Oil tank leakage detection method and system based on Internet of Things

Abstract

The present invention relates to the technical field of electronic digital data processing, and specifically to an oil storage tank leakage detection method and system based on the Internet of Things, including: collecting multiple parameter data of each oil storage tank at different times, and taking the multiple parameter data of each oil storage tank at the same time as a data point; dividing each data point into normal data points and abnormal data points, clustering the abnormal data points to obtain multiple clusters; inputting the data points of each oil storage tank into an improved machine learning model, outputting the leakage probability, and if the leakage probability exceeds a set threshold, it is determined to be an abnormal oil storage tank. The improved machine learning model includes a Gaussian kernel function, which is inversely correlated with the validity and leakage degree of the parameter; the validity characterizes the abnormal degree of the parameter, and the leakage degree characterizes the difference between the cluster and the normal data point. The present invention solves the problem of low accuracy and efficiency in leakage detection of oil storage tanks.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and more specifically, to an oil tank leakage detection method and system based on the Internet of Things. Background Art

[0002] With the rapid development of the petroleum and chemical industry, oil tanks, as important oil storage facilities, play a vital role in the oil storage process. However, due to the long-term exposure of oil tanks to complex environments such as high temperature and high pressure, they are prone to leakage, resulting in oil leakage. Oil tank leakage not only causes serious pollution to the environment, but may also cause safety accidents such as fire and explosion, threatening the lives and property of the public and workers. Therefore, timely and accurate detection of oil tank leakage is the key to ensuring safe production.

[0003] At present, oil tank leakage detection technology mainly includes detection methods based on physical parameters such as liquid level changes, pressure monitoring and temperature changes. Although these traditional methods can reflect the abnormal conditions of oil tanks to a certain extent, due to their limitations, they usually can only rely on a single physical parameter and lack comprehensive consideration of multiple influencing factors. Especially when the liquid level changes are small or interfered by environmental factors, traditional methods often have the risk of misjudgment or missed detection. In addition, with the continuous increase in the number and scale of oil tanks, manual inspection methods are inefficient and difficult to achieve real-time monitoring.

[0004] The patent application document with application publication number CN114995307A discloses a safety monitoring system for oil storage warehouses. The patent application document detects the combustible organic matter and static electricity in the air by monitoring the matching of liquid level changes and flow rates, and promptly identifies oil leakage or static electricity risks, ensures safety, and issues corresponding alarms.

[0005] However, the above technical solution mainly relies on the matching of liquid level changes and flow rates to detect oil tank leaks, and does not fully consider other parameters, such as the potential impact of oil tank temperature, air pressure, etc. on detection accuracy. It is unable to fully respond to complex leak detection needs, resulting in low accuracy and efficiency in oil tank leak detection. Summary of the invention

[0006] In order to solve the problem of low accuracy and efficiency in oil tank leakage detection raised in the above background technology, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides an oil tank leakage detection method based on the Internet of Things, comprising: collecting multiple parameter data of each oil tank at different times, and taking the multiple parameter data of each oil tank at the same time as a data point; inputting the data points of each oil tank into an improved machine learning model, outputting the leakage probability of the oil tank, and responding that the oil tank with a leakage probability higher than a set threshold is an abnormal oil tank; wherein the improved machine learning model includes a Gaussian kernel function, and the Gaussian kernel function is inversely correlated with the effectiveness and the leakage degree; the leakage degree characterizes the abnormality of the parameter; the effectiveness is: , where For the The validity of the parameters, is the total number of clusters, is the normalized exponential function, For the The first The data point The value of the item parameter, For the The normal data point The value of the item parameter, For the The total number of data points in the clusters, is the total number of normal data points in the data points; the cluster cluster is obtained by clustering the abnormal data points in the data points; wherein, The center distance between clusters and normal data points , where is the total number of parameter items.

[0008] The above technical solution can accurately identify abnormal patterns in the data, evaluate the leakage probability of the oil tank, introduce dynamic evaluation of effectiveness and leakage degree, improve the model's ability to handle complex data, and capture subtle abnormal signals between multiple feature data by collecting multiple parameter data of the oil tank in real time, and combining Gaussian kernel function and cluster analysis methods. This method not only improves the accuracy and sensitivity of anomaly detection, but also enhances the adaptability of the system, ensuring that safety risks can be effectively prevented in different environments and conditions, and improving the safety and reliability of oil tank monitoring.

