Cable fire prediction method based on cable thermal degradation characteristic gas concentration data

Abstract

The invention discloses a cable fire prediction method based on cable thermal degradation characteristic gas concentration data, and relates to the technical field of cable fire prediction, and the method comprises the steps: building a data fusion model, and processing first data; carrying out data fusion and error analysis; a fire prediction model is established to generate fire occurrence probability information. According to the method, the data fusion model is established, the concentration data of the gas released in the cable thermal degradation extremely early stage process is used as the characteristic quantity for data fusion, and the fire prediction model is established for prediction, so that more accurate and real-time cable fire prediction can be provided; early early warning is realized to reduce cable fire risk and guarantee stable operation of a power system.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fire prediction, and in particular to a cable fire prediction method based on cable thermal degradation characteristic gas concentration data. Background Art

[0002] As a key medium for power systems and information transmission, cables are vital to the infrastructure of modern society. Their reliability directly affects the stability of energy supply and the continuity of information transmission. However, cables may experience thermal degradation during long-term operation. This is due to overload, poor contact, high ambient temperature or material aging, which causes the internal temperature of the cable to rise, thereby affecting its insulation performance. Thermal degradation will not only shorten the service life of the cable, but may also cause cable fires. Cable fires can lead to power supply interruptions and cause significant economic losses. What's more serious is that they may cause secondary disasters such as the release of toxic smoke and the spread of fire, posing a serious threat to personnel safety and the environment.

[0003] In order to prevent cable thermal degradation and fire accidents, existing prediction algorithms mainly focus on temperature monitoring, load analysis and material property evaluation. Although these algorithms can evaluate the operating status of cables to a certain extent, they often have certain limitations. For example, some algorithms may rely on a large amount of historical data and are not sensitive enough to real-time changes. Therefore, the development of a new real-time prediction algorithm is of great significance to improving the safety and reliability of cable operation. 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: the existing cable fire prediction method is not responsive, relies on a large amount of historical data, and cannot evaluate the changes in the cable operating status in real time, and how to improve the real-time and accuracy of cable fire prediction.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a cable fire prediction method based on cable thermal degradation characteristic gas concentration data, comprising establishing a data fusion model to process the first data; performing data fusion and error analysis; and generating fire occurrence probability information by establishing a fire prediction model.

[0007] As a preferred solution of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, wherein: the data fusion model includes feature extraction, preprocessing, preliminary fusion and error analysis.

[0008] As a preferred solution of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, wherein: the processing of the first data includes feature extraction and preprocessing of different cable thermal degradation characteristic gas sensor cluster node data.

[0009] As a preferred solution of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, the data fusion includes preliminary fusion of the pre-processed data at the cluster head node based on a data fusion model.

[0010] As a preferred solution of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, the error analysis includes classifying the sensor data based on a data fusion model and performing error analysis.

[0011] As a preferred scheme of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, wherein: the fire prediction model includes judging various probability situations of fire risks, and outputting the levels of no fire probability P1, smoldering probability P2 and open flame probability P3, which are set to large, medium and small, and the level of fire probability P is divided into four levels: general, large, major and extremely major.

[0012] As a preferred embodiment of the cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in the present invention, wherein: the generation of fire occurrence probability information includes establishing a fire occurrence probability discrimination rule base based on the fused data, and there are 27 results for the fire probability P level, of which 5 results are general, 6 results are large, 9 results are major and 7 results are particularly major; when at least two of the no-fire probability P1, smoldering probability P2, and open flame probability P3 are small, and the open flame probability P3 level is not medium or large, the output fire probability P level is general; when at least two of the no-fire probability P1, smoldering probability P2, and open flame probability P3 are large, and the open flame probability P2 level is not small, or when the no-fire probability P1 level is small, the smoldering probability P2 level is large, and the open flame probability P3 level is medium, the output fire probability P level is particularly major.

[0013] Another object of the present invention is to provide a cable fire prediction system based on cable thermal degradation characteristic gas concentration data, which can solve the problem of insensitive response of current cable fire prediction by establishing a data fusion model, performing feature extraction and data preprocessing on the first data.

