Spherical tank structure damage identification method based on optical fiber sensing network information

Abstract

The invention discloses a spherical tank structure damage identification method based on optical fiber sensing network information. In the service process of a spherical tank, damage such as local cracks, corrosion and fatigue cracking may occur, and a traditional detection method has the problems of low efficiency, limited range, low sensitivity to tiny damage and the like. According to the method, the load and damage characteristics of the spherical tank structure are analyzed, quasi-distributed optical fiber sensing devices and layout are designed, an effective multi-scale optical fiber sensing information network is established, and the damage of the spherical tank structure is identified by sensing strain and temperature information, a structure theory and a data analysis method. A spherical tank structure damage source mode recognition method is developed based on continuously monitored mass data information and a BP neural network algorithm, damage types, positions and degrees are determined, and corresponding damage sensitivity indexes and early warning thresholds are established. According to the invention, the problem of difficulty in positioning and identifying the damage of the spherical tank structure is solved, the accuracy and efficiency of online monitoring are improved, and reliable technical guarantee is provided for safe operation of the spherical tank.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent structural health monitoring, and relates to a spherical tank structure damage diagnosis and identification method driven by optical fiber sensing network information. Background Art

[0002] As an important pressure-bearing vessel, spherical tanks are widely used in the petroleum, chemical, and gas industries to store various flammable and explosive media, such as liquefied gas and liquefied petroleum gas. However, during long-term operation, spherical tanks are susceptible to damage such as localized cracks, corrosion, and fatigue cracking due to environmental factors and the storage and transportation media. These damages can lead to serious leakage accidents, polluting the environment while causing significant economic losses and casualties. Therefore, conducting long-term, continuous, and effective damage detection and health monitoring of spherical tank structures to promptly identify and address potential hidden defects is extremely important for ensuring the safe operation of spherical tanks.

[0003] Traditional spherical tank inspection methods mainly include macroscopic inspection, non-destructive testing (such as ultrasonic testing, X-ray testing, etc.), and stress analysis-based methods. Macroscopic inspection mainly relies on manual visual inspection of the appearance of the spherical tank. Although it is simple and direct, it is difficult to detect minor internal damage and early damage, and the detection range is limited, making it impossible to fully assess the health of the spherical tank. Non-destructive testing technology is widely used in spherical tank inspection. For example, ultrasonic testing can detect defects inside the spherical tank, and X-ray testing can intuitively display the damage of the internal structure. However, these methods usually require shutdown inspection, have low detection efficiency, and are limited in detection range. Some difficult-to-reach areas or minor damage may not be accurately detected. In addition, these traditional detection methods can only provide detection results at a certain moment, and cannot achieve real-time and continuous monitoring of the spherical tank structure, making it difficult to capture the occurrence and development of damage in a timely manner.

[0004] Fiber optic sensors offer advantages such as high sensitivity, immunity to electromagnetic interference, fire and explosion resistance, waterproofness and moisture resistance, absolute measurement, corrosion resistance, compact size, light weight, and easy integration and networking. They enable high-precision, continuous measurement of physical quantities such as structural strain, temperature, and pressure. Fiber Bragg grating (FBG) strain sensors, in particular, can accurately and precisely sense changes in structural strain. Encapsulated FBG sensors effectively improve their survivability, measurement sensitivity, and stability, and coefficient calibration can be used to correct for strain transfer errors caused by encapsulation. These advantages make FBG sensors an ideal choice for spherical tank structural damage monitoring. In addition to advanced sensor technology, finite element simulation analysis, which establishes a finite element model of the tank's service state, can predict stress distribution under different operating conditions, providing a theoretical basis for optimal sensor placement. However, while existing detection technologies and methods have achieved some success in spherical tank damage detection, some shortcomings remain. For example, achieving comprehensive, real-time monitoring of the entire tank structure without disrupting normal operation remains an urgent challenge.

[0005] Therefore, this paper develops a spherical tank structure damage identification method based on fiber optic sensing network information. It utilizes the high sensitivity, high precision and stability of FBG sensors and the flexible integrated monitoring capabilities of the sensing network, combined with advanced signal data processing and damage identification algorithms, to achieve rapid, accurate and online continuous tracking and monitoring of spherical tank structure damage, which has important practical significance and broad application prospects for the safe operation and maintenance management of spherical tank structures and disaster risk warning. Summary of the Invention

[0006] The present invention discloses a method for identifying damage to spherical tank structures based on fiber optic sensor network information, providing an effective, reliable, and stable sensing technology for local and full-scale and full-process monitoring of spherical tank structures. A durable fiber optic sensor network is developed, and structural theory and data processing algorithms are integrated to achieve intelligent characterization of spherical tank service status information. The method combines source pattern recognition methods to determine the type, location, and extent of damage, and proposes indicators such as damage index, damage location deviation, and warning threshold, providing an effective and reliable early warning mechanism for spherical tank structure service safety risks. This method solves the problem of difficulty in locating and identifying damage to spherical tank structures, improves the accuracy and efficiency of detection, and provides a scientific, reliable, feasible, and low-cost technical guarantee and information-based and intelligent management method for the safe operation of spherical tanks.

