An on-line monitoring system for grounding resistance of oil depot storage tank

Abstract

This invention discloses an online monitoring system for the grounding resistance of oil depot storage tanks, belonging to the field of oil depot safety monitoring technology. The system utilizes a distributed acquisition configuration of four measurement points per tank to achieve non-contact online measurement through a grounding resistance acquisition module. The signal conditioning module integrates an adaptive interference suppression algorithm to dynamically identify and suppress complex interferences such as power frequency interference and electromagnetic radiation. The main control module accurately separates the resistance values ​​of each down conductor through matrix solving of a system of four linear equations, and constructs an aging corrosion prediction model using a fusion algorithm of random forest and XGBoost to predict potential hazards. The early warning module employs a three-level intelligent early warning mechanism with an alarm response time ≤3 seconds, and can dynamically adjust the early warning threshold according to the regional safety level and severe weather. Data is transmitted remotely via wireless communication, and the accompanying host computer platform supports multi-channel alarm push, data traceability, and intelligent management.

Description

Technical Field

[0001] This invention relates to the field of oil depot safety monitoring technology, specifically to an online monitoring system for the grounding resistance of oil depot storage tanks. Background Technology

[0002] As core facilities for storing flammable and explosive petroleum products, oil depot storage tanks rely heavily on their grounding systems. These systems are crucial safety barriers against fires and explosions caused by static electricity buildup and lightning strikes, directly determining the effectiveness of grounding protection. According to relevant requirements, the grounding resistance of storage tanks must be strictly controlled below 10Ω to ensure effective discharge of static charge and avoidance of safety risks. However, significant technical bottlenecks still exist in the field of oil depot storage tank grounding resistance testing. Existing oil depots generally use manual, periodic offline measurements, employing megohmmeters or clamp-on grounding resistance testers to perform single-point measurements by disconnecting the grounding lead when the tank is not in operation. This traditional testing method is severely out of step with the safety operation requirements of high-risk, large-capacity storage tank groups in the petrochemical industry, and its inherent limitations cannot meet the monitoring requirements of modern oil depots for all-weather, high-precision, and inherently safe operation.

[0003] Currently, existing traditional technologies and related online monitoring devices cannot solve the practical pain points of oil depot grounding monitoring: First, the detection mode is outdated and poses safety hazards. Manual offline measurement is time-consuming, and data lag makes it impossible to capture sudden faults such as grounding body corrosion and down conductor breakage in real time, creating blind spots in safety monitoring. Moreover, on-site measurement requires disconnecting the grounding circuit, which is cumbersome and poses an ignition risk in explosion-prone areas. Second, the measurement accuracy and anti-interference capabilities are insufficient. Manual operation is easily affected by environmental factors such as soil moisture and temperature. Existing online monitoring devices have weak anti-electromagnetic interference capabilities, low data transmission reliability, and lack explosion-proof certification, making them inadequate for monitoring. The system is not suitable for the complex and hazardous environment of oil depots; thirdly, it lacks the ability to locate faults and predict potential hazards, and cannot accurately locate the branch circuits of the four grounding leads of a single tank. After a fault occurs, manual inspection is required section by section, which is inefficient. At the same time, it lacks the ability to predict the aging and corrosion of the grounding system, and can only respond to faults passively; fourthly, the early warning mechanism is imperfect. The existing technology only uses fixed threshold alarms and does not have a graded early warning design. The alarm response time exceeds 5 minutes, and there is no multi-channel push function, which makes it difficult to support rapid emergency response. At the same time, it cannot dynamically adjust the early warning threshold according to the safety level of different areas of the oil depot and severe weather, which easily leads to over-warning or under-warning. Summary of the Invention

[0004] The technical problem to be solved by this invention is to overcome the above-mentioned defects in the existing technology and provide an online monitoring system for the grounding resistance of oil depot storage tanks. This system enables non-contact online measurement through a distributed acquisition configuration of four measurement points per tank using a grounding resistance acquisition module. The signal conditioning module integrates an adaptive interference suppression algorithm, which can dynamically identify and suppress complex interferences such as power frequency interference and electromagnetic radiation. The main control module accurately separates the resistance values ​​of each lead-in line through matrix solving of a four-element linear equation system, and constructs an aging corrosion prediction model by combining random forest and XGBoost fusion algorithms to predict potential hazards. The early warning module adopts a three-level intelligent early warning mechanism with an alarm response time of ≤3 seconds, and can dynamically adjust the early warning threshold according to the regional safety level and severe weather. Data transmission is achieved through LoRa explosion-proof wireless communication, and the supporting upper computer platform supports multi-channel alarm push, data traceability, and intelligent management.