[0014] As a preferred solution of the cable fire prediction system based on cable thermal degradation characteristic gas concentration data described in the present invention, it includes: a data processing module, an error analysis module, and a fire probability prediction module; the data processing module is used to establish a data fusion model, and perform feature extraction and data preprocessing on the first data; the error analysis module is used for data fusion and error analysis, and classifies and analyzes the processed data; the fire probability prediction module is used to establish a fire prediction model, generate fire occurrence probability information according to the fused data, output the probability of no fire, the probability of smoldering and the probability of open flame, and finally evaluate the fire probability, and output general, large, major and particularly major fire level probability information.

[0015] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a cable fire prediction method based on cable thermal degradation characteristic gas concentration data.

[0016] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a cable fire prediction method based on cable thermal degradation characteristic gas concentration data.

[0017] Beneficial effects of the present invention: The cable fire prediction method based on cable thermal degradation characteristic gas concentration data provided by the present invention establishes a data fusion model and processes the first data, and collects the cable thermal degradation characteristic gas concentration data in real time, so as to realize the comprehensive integration of multiple sensor data, which not only ensures the diversity of data and the extensiveness of sources, but also effectively improves the overall quality of data. Feature extraction and preprocessing are performed in the processing process, which can eliminate obvious outliers and ensure that the data used in subsequent analysis has higher accuracy and reliability. At the same time, the integration of data can reveal the relationship between different gas concentrations, so as to provide more comprehensive information to support fire warning, lay a solid foundation for subsequent fire risk prediction, and improve the system's response speed and accuracy to potential fires. In the classification and error analysis of the processed data, the cluster head node based on the data fusion model is used to classify the received data. The sensor data can be effectively classified, and errors and unreliable data can be further eliminated. The data of different sensors can be distinguished and integrated by applying logical algorithms, so as to obtain more accurate statistical results. In particular, the error analysis link can identify which sensor data behave abnormally under a specific environment or condition, and then correct the influence of these data, thereby realizing the refinement of data analysis, making the data based on which the fire prediction model is input have higher credibility and accuracy, and effectively reducing the possible risk of erroneous judgment. By constructing a fire prediction model and generating fire probability information based on the fused data, a comprehensive assessment of different fire risk levels can be achieved. It not only provides a multi-dimensional fire risk assessment, but also simplifies complex data relationships into clear output levels by establishing a fire probability discrimination rule base. The present invention achieves better results in terms of efficiency, accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0019] Figure 1 An overall flow chart of a cable fire prediction method based on cable thermal degradation characteristic gas concentration data provided by the first embodiment of the present invention.

[0020] Figure 2 A data fusion model based on K-Means clustering algorithm optimization is provided for a cable fire prediction method based on cable thermal degradation characteristic gas concentration data in the first embodiment of the present invention.

[0021] Figure 3A structure diagram of a fire prediction model based on a fuzzy neural network for a cable fire prediction method based on cable thermal degradation characteristic gas concentration data provided in the first embodiment of the present invention.

[0022] Figure 4 A schematic diagram of a module of a cable fire prediction system based on cable thermal degradation characteristic gas concentration data provided by the third embodiment of the present invention. DETAILED DESCRIPTION

[0023] 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.

[0024] Example 1, reference Figure 1-Figure 3 , which is an embodiment of the present invention, provides a cable fire prediction method based on cable thermal degradation characteristic gas concentration data, comprising:

[0025] S1: Establish a data fusion model to process the first data.

[0026] Furthermore, a data fusion model includes feature extraction, preprocessing, preliminary fusion and error analysis.

[0027] It should be noted that processing the first data includes extracting and preprocessing the characteristics of different cable thermal degradation characteristics of the gas sensor cluster node data.

[0028] It should also be noted that there are many ways to establish a data fusion model, such as a K-Means clustering algorithm or an IQR-based outlier elimination method.