[0007] The technical solution of the present invention is:

[0008] A spherical tank structure damage identification method based on fiber optic sensor network information is characterized by: designing the structural form, multi-parameter sensing capability, solar radiation protection function, and full-scale sensing network formation mode covering both local and overall areas based on the structural and material characteristics, stress distribution patterns, and possible damage types of the spherical tank; designing a dense sensor point layout for high-damage areas (such as welding locations) and critical locations (such as the top and bottom); integrating multiple quasi-distributed FBG sensors in series and parallel to form a highly redundant multi-scale fiber optic sensing network to improve the robustness of the sensing network, allowing for partial failure of some measuring points, and using a correlation reconstruction algorithm to achieve real-time continuous monitoring of the full-scale and local multi-parameter information of the spherical tank structure; identifying spherical tank damage characteristics based on sensed strain and temperature information, combining signal filtering and noise reduction technology, structural theory, and data processing algorithms, and establishing a source pattern recognition method for the spherical tank structure using a BP neural network to quantify damage; and establishing a damage sensitivity index and risk warning threshold based on continuous monitoring data information and the source pattern recognition method in combination with statistical analysis to provide effective and reliable data and technical support for the safe operation, online diagnosis of damage, and safety risk warning of the spherical tank structure.

[0009] The FBG sensor includes a base, two sensing fibers mounted on the base, and a stretching rod. The two sensing fibers are arranged symmetrically in a cross pattern, with several FBGs attached to each fiber. Stretching rods are located at both ends of each fiber, and the strain sensitivity of the FBG sensor is calibrated at the stretching rods. The two cross-shaped sensing fibers can simultaneously monitor longitudinal and transverse strain and temperature changes and are encapsulated in flexible silicone rubber.

[0010] The FBG sensors are positioned in areas prone to damage, such as stress concentration areas on the shell, the connection between the support and shell, weld locations, around manholes, and at the shell's equator. This optimizes sensor placement and determines the optimal placement and quantity of sensors to achieve precise monitoring of weak points in the overall tank structure while avoiding interference between sensors. The FBG sensors are connected to the transmission optical fiber using flanges or welding, ensuring a reliable connection and low signal attenuation, making the fiber-optic sensing network easy to replace, maintain, and upgrade.

[0011] The damage identification method utilizes signals collected by FBG sensors and performs preprocessing operations such as noise removal and filtering. Based on the full-process variation information of the spherical tank structure, such as strain and temperature, inverted based on strain transfer theory and heat conduction theory, structural theory and data analysis are combined to identify the damage characteristics of the spherical tank. A source pattern recognition method is established using a BP neural network optimization model to accurately identify the location and extent of damage. Based on massive monitoring information and an analysis method driven by data-physical models, a damage sensitivity index and risk warning threshold are established. The damage location and extent of the spherical tank under test are compared and analyzed. When the damage sensitivity index and risk warning threshold are reached, a warning signal is issued.

[0012] Effects and Benefits of the Invention: This invention provides a method for identifying structural damage in spherical tanks based on information from a fiber-optic sensing network. Compared to traditional detection methods, it offers advantages such as continuous online monitoring and real-time early warning of damage and disaster risks. FBG sensors, with their high sensitivity and precision, can accurately sense minute strains and temperature changes in the tank structure. By rationally arranging sensors and establishing an efficient fiber-optic sensing network, multi-point real-time monitoring is achieved, improving both comprehensiveness and accuracy. This method enables real-time monitoring during normal tank operation, avoiding the safety hazards associated with traditional detection. It also promptly detects and addresses potential damage, reducing accident risks, extending the tank's service life, and lowering replacement costs. Combining signal processing and pattern recognition technologies, it automatically identifies damage states, accurately locates and assesses the extent of damage, and implements intelligent monitoring and early warning. FBG sensors are resistant to electromagnetic interference and corrosion, adapting to complex operating conditions. This method is applicable to tanks of varying specifications and environmental conditions. In the long term, this method reduces maintenance frequency and repair costs, extending equipment life, and significantly reducing maintenance costs throughout the tank's lifecycle. Furthermore, the monitoring system collects and stores operational data in real time, providing a scientific basis for maintenance management and further improving operational efficiency and management. In summary, the present invention effectively solves the problem of structural damage detection of spherical tanks, provides reliable protection for the safe operation of spherical tanks, and has broad application prospects and important practical significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic diagram of a cross-shaped FBG sensor;