[0009] Further, Data points and Gaussian kernel function of data points for: ; In the formula, The natural constant The exponential function with base , To preset hyperparameters, For the The data point The value of the item parameter, For the The data point The value of the item parameter, is the total number of parameter items, For the The validity of the parameters, For the The degree of leakage of item parameters.

[0010] The above technical solution introduces the Gaussian kernel function, which comprehensively considers the effectiveness and leakage degree when calculating the similarity between data points, thereby enhancing the model's sensitivity to subtle data differences. In this way, the system can more accurately identify abnormal data points in complex multidimensional data spaces, especially when multiple factors interact. The introduction of the Gaussian kernel function makes the distance measurement between data more adaptable to nonlinear relationships, improves the accuracy and robustness of the model, and thus can identify potential abnormal risks earlier and more effectively in oil tank leakage detection, improving overall safety and response efficiency.

[0011] Furthermore, the leakage degree is: ; In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The total number of data points in the clusters, is the total number of clusters.

[0012] The above technical solution can more accurately evaluate the degree of deviation between data points and normal data by introducing a method for calculating the degree of leakage, thereby effectively identifying potential anomalies. By comparing the differences between data points in different clusters and normal data points, the system can quantify the degree of anomaly of each data point, thereby providing a more detailed basis for anomaly detection.

[0013] Furthermore, the leakage degree is: ; In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the item parameter, For the The normal data point The value of the item parameter, is the total number of clusters.

[0014] The above technical solution can dynamically evaluate the abnormality of data points by calculating the leakage degree, especially in the context of cluster analysis, accurately measuring the difference between data points and normal data. By quantifying the difference between data points within each cluster and normal data points, the system can effectively identify abnormal data points and further optimize the accuracy of anomaly detection. The normalization of the leakage degree ensures that the evaluation results remain consistent and stable under different data set sizes and types, and enhances the system's sensitivity to data changes.

[0015] Furthermore, the plurality of parameter data include: liquid level data, air pressure data, temperature data and oil and gas concentration data of the oil storage tank.

[0016] Furthermore, the DBSCAN algorithm is used to divide each data point into normal data points and abnormal data points.

[0017] The above technical solution divides the data points by introducing the DBSCAN algorithm. The technical solution can automatically identify normal and abnormal data based on the density characteristics of the data points. The algorithm does not rely on the preset number of categories, but automatically classifies the areas with higher density as normal data and the areas with lower density as abnormal data by analyzing the local density differences between the data points. This density-based clustering method can effectively cope with complex data distribution, reduce the need for human intervention, and improve the accuracy of anomaly detection and the adaptability of the system.

[0018] Furthermore, the machine learning model is a support vector machine model.

[0019] Furthermore, it also includes training the machine learning model, specifically: inputting the training set into a pre-built machine learning model for training, and during the training process, calculating the loss between the output prediction value and the label; adjusting the model parameters using the gradient descent method to minimize the prediction error; iteratively adjusting the parameters of the machine learning model until the loss is less than the set value or reaches the set number of training times, and finally obtaining a trained machine learning model.

[0020] In a second aspect, the present invention provides an oil tank leakage detection system based on the Internet of Things, comprising a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, any one of the above-mentioned oil tank leakage detection methods based on the Internet of Things is implemented.