[0029] As a preferred embodiment, a data fusion model is established in this embodiment based on the K-Means clustering algorithm, and its principle steps are as follows:

[0030] The data fusion model optimized based on K-Means clustering algorithm is as follows Figure 2 As shown, first, the first data of the cable thermal degradation characteristic gas sensor are collected centrally, including but not limited to CO concentration data, CO2 concentration data, CH4 concentration data, H2 concentration data and smoke concentration data, etc., and feature extraction is performed. Subsequently, a simple and effective low-level preprocessing is performed on the raw data collected by each sensor node to eliminate obvious outliers and ensure the accuracy of the data.

[0031] The preprocessed data will be sent to the cluster head node for preliminary fusion. In the preliminary fusion stage, the data will be screened again to further eliminate possible abnormal data. The processed sensor data will be sorted according to its numerical value to determine the maximum, middle and minimum values, and these values ​​will be defined as cluster centers.

[0032] Next, all sensor data are classified into the nearest cluster center according to the minimum distance principle, and the value of each cluster center is recalculated. In this process, the three cluster centers determined will be used for error analysis, specifically calculating the distance between the maximum and minimum cluster centers and the middle cluster center.

[0033] Finally, according to the preset allowable error range, the distance from each cluster center to the middle cluster center is compared with the error range. If the distance from a cluster center to the middle cluster center exceeds the error range, the relevant data group will be deleted. Otherwise, it will be retained and input into the fire prediction model. The data fusion model optimized based on the K-Means clustering algorithm ensures the reliability of the data, thereby providing a more accurate basis for subsequent fire predictions.

[0034] It should also be noted that in another feasible embodiment, a data fusion model is established by adopting an outlier elimination method based on IQR, and its principle steps are as follows:

[0035] First, the data of the cable thermal degradation characteristic gas sensor are collected, including but not limited to CO concentration data, CO2 concentration data, CH4 concentration data, H2 concentration data and smoke concentration data. Next, relevant features are extracted from the collected sensor data for further processing. On this basis, the collected data is pre-processed at a low level to remove obvious outliers. Then, the first quartile Q1 and the third quartile Q3 of the data set are calculated, and the interquartile range IQR is calculated, that is, the value of Q3 minus Q1. According to the calculated IQR, the standard for identifying outliers is lower than Q1 minus 1.5 For data with a value higher than Q3 plus 1.5 times IQR or higher than Q3 plus 1.5 times IQR, the identified abnormal data will be eliminated to ensure the accuracy and authenticity of the remaining data. For the data without abnormality after this step, further data fusion is performed at the cluster head node. During the data fusion process, the fused data is used for error analysis to ensure that the data error is within the allowable range, so that more accurate subsequent prediction analysis can be performed. The outlier elimination and data fusion model based on the IQR method makes the final data more reliable and accurate through strict outlier identification and screening, which effectively improves the accuracy and real-time performance of fire prediction.

[0036] S2: Perform data fusion and error analysis.

[0037] Furthermore, data fusion includes preliminary fusion of the preprocessed data at the cluster head node based on a data fusion model, and further eliminating some abnormal data.

[0038] It should be noted that the error analysis includes classifying the sensor data received by the cluster head node based on a data fusion model and performing error analysis based on the error range.

[0039] It should also be noted that, based on a data fusion model, as a preferred embodiment, a data fusion model optimized based on the K-Means clustering algorithm is adopted in this embodiment. After the central node receives the data sent by each cluster member node, in view of the fact that there may be some deviation data in the data that the cluster member nodes cannot eliminate, a reasonable weight value is assigned to the data of each cluster member node according to the degree of support of each cluster member node for the collected concentration data, a high weight is assigned to the data with high reliability, and a low weight is assigned to the data with low reliability, so as to achieve the best fusion effect.

[0040] S3: Generate fire occurrence probability information by establishing a fire prediction model.

[0041] Furthermore, a fire prediction model includes judging various probability situations of fire risk, outputting the levels of no fire probability P1, smoldering probability P2 and open flame probability P3 as large, medium and small, and dividing the level of fire probability P into four levels: general, large, major and extremely major.