[0005] The technical solution adopted by the present invention to solve its technical problem is as follows: an online monitoring system for grounding resistance of oil depot storage tanks, the system comprising: a grounding resistance acquisition module, a signal conditioning module, a main control module, a wireless communication module, an early warning module, and a host computer monitoring platform; The grounding resistance acquisition module adopts a single tank with four measurement points distributed acquisition configuration, which is used to inject detection signals into each grounding lead-down circuit of the storage tank and acquire the corresponding voltage and current signals. The signal conditioning module is used to amplify, filter, and isolate the signals collected from each measuring point, and adopts an adaptive interference suppression algorithm to automatically identify the types of power frequency interference and equipment electromagnetic radiation interference on site, and dynamically adjust the filtering parameters. The main control module is used to receive standard signals, construct a four-element linear equation system based on the comprehensive resistance value of multiple measurement points to solve the actual resistance value of each down conductor in order to achieve accurate fault location of the branch circuit, construct a grounding system aging / corrosion prediction model based on historical data to predict the trend of hidden dangers, calculate the grounding resistance value, judge the threshold and generate control commands and early warning signals. The wireless communication module is used for data uploading and command reception; The early warning module is used to execute corresponding early warning operations based on a three-level intelligent early warning mechanism, with an alarm response time of ≤3 seconds; The host computer monitoring platform is used for data management, intelligent monitoring, and pushing alarm information through multiple channels.

[0006] Preferably, the grounding resistance acquisition module adopts a single-tank four-point distributed acquisition configuration. For the four independent grounding down conductors of a single oil depot tank, dedicated acquisition points are set up at the key connection nodes between each down conductor and the tank body and the grounding grid. Each acquisition point adopts a frequency-injection type dual-electrode non-decoupled measurement structure. Low-frequency detection signals are synchronously injected into the independent circuits corresponding to the four grounding down conductors through a frequency-injection excitation unit, and the original electrical signals of voltage and current of each circuit are acquired in real time.

[0007] Preferably, the main control module first collects the comprehensive grounding resistance data of four measuring points in a single tank. Combining the electrical relationship between each measuring point and the grounding down conductor, a four-element linear equation system is constructed with the actual resistance of the four grounding down conductors as unknown parameters. The measured comprehensive resistance value of each measuring point is used as the known input of the equation system. The equation system is modeled according to the electrical topology and conduction rules of the grounding loop. Then, the equation system is accurately solved by matrix solving and numerical iteration algorithms to obtain the real resistance value of each grounding down conductor one by one, clearly distinguishing the resistance state differences of each down conductor. After the solution is completed, the resistance value of a single down conductor is compared with the preset safety threshold one by one to quickly locate the down conductor number and specific fault location with excessive resistance. The accurate location of branch faults can be achieved without on-site inspection of each circuit, effectively improving the efficiency and accuracy of grounding system hidden danger investigation.

[0008] Preferably, the four-element linear equation system is as follows: ,in, R 1 , R 2 , R 3 , R 4 This represents the actual resistance values ​​of the first to fourth grounding leads of the storage tank. R 01 , R 02 , R 03 , R 04 The combined grounding resistance value is measured at four distributed measuring points. a ij ( i,j = 1,2,3,4 The coefficient matrix elements are determined by the electrical connection topology between the measuring point and the grounding down conductor, and the signal conduction path. The equations are solved precisely using matrix solving and numerical iteration algorithms, resulting in a simplified matrix form: AR = R 0 Solving by matrix inversion yields: R = A -1 R 0 ,in, A It is a fourth-order correlation coefficient matrix, which represents the electrical coupling relationship between each measuring point and each grounding down conductor; A -1 It is a coefficient matrix A The inverse matrix, R This is the column vector of grounding lead resistance, representing the final solution result. R 0It is the column vector of the measured comprehensive resistance at the measurement point.

[0009] Preferably, the main control module integrates real-time and historical resistance values ​​of each grounding down conductor, ambient temperature and humidity, and usage duration information of the grounding device based on long-term collected and stored historical monitoring data. It then uses machine learning algorithms to build a prediction model for the aging and corrosion of the grounding system. The latest monitoring data is continuously input into the model for iterative optimization, and the model fitting parameters are gradually corrected to improve the prediction accuracy. Through quantitative analysis of key indicators such as resistance change rate, environmental corrosion influencing factors, and equipment service life loss, the module deduces the evolution law of grounding corrosion degree and down conductor aging rate, predicts the performance degradation trend and potential hidden danger nodes of the grounding system in advance, and clarifies the time cycle and risk level of hidden danger development.