[0021] The beneficial effects of the present invention are: The present invention can accurately identify the abnormal state of oil storage tanks and improve the sensitivity and accuracy of leak detection by comprehensively collecting multiple real-time parameters and combining them with an improved machine learning model. By adopting Gaussian kernel functions and technical means such as leakage degree and effectiveness, the abnormal recognition ability of the model is further optimized, so that the system can effectively distinguish between normal and abnormal situations in complex and changeable environments. At the same time, the density characteristics of data points are automatically divided in combination with the DBSCAN algorithm, so that the detection process does not need to rely on manually set thresholds, which enhances the adaptability and robustness of the system. This method can provide efficient and reliable technical support for the safety monitoring of oil storage tanks through the analysis of real-time data streams, significantly improve the timeliness and accuracy of leak warnings, reduce environmental risks and ensure the safe operation of oil storage tanks. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 is a flow chart schematically showing an oil tank leakage detection method based on the Internet of Things according to an embodiment of the present invention; Figure 2 is a structural block diagram schematically showing an oil tank leakage detection system based on the Internet of Things according to an embodiment of the present invention; Figure 3 Schematically shows a comparison diagram of ROC curves of the support vector machine model before and after improvement according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] An implementation example of an oil tank leakage detection method based on the Internet of Things.

[0026] like Figure 1 As shown, the flow chart of the oil tank leakage detection method based on the Internet of Things in an embodiment of the present invention includes the following steps: S1: Collect multiple parameter data of each oil storage tank at different times, and take the multiple parameter data of each oil storage tank at the same time as a data point.

[0027] In one embodiment, multiple parameter data of each oil storage tank at different times are collected, and the parameter data of all oil storage tanks at the same time are stored and analyzed as a complete data point. This method can not only effectively capture the state changes of each oil storage tank, but also realize the overall monitoring and comparison of multiple oil storage tanks, thereby providing more comprehensive data support for the safety management of the oil storage system; Specifically, the parameter data collected in this embodiment include the liquid level data, air pressure data, temperature data and oil and gas concentration data of the oil storage tank. These parameter information of each oil storage tank at different time points will be recorded as a separate sample point, so that when performing data analysis, not only the status of a single oil storage tank can be considered, but also potential systemic problems and hidden dangers can be further discovered by comparing the parameter data of different oil storage tanks.

[0028] After data collection is completed, data preprocessing is required to ensure data accuracy and consistency. Data preprocessing is a crucial step in any data analysis process, which can significantly improve the accuracy and credibility of subsequent model analysis. The specific preprocessing process includes but is not limited to: removing noise data, filling missing values, processing outliers, standardizing various parameter data, etc. Through these preprocessing steps, the interference of different sensors or environmental conditions on data collection can be eliminated, ensuring the reliability of all data points.

[0029] S2: Input the data points of each oil storage tank into the improved machine learning model, and output the leakage probability of the oil storage tank; wherein the improved machine learning model includes a Gaussian kernel function.

[0030] In one embodiment, the improved machine learning model includes a Gaussian kernel function, and the machine learning model is a support vector machine model; taking the support vector machine model as an example, the machine learning model is trained as follows: The training set is input into a pre-built machine learning model for training. During the training process, the loss between the output prediction value and the label is calculated, and the label is set manually; the model parameters are adjusted using the gradient descent method to minimize the prediction error; the parameters of the machine learning model are iteratively adjusted until the loss is less than the set value or the set number of training times is reached, and finally a trained machine learning model is obtained.

[0031] like Figure 3 As shown, a comparison diagram of the ROC curves of the support vector machine model before and after the improvement of the embodiment of the present invention.

[0032] Among them, Data points and Gaussian kernel function of data points for: ; In the formula, The natural constant The exponential function with base , To preset hyperparameters, For the The data point The value of the item parameter, For the The data point The value of the item parameter, is the total number of parameter items, For the The validity of the parameters, For the The degree of leakage of item parameters.

[0033] By introducing the Gaussian kernel function and combining it with a specific validity and leakage degree, the similarity between different data points can be accurately measured. This approach enables the model to more sensitively capture subtle differences between data points, especially when dealing with data with multiple influencing factors. By adjusting the hyperparameters in the similarity calculation, the model can more effectively distinguish between normal and abnormal data points, thereby improving the accuracy of classification and prediction.

[0034] The leakage degree is: ; In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The total number of data points in the clusters, is the total number of clusters.

[0035] By introducing a method to calculate the degree of leakage, this technical solution can dynamically evaluate the degree of abnormality of each data point in different clusters. This method quantifies the abnormal characteristics of data points by comparing the differences between the internal and normal data of the cluster, helping the model to identify potential abnormal patterns. Specifically, the degree of leakage can reveal the inconsistency and discreteness of data points in multidimensional space, allowing the model to more accurately detect data that deviates from normal patterns, thereby improving the accuracy of anomaly detection and classification tasks.