[0042] It should be noted that generating fire probability information includes establishing a fire probability discrimination rule base based on the fused data. There are 27 fire probability levels, including 5 general, 6 large, 9 major and 7 particularly major results. When at least two of the no-fire probability P1, smoldering probability P2 and open flame probability P3 are small, and the open flame probability P3 is not medium or large, the output fire probability P level is general. When at least two of the no-fire probability P1, smoldering probability P2 and open flame probability P3 are large, and the open flame probability P2 is not small, or when the no-fire probability P1 is small, the smoldering probability P2 is large and the open flame probability P3 is medium, the output fire probability P level is particularly major.

[0043] It should also be noted that there are many ways to establish a fire prediction model, such as based on a fuzzy neural network or a decision tree.

[0044] As a preferred embodiment, a fire prediction model is established in this embodiment based on a fuzzy neural network, and its principle steps are as follows:

[0045] The structure diagram of the fire prediction model based on fuzzy neural network is as follows Figure 3As shown, fuzzy logic reasoning is combined with neural network in series, including input layer, hidden layer, output layer, definition of fuzzy set and fuzzy rule base. The fuzzy rule base is a rule base for distinguishing the probability of fire occurrence. The neural network algorithm simulates the human brain neural network. People can quickly discover fires by visual thinking, and can self-learn to adapt to changes in various environments, showing strong fault tolerance. The fuzzy logic algorithm is an algorithm designed to imitate the logical thinking method of people. It has strong recognition ability for nonlinear structural problems such as fire prediction. Through the adaptive learning ability of real-time feedback of neural network combined with fuzzy rules, the fuzzy probability of fire occurrence is output, and finally a more accurate prediction of the cable fire level is made through defuzzification.

[0046] According to the data processed by the data fusion model, the number of input layer nodes of the neural network is determined to be 5, and the input variables are CO concentration, CO2 concentration, CH4 concentration, H2 concentration, and smoke concentration. At the same time, the number of output layer nodes of the neural network is determined to be 3, and the output variables are open flame probability P1, smoldering probability P2, and no fire probability P3.

[0047] Calculate the number of hidden layer nodes, expressed as:

[0048]

[0049] Among them, n H is the number of hidden layer nodes, n1 is the number of input nodes, n2 is the number of output nodes, n is an integer between 1 and 10, so the number of hidden layer nodes is determined to be 8.

[0050] After determining the number of nodes in each layer of the neural network, the neural network model needs to be trained. Using the error gradient descent principle of the neural network, continuous cycles and iterations are used to obtain the optimal weights and thresholds and improve the neural network structure. The specific steps include:

[0051] A1, initialization, select the maximum allowable error Emax acceptable during the neural network training process, and set the weight coefficient w ij , v jk and threshold θ j , r t Initialize to small random values.

[0052] A2, input training sample (A1, Y1), A1 is the input value, Y1 is the expected output value.

[0053] A3, calculate the hidden layer output, expressed as:

[0054]

[0055] Among them, b j is the output of the jth node in the hidden layer, f is the activation function used to introduce nonlinearity, sj is the input signal of the jth node in the hidden layer, m is the total number of nodes in the input layer, and w ij is the connection weight from the i-th node in the input layer to the j-th node in the hidden layer, x i is the input value of the i-th node in the input layer, θ j is the threshold of the jth node in the hidden layer, and p is the total number of nodes in the hidden layer.

[0056] A4, calculate the output layer output, expressed as:

[0057]

[0058] in, is the output prediction value, l k is the input signal of the kth node in the output layer, ν jk is the connection weight from the jth node in the hidden layer to the kth node in the output layer, r k is the threshold of the kth node in the output layer, and n is the total number of nodes in the output layer.

[0059] A5, calculate the correction error of each unit in the output layer, expressed as:

[0060]

[0061] Among them, d k is the correction error of the kth node in the output layer, is the predicted value of the kth node in the output layer, y k is the expected output value.

[0062] A6, calculates the correction error of each unit in the hidden layer, expressed as:

[0063]

[0064] Among them, e j is the correction error of the jth node in the hidden layer.