[0010] Preferably, the specific steps for the main control module to build a prediction model for the aging and corrosion of the grounding system using machine learning algorithms are as follows: Continuously collect historical resistance values ​​of each grounding down conductor, real-time temperature and humidity on site, commissioning time of grounding device and historical maintenance data, remove abnormal outliers, fill in missing data, and standardize and normalize data of different dimensions including resistance, temperature and humidity and time to eliminate the impact of data magnitude differences on model training. The monthly average resistance change rate, temperature and humidity corrosion weight, service time aging coefficient, and resistance fluctuation variance features were extracted from the preprocessed data. Pearson correlation analysis was used to screen out the feature parameters that are strongly correlated with the aging and corrosion of the grounding system. Redundant and invalid features were removed to determine the model input feature set. Using the selected effective features as model input, and the future change value of grounding resistance, the degree of corrosion and aging, and the probability of the occurrence of hidden dangers as model output, we initialize the random forest regression and XGBoost gradient boosting regression fusion model, and set the initial hyperparameters including the number of basic decision trees, learning rate, and maximum depth. The dataset was divided into training and testing sets in a 7:3 ratio. The fusion model was iteratively trained using the training set. Hyperparameters, including the number of decision trees, regularization parameters, and learning rate, were optimized through grid search and 5-fold cross-validation to minimize the mean squared error of prediction. The accuracy of the trained model was verified using test set data. The error between the predicted resistance value and the actual monitored value was compared. The model weight parameters were calibrated in combination with the actual working conditions of corrosion and aging of the grounding device at the oil depot. The calibrated fusion model is deployed to the main control module, and the latest monitoring data is accessed in real time for online predictive analysis. New monitoring data and fault cases are continuously accumulated, and the model is incrementally trained and its parameters are updated regularly.

[0011] Preferably, the main control module calculates the comprehensive grounding resistance value of the storage tank strictly according to Ohm's law based on the voltage and current signals acquired in real time from each acquisition point. At the same time, it compares the calculated comprehensive resistance value, the resistance value of each branch lead-down line with the dynamically adjusted warning threshold one by one, based on the comparison results and the risk level predicted by the model. According to the comparison results and the risk level predicted by the model, it generates corresponding equipment control commands, complete monitoring data and graded warning signals, and simultaneously completes the sorting and packaging of monitoring data and the triggering output of warning signals to ensure that the system accurately reflects the real-time status of the grounding system.

[0012] Preferably, the early warning module is used to execute corresponding early warning operations based on a three-level intelligent early warning mechanism, which specifically includes: Level 1 warning: When the grounding resistance value is between 8 and 10 Ω, a warning sound and light will be issued. Level 2 alarm: When the grounding resistance value is >10Ω, a warning audible and visual alarm will be issued; Emergency alarm: If the grounding resistance value is >30Ω, the equipment malfunctions, or the data transmission is interrupted, an emergency audible and visual alarm will be issued, with an alarm response time of ≤3 seconds.

[0013] Preferably, the early warning module is designed with an adaptive adjustment mechanism for early warning levels, which can dynamically adjust the early warning thresholds at each level according to the safety level of different areas of the oil depot and severe weather conditions.

[0014] The beneficial effects of the online monitoring system for grounding resistance of oil depot storage tanks of the present invention are as follows: This invention achieves non-contact online measurement through a distributed acquisition configuration of four measurement points per tank using a grounding resistance acquisition module. Utilizing matrix solving technology for a quaternary linear equation system, it can accurately separate the actual resistance values ​​of the four grounding leads per tank, enabling precise fault location and significantly improving fault diagnosis efficiency. A aging corrosion prediction model constructed using a fusion algorithm of random forest and XGBoost can predict potential grounding system hazards in advance based on historical monitoring data and real-time operating conditions. A three-level intelligent early warning mechanism is provided, reducing alarm response time to less than 3 seconds. The warning threshold can be dynamically adjusted according to the safety level of different areas of the oil depot and severe weather conditions, avoiding over-warning or under-warning. Accurate alarm information is delivered through multiple channels, including SMS, platform pop-ups, on-site audio-visual displays, and emergency terminal push notifications. Furthermore, LoRa explosion-proof wireless communication technology ensures stable operation and data transmission in the hazardous environment of the oil depot. The upper-level monitoring platform integrates remote monitoring, data traceability, and intelligent control functions, significantly improving the safety assurance capabilities and intelligent management level of the oil depot's tank grounding. Attached Figure Description

