptively adjusted based on data fluctuations. This model can more accurately identify and remove outliers caused by sensor failures or environmental interference, providing a more reliable data foundation for subsequent feature extraction.

[0088] In some implementations, after removing outliers, a feature extraction algorithm based on a combination of time-frequency analysis and deep learning is used to extract feature vectors from the cleaned data that can characterize potential unsafe factors in the infrastructure construction process. In the wind turbine blade manufacturing process, time-frequency analysis can convert vibration data from the time domain to the frequency domain, clearly presenting the vibration characteristics at different frequencies, while deep learning algorithms (such as convolutional neural networks) can automatically learn complex patterns and abstract features in the data, extracting key features of unsafe factors such as blade cracks and imbalance from large amounts of data. These feature vectors contain time domain features (such as mean and variance), frequency domain features (such as frequency peaks), and abstract features learned through deep neural networks (such as characteristic patterns of blade structural damage), laying a solid foundation for the subsequent identification of unsafe factors.

[0089] In a further embodiment, step S4 includes the following process:

[0090] Step S41: Analyze other factors associated with the current unsafe factor and map the other factors as variables to the Bayesian network nodes;

[0091] Step S42: Determine directed edges between nodes based on the causal relationships, temporal sequences, and spatial position relationships between different physical quantities, draw a factor association network, and adjust the positions of nodes and edge layouts based on the actual logical relationships in the infrastructure construction process;

[0092] Step S43: For each node, estimate its conditional probability table based on historical data and expert experience, and analyze the probability of occurrence of each factor in different situations;

[0093] Step S44: Sort the occurrence probability of each factor in different situations from large to small to mine potential unsafe factors.

[0094] It should be noted that, in some embodiments, a model based on Granger causality test combined with conditional entropy is recommended for calculating the strength of causal relationships in factor association network models. This model uses Granger causality test to determine the causal direction between factors and then uses conditional entropy to measure the strength of causal relationships. The formula for calculating the strength of causal relationships is:

[0095]

[0096] in, Representation factors Factors The strength of the causal relationship, For known factors Conditional factors The conditional entropy of Without considering factors Historical information time factors This model can more accurately explore the causal relationship between different physical quantities and improve the mining of unsafe factor chains.

[0097] like Figure 2 As shown, the present invention also discloses a system for identifying unsafe factors in the process of new energy infrastructure construction, including the following six units: a multi-source heterogeneous data aggregation and fusion unit, which is connected to various sensors distributed at the new energy infrastructure site for receiving and integrating multi-source data such as physical parameters and images from different sensors; an adaptive data cleaning and feature extraction unit, whose input is connected to the output of the multi-source heterogeneous data aggregation and fusion unit, and is used to clean the aggregated data and extract features based on a specific algorithm; a dynamic feature library construction and maintenance unit, whose input is connected to the output of the adaptive data cleaning and feature extraction unit, and is responsible for constructing and dynamically updating the feature library based on historical cases, etc. Unsafe factor feature library; Similarity matching and preliminary judgment unit, whose input is connected to the output of the adaptive data cleaning and feature extraction unit and the dynamic feature library construction and maintenance unit, and makes a preliminary judgment of unsafe factors through a specific similarity algorithm; Multi-factor association mining and analysis unit, whose input is connected to the output of the similarity matching and preliminary judgment unit, and uses the constructed factor association network model to analyze the associated factors of unsafe factors; Comprehensive evaluation and visualization presentation unit, whose input is connected to the output of the similarity matching and preliminary judgment unit and the multi-factor association mining and analysis unit, uses a specific comprehensive evaluation algorithm to evaluate the severity of unsafe factors, outputs the results in a visual form, and generates a report.

[0098] The identification of unsafe factors in traditional new energy infrastructure relies on manual inspections or single-sensor monitoring, which suffers from problems such as incomplete coverage and poor real-time performance. This system uses temperature, pressure, vibration sensors, and image acquisition equipment through multi-source data collection steps to achieve comprehensive monitoring of the construction environment. Taking the construction of wind farm towers as an example, vibration sensors use high-frequency sampling to capture subtle equipment anomalies, and image acquisition equipment monitors personnel operations in real time, solving the problems of low efficiency and easy omissions in manual inspections. At the same time, the sampling frequency is dynamically adjusted to ensure accurate data collection in key areas, avoiding data loss caused by the unreasonable layout of traditional sensors and overcoming the shortcomings of existing technologies such as incomplete monitoring and low data quality.

[0099] In the data processing and analysis stage, existing technologies mostly use simple threshold judgments or traditional statistical methods, which are unable to deeply explore complex data relationships and have poor adaptability to newly emerging unsafe factors. The dynamic feature library constructed by this method is based on historical cases, industry standards and expert experience, combined with an incremental learning update mechanism, which can promptly incorporate new risk features brought about by the application of new energy storage materials, and is more timely than traditional static feature libraries. In addition, the multi-factor association analysis model based on the Bayesian network can accurately sort out the cause-and-effect relationship of various factors. For example, in the operation and maintenance of wind farms, it can clearly analyze the relationship between abnormal blade vibration and gearbox failure and wind speed changes, which changes the limitations of traditional methods of isolated data analysis and greatly improves the accuracy and comprehensiveness of unsafe factor identification.

[0100] In terms of result output and application, existing technologies lack systematicity in assessing the severity of unsafe factors, and the results are presented in a single format, making it difficult to assist in decision-making. This system uses a comprehensive assessment algorithm that integrates fuzzy hierarchical analysis and grey correlation analysis to quantify the severity of unsafe factors from multiple dimensions such as safety risks, economic losses, and project progress. At the same time, it uses a visual interface to display the risk level, location, and scope of impact, and generates a report containing detailed data and analysis basis. Compared with the traditional method of only providing simple alarm information, it can help staff quickly grasp the overall risk picture and formulate scientific response strategies, effectively making up for the shortcomings of existing technologies in risk assessment and decision-making support, and providing a strong guarantee for the safe and efficient construction of new energy infrastructure.