[0065] A7, calculate the new connection weight between the hidden layer and the output layer, expressed as:

[0066] v jk (L+1)=v jk (L)+αd k b j

[0067] r k (L+1)=r k (L)+αd k

[0068] Among them, v jk(L) is the connection weight from the jth node in the current hidden layer to the kth node in the output layer, v jk (L+1) is the updated connection weight from the jth node in the hidden layer to the kth node in the output layer, r k (L) is the threshold of the kth node in the current output layer, r k (L+1) is the threshold of the kth node in the updated output layer, and α is the learning rate from the hidden layer to the output layer.

[0069] A8, calculate the new connection weight between the input layer and the hidden layer, expressed as:

[0070] w ij (L+1)=w ij (L)+βe j x i

[0071] θ j (L+1)=θ j (L)+βe j

[0072] Among them, w ij (L) is the connection weight from the current input layer i-th node to the hidden layer j-th node, w ij (L+1) is the updated connection weight from the i-th node in the input layer to the j-th node in the hidden layer, θ j (L) is the threshold of the jth node in the current hidden layer, θ j (L+1) is the threshold of the jth node in the updated hidden layer, and β is the learning rate from the input layer to the hidden layer.

[0073] A9, repeat the process from step A3 to step A8 for the 2nd to Nth groups of training samples to complete the first training.

[0074] A10, start the second training, repeat the process of the first training, when the error E is less than the target value, or the number of training times reaches the maximum value, the training ends.

[0075] After completing the construction of the neural network model, it is combined in series with the fuzzy logic algorithm to fuzzify the input and output variables of the fuzzy logic algorithm. The input of the fuzzy logic algorithm is the output of the neural network, and the output is the probability of fire occurrence. The fuzzy levels of the no-fire probability P1, smoldering probability P2 and open flame probability P3 output by the neural network are defined as large (L), medium (M) and small (S). The output fire probability P has four fuzzy levels: general (S), large (M), major (L) and extremely major (B).

[0076] Define a fuzzy discrimination rule base, that is, a fire occurrence probability discrimination rule base. Fuzzy discrimination rules are usually expressed in the form of "IF (a set of conditions) THEN (a set of results are derived)". For example, when the no-fire probability level is S, the smoldering probability level is S, and the open flame probability level is S, the output fire occurrence probability is S. There are 27 results in total. Please refer to Table 1 for details.

[0077] Table 1 Fuzzy discrimination rule base

[0078]

[0079]

[0080] Finally, the center of gravity method is used to defuzzify the output fire probability to obtain accurate fire identification results, which can be expressed as:

[0081]

[0082] Among them, u is the output of the defuzzification process, that is, the centroid (center of gravity) of the fuzzy set, μ(u a ) is the membership function, indicating the a-th element u a The degree of membership in a fuzzy set, A is the total number of elements in the fuzzy set.

[0083] It should also be noted that in another feasible embodiment, a fire prediction model is established based on a decision tree, and its principle steps are as follows:

[0084] First, in the data preparation stage, various characteristic data related to fire need to be collected. The data processed by the data fusion model include but are not limited to CO concentration, CO2 concentration, CH4 concentration, H2 concentration and smoke concentration. At the same time, the corresponding fire occurrence situation needs to be obtained as label data for subsequent model training. By calculating the information gain or Gini coefficient of the feature, the features that are highly correlated with the fire risk can be identified. For example, if it is found that the prediction effect of CO concentration on the occurrence of fire is significantly higher than other features, then this feature should be preferentially selected as the root node of the decision tree. When constructing the decision tree, the system will recursively divide the selected features from top to bottom to form a series of decision nodes and subsets until the node purity conditions are met, such as reaching a specific sample. The threshold of this number or information gain is reduced to a certain level. Finally, through this layer-by-layer analysis method, the decision tree will generate multiple leaf nodes, each leaf node corresponds to a set of fire probability prediction results, which can be tested through cross-validation in the model evaluation stage. The evaluation indicators include classification accuracy and confusion matrix. In order to ensure the effectiveness and generalization ability of the model, it is important to implement pruning operations in the later stage of model construction to remove unnecessary branches to reduce the risk of overfitting. After completing the model training, the decision tree can process specific input data and make real-time fire probability predictions for the newly collected gas concentration data based on decision rules. Finally, the probability of fire occurrence is output in the form of fuzzy levels to help decision makers evaluate and respond to potential fire risks.