[0015] Figure 1This is a structural block diagram of an embodiment of the online monitoring system for grounding resistance of oil depot storage tanks of the present invention; Figure 2 This is a flowchart illustrating the construction of a prediction model for the aging and corrosion of a grounding system in an embodiment of the online monitoring system for grounding resistance of oil depot storage tanks according to the present invention. Figure 3 This is a flowchart of an embodiment of the online monitoring system for grounding resistance of oil depot storage tanks according to the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0017] Example This invention provides an online monitoring system for the grounding resistance of oil depot storage tanks, comprising: a grounding resistance acquisition module, a signal conditioning module, a main control module, a wireless communication module, an early warning module, and a host computer monitoring platform. The grounding resistance acquisition module employs a distributed acquisition configuration of four measurement points per tank to achieve non-contact online measurement. The signal conditioning module integrates an adaptive interference suppression algorithm, which can dynamically identify and suppress complex interferences such as power frequency interference and electromagnetic radiation. The main control module accurately separates the resistance values ​​of each downlead through matrix solving of a system of four linear equations, and constructs an aging corrosion prediction model by combining random forest and XGBoost fusion algorithms to predict potential hazards. The early warning module adopts a three-level intelligent early warning mechanism with an alarm response time ≤3 seconds, and can dynamically adjust the early warning threshold according to the regional safety level and severe weather. Data transmission is achieved through LoRa explosion-proof wireless communication, and the host computer platform supports multi-channel alarm push, data traceability, and intelligent management.

[0018] The grounding resistance acquisition module is deployed near key nodes of the tank grounding lead-down line. The signal conditioning module is integrated with the acquisition module and installed nearby to reduce signal transmission loss. The main control module and wireless communication module are integrated into an explosion-proof control box and installed in a safe area of ​​the tank area. The early warning module is deployed at the on-site duty points of the tank and the central control room of the oil depot. The upper computer monitoring platform is deployed in the central control room of the oil depot to realize remote centralized monitoring. All field modules are equipped with explosion-proof shells, anti-corrosion coatings and waterproof sealing structures to adapt to the flammable, explosive, humid, corrosive and dusty outdoor working conditions of the oil depot, ensuring long-term stable operation of the system.

[0019] The grounding resistance acquisition module adopts a distributed acquisition configuration with four measurement points per tank. It is used to inject detection signals into each grounding lead-down circuit of the storage tank and acquire the corresponding voltage and current signals. The specific implementation process is as follows: Measurement point layout rules: For the four independent grounding down conductors that are equipped for a single oil depot tank, one dedicated data acquisition point is set up at each of the four key electrical nodes: the tank connection end of each down conductor, the middle section of the down conductor, the connection end between the down conductor and the grounding grid, and the connection point of the main grounding grid. A total of four sets of independent data acquisition points are formed for a single tank, which fully covers the entire path of the grounding circuit.

[0020] Measurement structure design: Each measuring point adopts a frequency conversion injection type dual electrode non-desolvable measurement structure, which includes an excitation electrode and a detection electrode. It can realize online measurement without disassembling the grounding lead and without interrupting the grounding circuit, completely solving the safety hazards of traditional measurement that require disconnecting the electrolytic wire.

[0021] Signal injection method: The module has a built-in frequency conversion excitation unit that synchronously injects low-frequency detection signals into the independent circuits corresponding to the four grounding leads. The signal frequency avoids the 50Hz power frequency and harmonic frequency bands on site to avoid interference superposition and circuit resonance, thus ensuring the purity of the acquired signal.

[0022] Signal acquisition content: Real-time synchronous acquisition of raw AC voltage and AC current signals of each grounding circuit. Continuous sampling mode is adopted, and the sampling frequency matches the on-site monitoring requirements to ensure that the acquired data is complete, real-time, and synchronous, providing accurate basic data for subsequent resistance calculation.

[0023] The signal conditioning module is responsible for standardizing the raw weak signal and incorporates an adaptive interference suppression algorithm. The specific implementation process is as follows: Low-noise linear amplification: For the weak microvolt-level electrical signals output by the acquisition module, a low-noise amplification circuit is used to linearly increase the amplitude. There is no signal distortion or additional noise introduced during the amplification process, and the weak signal is amplified to the standard amplitude range that is compatible with the main control module.

[0024] Adaptive interference suppression: Real-time scanning of the electromagnetic interference spectrum at the site, automatic identification of types such as power frequency interference, equipment electromagnetic radiation interference, and ground loop interference, dynamically adjusting the filter bandwidth, cutoff frequency, and gain parameters, and accurately filtering out invalid interference signals through multi-stage filter circuits while retaining valid detection signals.

[0025] Dual electrical isolation: The dual isolation design, which combines opto-isolation and magnetic isolation, blocks interference from ground potential difference, surge voltage, and high-voltage interference, protecting the back-end main control module from damage by high-voltage electricity on site, while ensuring the electrical independence of signal transmission.