[0036] The validity is: , where For the The validity of the parameters, is the total number of clusters, is the normalized exponential function, For the The first The data point The value of the item parameter, For the The normal data point The value of the item parameter, For the The total number of data points in the clusters, is the total number of normal data points in the data points; the cluster cluster is obtained by clustering the abnormal data points in the data points, For the The center distance between clusters and normal data points; By introducing the calculation method of validity, the importance and representativeness of each data item in the overall data structure can be quantified. Validity reflects the contribution of data points in cluster analysis. By comparing the difference between the internal data of the cluster and the normal data, combined with normalization processing, the validity of each data item can be more accurately evaluated. This method effectively enhances the ability to identify abnormal data points, ensuring that the model can identify the most representative features in complex data sets, thereby improving the accuracy of classification and the robustness of the model.

[0037] The DBSCAN algorithm is used to divide each data point into normal data points and abnormal data points.

[0038] By using the DBSCAN algorithm to divide data points, this technical solution can effectively distinguish normal data points from abnormal data points. The DBSCAN algorithm is a density-based clustering method that identifies local density differences in data points, identifies areas with higher density as normal data points, and divides areas with lower density as abnormal data points. This method does not rely on the number of data categories set in advance, can automatically identify abnormal patterns in data, and effectively deal with the challenges of noise and irregular distribution.

[0039] Among them, The center distance between clusters and normal data points , where is the total number of parameter items.

[0040] By introducing the calculation of the center distance between clusters and normal data points, this technical solution can quantify the distance difference between clusters and normal data, and further improve the ability to identify abnormal data points. This method reflects the similarity between clusters and normal data by calculating the center distance between the center of clusters and normal data points. This technology can effectively identify clusters that are significantly different from normal data, so that abnormal data points can be accurately classified.

[0041] In another embodiment, the leakage degree is: ; In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the item parameter, For the The first normal data point The value of the item parameter, is the total number of clusters.

[0042] By introducing a new method for calculating the degree of leakage, this technical solution can more accurately measure the difference between abnormal data points and normal data points in a data set. This calculation method reveals the degree of abnormality of a cluster by comparing the difference between the average value of the data within the cluster and the average value of the normal data points. Normalization makes the assessment of the degree of leakage more objective and consistent, and can eliminate the impact of different data scales or ranges. This technical solution improves the sensitivity of the model in a complex data environment, enabling it to more accurately identify potential abnormal patterns, optimize anomaly detection capabilities, and enhance the system's adaptability and reliability to various data changes.

[0043] S3: In response to the oil storage tank having a leakage probability higher than a set threshold being an abnormal oil storage tank.

[0044] The above threshold value may be set to 0.6, and of course, it may be determined according to actual conditions.

[0045] The purpose of setting the threshold is to provide a clear standard for abnormal detection of oil tanks. When the probability of leakage exceeds the threshold, the system can immediately identify and mark the abnormal oil tank, providing effective alarms and timely decision support for operators. The process of setting the threshold is dynamically adjusted and can be continuously optimized according to different environmental conditions and historical data, thereby ensuring that the system can maintain the best detection effect under different working conditions.

[0046] In addition, the setting of thresholds is not only for simple anomaly detection, but also helps to enhance the fault tolerance and flexibility of the system. By properly adjusting the thresholds, the system can flexibly respond to different levels of abnormal risks according to actual conditions. From minor anomalies to serious failures, the progressive alarm mechanism can help decision makers make more reasonable and faster responses. For example, for initial anomalies, the system may issue an early warning signal and recommend inspections; and for situations where the probability of leakage is seriously exceeded, the system will immediately activate the emergency response mechanism.

[0047] In one embodiment, when the leakage probability of an oil tank exceeds a preset safety threshold, the system will automatically identify the oil tank as an abnormal oil tank. This design can not only monitor the operating status of the oil tank in real time, but also quickly respond to potential risks, so that measures can be taken as soon as possible when an abnormality occurs to prevent safety accidents. This process ensures that the oil tank monitoring system can remain efficient and accurate when facing large-scale data through automated data analysis and intelligent judgment mechanisms, thereby maximizing the system's response speed and accuracy.