[0085] Example 2 is an embodiment of the present invention, which provides a cable fire prediction method based on cable thermal degradation characteristic gas concentration data. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0086] In this embodiment, a data fusion model based on the K-Means clustering algorithm was established, which was mainly used to process the data collected by the cable thermal degradation characteristic gas sensor to improve the accuracy and reliability of fire prediction. First, the experimental group selected multiple advanced gas sensors, including CO, CO2, CH4, H2 and smoke concentration sensors, and constructed a multi-sensor network for collecting fire-related data. These sensors were distributed in an actual climate environment to obtain real-time data. The experiment lasted for 24 hours.

[0087] After data collection, feature extraction is first performed to build a basic data set based on the gas concentration values ​​collected by the sensor. Then, the raw data is preprocessed at a low level to remove obvious outliers based on the concentration value of each gas to improve the accuracy and reliability of the data. During this process, preliminary statistical analysis is performed on all gas concentration data to ensure that the removed outliers are within the standard deviation range.

[0088] The preprocessed data is initially fused by the cluster head node. At this stage, the K-Means algorithm is used to further cluster the data, identify the central trend, and assign a reasonable weight value to each detected cluster data. By assigning higher weights to data with high reliability, it is ensured that the final retained data can accurately reflect the actual environmental status. After the fusion is completed, the central node analyzes the data of each cluster and performs a detailed error analysis on the clustering results, thereby optimizing the final output data set to ensure that it is suitable for subsequent fire risk prediction.

[0089] Next, the obtained fusion data was used to construct a fire prediction model based on fuzzy neural network. This model combines fuzzy logic reasoning with neural network. The input layer is set with five nodes, representing the concentration variables of CO, CO2, CH4, H2 and smoke respectively. The number of hidden layer nodes is calculated by empirical formula and determined to be 8 to enhance the learning ability and fault tolerance of the network. By setting connection weights and thresholds, the neural network can continuously adjust while learning new data to ensure that the output layer reflects the accurate values ​​of the probability of open fire, smoldering and no fire.

[0090] Finally, in the output stage of the model, the fuzzy logic algorithm is comprehensively used to fuzzify the output of the neural network. By defining the fuzzification level and the corresponding fuzzy discrimination rule base, the effective output of the fire probability is finally achieved. The center of gravity method is used for defuzzification processing to obtain more accurate fire risk discrimination results, thus providing an important basis for fire prevention and response.

[0091] Refer to Table 2 for a comparative analysis of the experimental data.

[0092] Table 2 Experimental data record table

[0093]

[0094] By analyzing the experimental data in Table 2, it can be found that the data fusion model optimized based on the K-Means clustering algorithm adopted in this embodiment shows advantages over the traditional fire warning system. Table 2 shows the gas concentration data collected by different sensors in the same time period and their corresponding risk levels. First, the gas concentrations recorded by sensors A, B and E are all in a relatively safe range, and the risk level is low, which indicates that the fire risk in these areas is relatively small. However, the data from sensor C shows that the risk level is medium, especially the CO, CO2 and smoke concentrations have increased to a certain extent, indicating that the fire hazard in this area is gradually increasing. In addition, the data from sensors D and F show that the risk level is high and particularly serious, among which the CO concentration is particularly prominent, reflecting that the fire risk at this point is increased and needs to be paid attention to.

[0095] Through this data analysis, it can be clearly seen that the implementation of the data fusion model based on the K-Means clustering algorithm ensures reliable data output, which is not available in traditional fire warning systems. Especially in the data processing process, the elimination of outliers and the innovation of data weighting methods have effectively improved the accuracy and availability of data. Especially in multi-sensor systems, the diversity and complexity of data often make traditional methods face challenges such as tuning and poor adaptability. The data fusion model optimized based on the K-Means clustering algorithm of the present invention fully utilizes the advantages of data obtained by different sensors through precise data fusion and weight distribution.