[0026] Standard signal output: The signal, after amplification, filtering and isolation, is converted into a stable standard electrical signal and transmitted to the main control module without drift or attenuation, providing reliable data support for resistance calculation and fault diagnosis.

[0027] The main control module receives standard signals, constructs a four-element linear equation system based on the comprehensive resistance values ​​of multiple measurement points to solve the actual resistance values ​​of each down conductor to achieve accurate fault location in the branch circuit, constructs a grounding system aging / corrosion prediction model based on historical data to predict potential hazards, calculates grounding resistance values, determines thresholds, and generates control commands and early warning signals. Its specific implementation process is as follows: Data Acquisition and Modeling: Real-time acquisition of comprehensive grounding resistance data from four measuring points on a single tank. This data, combined with the series and parallel electrical topology of the measuring points and grounding down conductors, and the signal conduction impedance characteristics, is used to model the actual resistance of the four grounding down conductors. R 1 , R 2 , R 3 , R 4 For unknown parameters, construct a system of four linear equations: ,in, R 1 , R 2 , R 3 , R 4 The actual resistance of the four grounding leads; R 01 , R 02 , R 03 , R 04 The combined resistance was measured at four measuring points. a ij ( i,j = 1,2,3,4 The coefficient matrix elements are determined through on-site grounding topology mapping and electrical parameter calibration.

[0028] Matrix solution calculation: simplifying the system of equations into matrix form. AR = R 0 Obtained by matrix inversion R = A -1 R 0 It can quickly and accurately calculate the true resistance value of each grounding lead, and the calculation error meets the accuracy requirements of industrial monitoring.

[0029] Precise fault location: The resistance value of each lead wire is compared with the preset safety threshold one by one, and the lead wire number and physical location are automatically matched to directly lock the lead wire with excessive resistance and the fault node. There is no need to check each circuit on site, realizing the visualization and precise location of branch circuit faults.

[0030] The main control module uses machine learning algorithms to build an aging and corrosion prediction model based on multi-dimensional historical data, and the entire process is executed automatically. Specific implementation steps are as follows: Multi-dimensional data acquisition: Continuously collects historical resistance values ​​of each grounding down conductor, real-time temperature and humidity on site, soil resistivity, rainfall duration, corrosive gas concentration, grounding device commissioning duration, historical maintenance records, and other comprehensive data.

[0031] Data preprocessing: Outliers were removed using the Laida criterion, and missing data were filled in using linear interpolation; for data of different dimensions such as resistance, temperature, humidity, and duration, min-max normalization was used to eliminate the impact of differences in data magnitude on model training.

[0032] Feature selection optimization: Extract features such as monthly average resistance change rate, temperature and humidity corrosion weight, service life aging coefficient, and resistance fluctuation variance. Set thresholds through Pearson correlation analysis, retain feature parameters that are strongly correlated with grounding aging and corrosion, eliminate redundant and invalid features, and determine the optimal model input feature set.

[0033] Fusion model training: Using the selected effective features as input and the future change value of grounding resistance, corrosion aging level, and probability of hidden danger occurrence as output, initialize the random forest regression + XGBoost gradient boosting regression fusion model; divide the dataset into training and test sets in a 7:3 ratio, and optimize hyperparameters such as the number of decision trees, learning rate, and regularization parameters through grid search and 5-fold cross-validation to minimize the prediction mean square error and avoid model overfitting.

[0034] Model calibration and deployment: The accuracy of the model is verified using test set data, and the model weight parameters are calibrated in combination with the actual working conditions of corrosion and aging of the grounding device at the oil depot. The calibrated model is then deployed to the main control module and connected to the latest monitoring data in real time for online prediction.

[0035] Incremental updates and trend projections: Continuously accumulate new monitoring data and fault cases, and incrementally train and update the model on a monthly / quarterly basis; by quantitatively analyzing indicators such as resistance change rate, environmental corrosion factor, and equipment cycle loss, project the evolution of grounding electrode corrosion degree and down conductor aging rate, and predict the development cycle and risk level of hidden dangers in advance.

[0036] Based on the collected voltage and current signals, the main control module calculates the comprehensive grounding resistance and instantaneous resistance of the storage tank in real time according to Ohm's law, and also calculates the average resistance per unit time to improve data stability.

[0037] By combining the potential hazard trends output by the aging corrosion prediction model, and dynamically adjusting the early warning threshold based on the safety level of the oil depot area and severe weather conditions, instead of using a fixed threshold, the adaptability of the early warning is improved.

[0038] By comparing the overall resistance, branch lead resistance, and dynamic threshold, and combining them with the predicted risk level, control commands for the acquisition module, trigger commands for the early warning module, and data upload commands are generated. At the same time, complete monitoring data and graded early warning signals are packaged and transmitted synchronously to the wireless communication module and the early warning module.