[0101] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying unsafe factors in the process of new energy infrastructure construction, characterized by: The steps include: Step S1: Analyze unsafe factors in different construction stages and different accident types, construct unsafe factor feature vectors in different construction stages and different accident types, and thus construct a feature database of unsafe factors; Step S2: With the goal of maximizing coverage of the infrastructure area and minimizing the number of sensors, various sensors are deployed at the infrastructure site to collect real-time data and extract unsafe factor feature vectors. Step S3: Match the unsafe factor feature vectors collected in real time with the unsafe factor feature vectors in the feature database, and use a matching algorithm based on similarity measurement to calculate the similarity between the two. When the calculated similarity is greater than a set similarity threshold, it is preliminarily determined that the corresponding unsafe factor exists and is recorded as the current unsafe factor. Step S4: Analyze other factors associated with the current unsafe factor and construct a factor association network model. Consider the causal relationship, temporal sequence, and spatial position relationship between different physical quantities to determine the unsafe factor chain and explore potential unsafe factors. The step S4 comprises: Step S41: Analyze other factors associated with the current unsafe factor and map the other factors as variables to the Bayesian network nodes; Step S42: Determine directed edges between nodes based on the causal relationships, temporal sequences, and spatial position relationships between different physical quantities, draw a factor association network, and adjust the positions of nodes and edge layouts based on the actual logical relationships in the infrastructure construction process; Step S43: For each node, estimate its conditional probability table based on historical data and expert experience, and analyze the probability of occurrence of each factor in different situations; Step S44: sort the occurrence probabilities of various factors under different circumstances from large to small to mine potential unsafe factors; Step S5: Use the fusion model of fuzzy hierarchical analysis and grey relational analysis to comprehensively evaluate the severity of current unsafe factors and potential unsafe factors.

2. The method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 1, characterized in that: The step S1 comprises: Step S11: Analyze unsafe factors in different construction stages and accident types based on historical accident cases, industry standards, and expert experience, and determine the corresponding relationship between characteristics and unsafe factors; Step S12: Use the improved hierarchical clustering algorithm to preliminarily cluster the feature vectors of the infrastructure site. In the clustering process, the spatial correlation weights between features are introduced. The improved hierarchical clustering distance calculation formula is: ， in, and are two eigenvectors, and are vectors and No. The first dimension eigenvalues, For the The spatial correlation weight of the dimension, is the dimension of the feature vector, is the number of features under each dimension; Step S13: Input the clustering results in step S12 into the decision tree model for further classification. The decision tree model splits the nodes according to the information gain ratio of the features, classifies and organizes the features, and improves the feature database of unsafe factors.

3. The method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 2, characterized in that: The method further includes step S14, using an incremental learning algorithm to update the feature database of unsafe factors, and promptly incorporate the features of newly emerged unsafe factors into the feature database. The incremental learning process uses a weight update formula based on gradient descent: ， in, for The weight vector of the moment feature library model, is the learning rate, is the loss function About the weight vector gradient.

4. The method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 1, characterized in that: The step S2 comprises: Step S21: Arrange various sensors at the infrastructure construction site based on the spatial optimization layout model. The layout model formula is as follows: ， Constraints: , in, Indicates whether it is at the location Arrange sensors, Indicates the location The cost of deploying sensors, Indicates location Sensor pair area coverage capabilities, Indicates location The set of areas that can be covered by the sensor, Represents the set of all areas where the infrastructure site is divided; Step S22: Determine the sampling frequency of each sensor type based on the sensor type and the dynamic characteristics of the infrastructure construction site, and collect data on the physical parameters and real-time scenes at the infrastructure construction site at the specific sampling frequency; Step S23: extracting the unsafe factor feature vectors in the infrastructure construction site in real time.

5. The method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 4, characterized in that: The unsafe factor feature vector includes time domain features, frequency domain features and abstract features.

6. The method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 1, characterized in that: In step S5, the weights of different unsafe factors and their associated factors are determined by fuzzy hierarchical analysis, and the fuzzy judgment matrix elements are The calculation is based on expert evaluation of factors and factors The fuzzy evaluation of relative importance is then performed, and then the correlation between the real-time unsafe factors and the standard risk pattern is calculated through grey correlation analysis. The grey correlation calculation formula is: ， in, Reference sequence Compare with sequence The grey relational degree of and The reference sequence and i The comparison sequence is in The value under the indicator, is the resolution coefficient.

7. The system corresponding to the method for identifying unsafe factors in the process of new energy infrastructure construction according to claim 1 is characterized in that: It includes a multi-source heterogeneous data aggregation and fusion unit, an adaptive data cleaning and feature extraction unit, a dynamic feature library construction and maintenance unit, a similarity matching and preliminary judgment unit, a multi-factor association mining and analysis unit, and a comprehensive evaluation and visualization presentation unit; the multi-source heterogeneous data aggregation and fusion unit is used to receive and integrate physical parameters and image multi-source data from different sensors; the adaptive data cleaning and feature extraction unit is used to clean the aggregated data and extract features based on specific algorithms; the dynamic feature library construction and maintenance unit is responsible for constructing and dynamically updating the unsafe factor feature library based on historical cases; the similarity matching and preliminary judgment unit makes a preliminary judgment on unsafe factors through a similarity algorithm; the multi-factor association mining and analysis unit uses the constructed factor association network model to analyze the associated factors of unsafe factors; the comprehensive evaluation and visualization presentation unit uses a comprehensive evaluation algorithm to evaluate the severity of unsafe factors.

Images