[0096] Finally, a fire prediction model based on fuzzy neural network is constructed to make the determination of fire probability more flexible and adaptable. Through the reasoning mechanism of fuzzy logic, the network can still achieve efficient self-learning ability even in the face of uncertain and complex environmental changes, providing a more accurate basis for the prevention and control of cable fires. Experimental results show that based on the combination of improved data fusion model and prediction algorithm, the accuracy and timeliness of cable fires are significantly improved. This innovative solution effectively overcomes the shortcomings of existing technologies and has broad application prospects and market value.

[0097] Example 3, reference Figure 4 , which is an embodiment of the present invention, provides a cable fire prediction system based on cable thermal degradation characteristic gas concentration data, including a data processing module, an error analysis module, and a fire probability prediction module.

[0098] The data processing module is used to establish a data fusion model to perform feature extraction and data preprocessing on the first data; the error analysis module is used for data fusion and error analysis, and to classify and analyze the processed data; the fire probability prediction module is used to establish a fire prediction model, generate fire occurrence probability information based on the fused data, output the probability of no fire, the probability of smoldering and the probability of open flame, and finally evaluate the fire probability, and output the probability information of general, large, major and extremely major fire levels.

[0099] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0100] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0101] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0102] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0103] 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 cable fire prediction method based on cable thermal degradation characteristic gas concentration data, characterized in that: include: Establishing a data fusion model to process the first data; Conduct data fusion and error analysis; By establishing a fire prediction model, the fire occurrence probability information is generated.

2. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 1, characterized in that: The data fusion model includes feature extraction, preprocessing, preliminary fusion and error analysis.

3. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 2, characterized in that: The processing of the first data includes extracting and preprocessing the characteristics of different cable thermal degradation characteristics of the gas sensor cluster node data.

4. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 3 is characterized in that: The data fusion includes performing preliminary fusion of the pre-processed data at the cluster head node based on a data fusion model.

5. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 4 is characterized in that: The error analysis includes classifying the sensor data based on a data fusion model and performing error analysis.

6. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 5, characterized in that: The fire prediction model includes judging various probability situations of fire risks, outputting the levels of no-fire probability P1, smoldering probability P2 and open flame probability P3 as large, medium and small, and dividing the level of fire probability P into four levels: general, large, major and extremely major.

7. The cable fire prediction method based on cable thermal degradation characteristic gas concentration data according to claim 6, characterized in that: The generating of fire probability information includes establishing a fire probability discrimination rule base according to the fused data, and the fire probability P level has 27 results, of which 5 results are general, 6 results are large, 9 results are major and 7 results are particularly major; When at least two of the no-fire probability P1, smoldering probability P2, and open flame probability P3 are small, and the open flame probability P3 is not medium or large, the output fire probability P level is general; When at least two of the no-fire probability P1, smoldering probability P2, and open flame probability P3 are large, and the open flame probability P2 is not small, or when the no-fire probability P1 is small, the smoldering probability P2 is large, and the open flame probability P3 is medium, the output fire probability P level is particularly severe.

8. A system using the cable fire prediction method based on cable thermal degradation characteristic gas concentration data as described in any one of claims 1 to 7, characterized in that: Including data processing module, error analysis module, fire probability prediction module; The data processing module is used to establish a data fusion model to perform feature extraction and data preprocessing on the first data; The error analysis module is used for data fusion and error analysis, and performs classification and error analysis on the processed data; The fire probability prediction module is used to establish a fire prediction model, generate fire occurrence probability information based on the fused data, output the probability of no fire, smoldering combustion and open flame, and finally evaluate the fire probability, and output general, large, major and extremely major fire level probability information.

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 cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in 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 cable fire prediction method based on cable thermal degradation characteristic gas concentration data described in any one of claims 1 to 7 are implemented.

Images