[0039] The wireless communication module is responsible for bidirectional data interaction between the system and the host computer monitoring platform. It adopts a dual communication link backup design to ensure transmission reliability. Communication method: It adopts dual links of industrial-grade wireless cellular communication and wireless local area network communication. When the primary link fails, it automatically switches to the backup link to avoid data interruption.

[0040] Data transmission rules: Real-time monitoring data is uploaded at fixed intervals, while fault and early warning data are uploaded instantly to ensure that alarm information is not delayed; all transmitted data uses encryption protocols to meet the data security requirements of the oil depot.

[0041] Command reception and execution: Real-time reception of threshold adjustment, parameter configuration, remote control, and model update commands issued by the host computer, and immediate transmission to the main control module for parsing and execution, realizing remote closed-loop management and control.

[0042] Environmental adaptability: The communication module is explosion-proof and anti-electromagnetic interference, adapting to the complex communication environment of the tank farm, ensuring stable data transmission without packet loss.

[0043] The early warning module adopts a three-level intelligent early warning mechanism, coupled with a threshold adaptive adjustment function, with an alarm response time of ≤3 seconds. Its three-level early warning execution logic is as follows: Level 1 warning: When the grounding resistance value is between 8 and 10Ω, a warning sound and light will be triggered, with the light flashing slowly at a low frequency and the buzzer sounding intermittently, serving only as a status indication.

[0044] Level 2 warning: Grounding resistance value > 10Ω, triggering a warning audible and visual alarm, with lights flashing rapidly at medium frequency and a buzzer sounding continuously to remind maintenance personnel to pay attention.

[0045] Emergency Warning: If the grounding resistance value is greater than 30Ω, equipment failure, or data transmission interruption occurs, an emergency audible and visual alarm will be triggered. The lights will remain on at a high frequency, the buzzer will sound at a high decibel level, and the alarm terminal in the central control room will be activated simultaneously.

[0046] Rapid response design: The early warning signal is triggered directly by hardware circuitry with no software delay, ensuring a response time of strictly ≤3 seconds, which meets the safety and emergency requirements of oil depots.

[0047] Threshold adaptive adjustment: It can automatically tighten or loosen the warning thresholds at each level according to the high / medium / low safety level area of ​​the oil depot, severe weather such as thunderstorms / high temperature and humidity, etc., to adapt to different on-site control needs.

[0048] Early warning record storage: All early warning information is stored in real time, including the early warning time, storage tank number, down conductor point position, resistance value, and early warning level, facilitating subsequent operation and maintenance traceability.

[0049] The upper computer monitoring platform is the core of remote control, realizing the integrated functions of data management, visual monitoring, alarm push, and operation and maintenance management: Data storage: Adopting a dual-database architecture of real-time database + historical database, storing all monitoring data, early warning records, fault location results, model parameters, and equipment status information, and the data storage duration meets the industry operation and maintenance requirements.

[0050] Visual monitoring: Through the storage tank distribution schematic diagram, real-time grounding resistance value, historical change curve, fault point marking, and prediction trend chart, intuitively display the full status of the grounding system, and support single-tank / branch data switching and viewing.

[0051] Intelligent alarm push: Adopting multi-channel linkage push, including on-site sound and light linkage, pop-up window on the central control room large screen, SMS push, mobile APP notification, and email alarm, ensuring that alarm information reaches operation and maintenance personnel without omission.

[0052] Operation and maintenance management: Support equipment status monitoring, automatic generation of maintenance plans, historical data query, report export, and fault statistical analysis, assisting operation and maintenance personnel in efficient management.

[0053] Parameter configuration: Support adjustment of early warning thresholds, modification of model parameters, setting of acquisition frequencies, and configuration of communication parameters to adapt to different on-site requirements; configure hierarchical permission management to distinguish the permissions of administrators, operation and maintenance personnel, and observers, ensuring operation safety.

[0054] As Figure 3 shown, the specific working process of an on-line monitoring system for the grounding resistance of oil storage tanks provided by the present invention is as follows: System initialization: After the system is powered on, each module completes self-checking in sequence. After the status checks of the acquisition, conditioning, main control, communication, and early warning modules are normal, it automatically enters the all-weather monitoring mode.

[0055] Signal acquisition and injection: The grounding resistance acquisition module synchronously injects low-frequency detection signals into 4 down conductor loops, and the double-electrode structure collects the original voltage and current signals of each loop in real time.

[0056] Signal standardization processing: The original signal is transmitted to the signal conditioning module, which completes low-noise amplification, adaptive filtering, and double electrical isolation, and outputs a standard signal to the main control module.