[0048] The solution of the present invention realizes efficient detection and monitoring of oil tank leakage by combining Internet of Things technology and improved machine learning models. The method can collect multiple parameter data of oil tanks in real time, and accurately assess the leakage risk of oil tanks through intelligent analysis and processing of data. Using Gaussian kernel function and density clustering technology, the system can effectively identify and distinguish between normal and abnormal oil tanks, greatly improving the sensitivity to abnormal situations and detection accuracy. By training and optimizing machine learning models, the system can automatically adjust and improve prediction capabilities, reduce human intervention, and achieve early warning of potential leaks. This technology has important application value in improving oil tank safety and reducing environmental risks, and provides an innovative solution for intelligent monitoring of oil tanks.

[0049] Example of oil tank leakage detection system based on Internet of Things: like Figure 2 As shown, a structural block diagram of an oil tank leakage detection system based on the Internet of Things in an embodiment of the present invention includes a processor and a memory.

[0050] The present invention also provides an oil tank leakage detection system based on the Internet of Things. Figure 2 As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the oil tank leakage detection method based on the Internet of Things according to the present invention is implemented.

[0051] The IoT-based oil tank leakage detection system also includes other components familiar to those skilled in the art, such as a communication interface, and the configuration and functions of the components are known in the art, so they will not be described in detail here.

[0052] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

[0053] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly defined.

[0054] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. An oil tank leakage detection method based on the Internet of Things, characterized in that: include: Collect multiple parameter data of each oil storage tank at different times, and take the multiple parameter data of each oil storage tank at the same time as one data point; Inputting the data points of each oil storage tank into the improved machine learning model, outputting the leakage probability of the oil storage tank, and responding that the oil storage tank with a leakage probability higher than a set threshold is an abnormal oil storage tank; The improved machine learning model includes a Gaussian kernel function, which is inversely correlated with the effectiveness and the leakage degree; the leakage degree represents the abnormality of the parameter; The validity is: , where For the The validity of the parameters, is the total number of clusters, is the normalized exponential function, For the The first The data point The value of the item parameter, For the The normal data point The value of the item parameter, For the The total number of data points in the clusters, is the total number of normal data points in the data points; the cluster cluster is obtained by clustering the abnormal data points in the data points; Among them, The center distance between clusters and normal data points , where is the total number of parameter items.

2. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: No. Data points and Gaussian kernel function of data points for: ； In the formula, The natural constant The exponential function with base , To preset hyperparameters, For the The data point The value of the item parameter, For the The data point The value of the item parameter, is the total number of parameter items, For the The validity of the parameters, For the The degree of leakage of item parameters.

3. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: The leakage degree is: ； In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The first The data point The value of the parameter is the same as The normal data point The difference in the parameters, For the The total number of data points in the clusters, is the total number of clusters.

4. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: The leakage degree is: ； In the formula, For the The leakage degree of the item parameters, is the normalization function, For the The total number of data points in the clusters, is the total number of normal data points, For the The first The data point The value of the item parameter, For the The normal data point The value of the item parameter, is the total number of clusters.

5. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: The multiple parameter data include: liquid level data, air pressure data, temperature data and oil and gas concentration data of the oil storage tank.

6. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: The DBSCAN algorithm is used to divide each data point into normal data points and abnormal data points.

7. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: The machine learning model is a support vector machine model.

8. The oil tank leakage detection method based on the Internet of Things according to claim 1 is characterized in that: It also includes training the machine learning model, specifically: The training set is input into the pre-built machine learning model for training. During the training process, the loss between the output prediction value and the label is calculated; Use gradient descent to adjust model parameters to minimize prediction error; Iteratively adjust the parameters of the machine learning model until the loss is less than the set value or the set number of training times is reached, and finally a trained machine learning model is obtained.

9. The oil tank leakage detection system based on the Internet of Things is characterized by: The method comprises a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the oil tank leakage detection method based on the Internet of Things as described in any one of claims 1 to 8 is implemented.

Images