[0057] Core data calculation: The main control module calculates the comprehensive resistance according to Ohm's law, constructs a four-element linear equation system to calculate the resistance of the branch down conductors, and at the same time calls the prediction model to complete the prediction of the aging and corrosion trend.

[0058] Early warning and command generation: The main control module compares the calculated data, prediction results and dynamic thresholds to generate graded early warning signals and control commands.

[0059] On-site warning trigger: The warning module receives the signal and immediately triggers an audible and visual warning according to the three-level mechanism, with a response time of ≤3 seconds.

[0060] Remote data transmission: The wireless communication module encrypts and uploads monitoring data, early warning information, and fault location results to the host computer platform.

[0061] Remote control closed loop: The host computer platform completes data storage, visualization display, and multi-channel alarm push, and can also send remote commands to the main control module to realize the closed loop of the entire process of monitoring-early warning-control.

[0062] Continuous model optimization: The system executes the monitoring process in a loop, regularly accumulating new data to incrementally train the prediction model, continuously improving prediction accuracy and system stability.

[0063] This implementation method, through a full-process design including distributed acquisition, adaptive signal processing, precise calculation of branch resistance, intelligent hazard prediction, and hierarchical early warning, completely solves the problems of low efficiency, inability to monitor online, difficulty in fault location, and delayed hazard prediction in traditional oil depot storage tank grounding resistance manual measurement, comprehensively improving the safety operation and maintenance level and emergency response capability of oil depot storage tank grounding system.

Claims

1. An on-line monitoring system for grounding resistance of oil depot storage tank, characterized in that, The system includes: a grounding resistance acquisition module, a signal conditioning module, a main control module, a wireless communication module, an early warning module, and a host computer monitoring platform; The grounding resistance acquisition module adopts a single tank with four measurement points distributed acquisition configuration, which is used to inject detection signals into each grounding lead-down circuit of the storage tank and acquire the corresponding voltage and current signals. The signal conditioning module is used to amplify, filter, and isolate the signals collected from each measuring point, and adopts an adaptive interference suppression algorithm to automatically identify the types of power frequency interference and equipment electromagnetic radiation interference on site, and dynamically adjust the filtering parameters. The main control module is used to receive standard signals, construct a four-element linear equation system based on the comprehensive resistance value of multiple measurement points to solve the actual resistance value of each down conductor in order to achieve accurate fault location of the branch circuit, construct a grounding system aging / corrosion prediction model based on historical data to predict the trend of hidden dangers, calculate the grounding resistance value, judge the threshold and generate control commands and early warning signals. The wireless communication module is used for data uploading and command reception; The early warning module is used to execute corresponding early warning operations based on a three-level intelligent early warning mechanism, with an alarm response time of ≤3 seconds; The host computer monitoring platform is used for data management, intelligent monitoring, and pushing alarm information through multiple channels.

2. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1, characterized in that, The grounding resistance acquisition module adopts a distributed acquisition configuration of four measurement points per tank. For the four independent grounding down conductors of a single oil depot tank, dedicated acquisition measurement points are set up at the key connection nodes between each down conductor and the tank body and the grounding grid. Each measurement point adopts a frequency-injection type dual-electrode non-deinterlocking measurement structure. Low-frequency detection signals are synchronously injected into the independent circuits corresponding to the four grounding down conductors through the frequency-injection excitation unit, and the original electrical signals of voltage and current of each circuit are acquired in real time.

3. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1 or 2, characterized in that, The main control module first collects the comprehensive grounding resistance data of four measuring points in a single tank. Combining the electrical relationship between each measuring point and the grounding down conductor, it constructs a four-element linear equation system with the actual resistance of the four grounding down conductors as unknown parameters. The measured comprehensive resistance value of each measuring point is used as the known input of the equation system. The equation system is modeled according to the electrical topology and conduction rules of the grounding loop. Then, the equation system is accurately solved by matrix solving and numerical iteration algorithms to obtain the real resistance value of each grounding down conductor. The differences in resistance status of each down conductor are clearly distinguished. After the solution is completed, the resistance value of each down conductor is compared with the preset safety threshold one by one to quickly locate the down conductor number and specific fault location with excessive resistance. The accurate location of branch faults can be achieved without on-site inspection of each circuit, effectively improving the efficiency and accuracy of grounding system hidden danger investigation.

4. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 3, characterized in that, The system of four linear equations is as follows: ,in, R 1 , R 2 , R 3 , R 4 This represents the actual resistance values ​​of the first to fourth grounding leads of the storage tank. R 01 , R 02 , R 03 , R 04 The combined grounding resistance value is measured at four distributed measuring points. a ij ( i, j = 1,2,3,4 The coefficient matrix elements are determined by the electrical connection topology between the measuring point and the grounding down conductor, and the signal conduction path. The equations are solved precisely using matrix solving and numerical iteration algorithms, resulting in a simplified matrix form: AR = R 0 Solving by matrix inversion yields: R = A -1 R 0 ,in, A It is a fourth-order correlation coefficient matrix, which represents the electrical coupling relationship between each measuring point and each grounding down conductor; A -1 It is a coefficient matrix A The inverse matrix, R This is the column vector of grounding lead resistance, representing the final solution result. R 0 It is the column vector of the measured comprehensive resistance at the measurement point.

5. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1 or 2, characterized in that, The main control module integrates real-time and historical resistance values ​​of each grounding down conductor, ambient temperature and humidity, and usage duration information of the grounding device based on long-term collected and stored historical monitoring data. It uses machine learning algorithms to build a prediction model for the aging and corrosion of the grounding system. The latest monitoring data is continuously input into the model for iterative optimization, and the model fitting parameters are gradually corrected to improve the prediction accuracy. Through quantitative analysis of key indicators such as resistance change rate, environmental corrosion influencing factors, and equipment service life loss, the module deduces the evolution law of grounding body corrosion degree and down conductor aging rate, predicts the performance degradation trend and potential hidden danger nodes of the grounding system in advance, and clarifies the time cycle and risk level of hidden danger development.

6. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 5, characterized in that, The specific steps for the main control module to build a prediction model for the aging and corrosion of the grounding system using machine learning algorithms are as follows: Continuously collect historical resistance values ​​of each grounding down conductor, real-time temperature and humidity on site, commissioning time of grounding device and historical maintenance data, remove abnormal outliers, fill in missing data, and standardize and normalize data of different dimensions including resistance, temperature and humidity and time to eliminate the impact of data magnitude differences on model training. The monthly average resistance change rate, temperature and humidity corrosion weight, service time aging coefficient, and resistance fluctuation variance features were extracted from the preprocessed data. Pearson correlation analysis was used to screen out the feature parameters that are strongly correlated with the aging and corrosion of the grounding system. Redundant and invalid features were removed to determine the model input feature set. Using the selected effective features as model input, and the future change value of grounding resistance, the degree of corrosion and aging, and the probability of the occurrence of hidden dangers as model output, we initialize the random forest regression and XGBoost gradient boosting regression fusion model, and set the initial hyperparameters including the number of basic decision trees, learning rate, and maximum depth. The dataset was divided into training and testing sets in a 7:3 ratio. The fusion model was iteratively trained using the training set. Hyperparameters, including the number of decision trees, regularization parameters, and learning rate, were optimized through grid search and 5-fold cross-validation to minimize the mean squared error of prediction. The accuracy of the trained model was verified using test set data. The error between the predicted resistance value and the actual monitored value was compared. The model weight parameters were calibrated in combination with the actual working conditions of corrosion and aging of the grounding device at the oil depot. The calibrated fusion model is deployed to the main control module, and the latest monitoring data is accessed in real time for online predictive analysis. New monitoring data and fault cases are continuously accumulated, and the model is incrementally trained and its parameters are updated regularly.

7. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1 or 2, characterized in that, The main control module calculates the comprehensive grounding resistance of the storage tank based on the voltage and current signals acquired in real time from each measurement point, strictly following Ohm's law. Simultaneously, it combines the predicted trend of potential hazards output by the grounding system aging and corrosion prediction model. The calculated comprehensive resistance value, the resistance value of each branch lead-out line, and the dynamically adjusted warning threshold are compared one by one. Based on the comparison results and the risk level predicted by the model, corresponding equipment control commands, complete monitoring data, and graded warning signals are generated. The monitoring data is organized and packaged simultaneously, and the warning signals are triggered and output, ensuring that the system accurately reflects the real-time status of the grounding system.

8. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1 or 2, characterized in that, The early warning module is used to execute corresponding early warning operations based on a three-level intelligent early warning mechanism, which is specifically as follows: Level 1 warning: When the grounding resistance value is between 8 and 10 Ω, a warning sound and light will be issued. Level 2 alarm: When the grounding resistance value is >10Ω, a warning audible and visual alarm will be issued; Emergency alarm: If the grounding resistance value is >30Ω, the equipment malfunctions, or the data transmission is interrupted, an emergency audible and visual alarm will be issued, with an alarm response time of ≤3 seconds.

9. The online monitoring system for grounding resistance of oil depot storage tanks according to claim 1 or 2, characterized in that, The early warning module is designed with an adaptive adjustment mechanism for early warning levels, which can dynamically adjust the early warning thresholds at each level according to the safety level of different areas of the oil depot and severe weather conditions.

Images