[0014] Figure 2 This is a schematic diagram of a spherical tank structure damage monitoring system based on optical fiber sensor network information;

[0015] Figure 3 It is a schematic diagram of BP neural network;

[0016] Figure 4 This is a flow chart of a spherical tank structure damage identification method based on optical fiber sensor network information;

[0017] In the figure: 1. Base; 2. FBG; 3. Tensile rod; 4. Cross-shaped FBG sensor; 5. Spherical tank; 6. Transmission optical fiber; 7. Grating fiber demodulator; 8. Data processing center; 9 Around the manhole; 10. Equatorial area; 11. Connection between the pillar and the spherical tank; 12. Input layer; 13. Hidden layer; 14. Output layer. DETAILED DESCRIPTION

[0018] The technical solution of the present invention is described in detail below with reference to the accompanying drawings.

[0019] During the service process, spherical tanks may suffer from local cracks, corrosion, fatigue cracking and other damages. Traditional detection methods have problems such as low efficiency, limited range, and low sensitivity to minor damage. The purpose of the present invention is to provide a spherical tank structure damage identification system and method based on fiber optic sensing network information. By analyzing the load and damage characteristics of the spherical tank structure, quasi-distributed fiber optic sensor devices and layouts are designed, and an effective multi-scale fiber optic sensing information network is established. The perceived strain, temperature information, structural theory and data analysis methods are used to identify spherical tank structure damage. Based on the massive data information of continuous monitoring and the BP neural network algorithm, a spherical tank structure damage source pattern recognition method is developed to determine the damage type, location and degree, and establish corresponding damage sensitivity indicators and warning thresholds to achieve timely warnings. The present invention solves the problem of difficulty in locating and identifying spherical tank structure damage, improves the accuracy and efficiency of online monitoring, and provides a reliable technical guarantee for the safe operation of spherical tanks.

[0020] like Figure 1-Figure 2 The spherical tank structure damage identification system based on optical fiber sensing network information provided by the present invention includes: multiple cross-shaped FBG sensors 4, a fiber grating demodulator 7 and a data processing center 8; the multiple FBG sensors 4 are respectively arranged in the damage monitoring area of the spherical tank 5 to be tested, and the damage monitoring area includes the equatorial part 10, the area around the manhole 9, the welding position and the connection between the pillar and the spherical shell 11; the multiple FBG sensors 4 set in each damage monitoring area are connected in series through transmission optical fibers 6 to form an optical fiber sensing network; the transmission optical fibers 6 corresponding to the multiple damage monitoring areas are respectively connected to the fiber grating demodulator 7, and the fiber grating demodulator 7 is connected to the data processing center 8; the multiple FBG sensors 4 set in the equatorial part 10 are connected in series through the first transmission optical fiber, the multiple FBG sensors 4 set in the area around the manhole 9 are connected in series through the second transmission optical fiber, and the multiple FBG sensors 4 set in the connection between the pillar and the spherical shell 11 are connected in series through the third transmission optical fiber; the first transmission optical fiber, the second transmission optical fiber, and the third transmission optical fiber are respectively connected to the fiber grating demodulator 7.

[0021] like Figure 3The BP neural network includes an input layer, hidden layers, and an output layer. The three characteristics of longitudinal and transverse strain and temperature change serve as input data for the nodes in the input layer 12. Two hidden layers 13 process the input data, and the output layer 14 consists of two nodes that output the damage location and damage extent. The number of input layer nodes can be set based on the number of FBG sensors.

[0022] like Figure 4 The present invention provides a method for identifying damage to a spherical tank structure based on optical fiber sensing network information, comprising the steps of:

[0023] Based on the structural characteristics and stress distribution of the spherical tank to be tested, the monitoring area is determined, including high-incidence areas and key parts. The number and arrangement spacing of FBG sensors are determined according to the specific size of the spherical tank, and the FBG sensors are deployed to form a fiber optic sensing network. Each FBG sensor monitors the longitudinal and transverse strain and temperature changes in real time and transmits the monitored reflection spectrum signal to the fiber grating demodulator. The fiber grating demodulator receives the monitored real-time reflected light wavelength increment signal Δλ B (t), transmitted to the data processing center, and converted into the real-time strain ε(t) of the structure using the following formula;

[0024]

[0025] The data processing center extracts effective signal features reflecting strain and temperature changes, inputs them into the constructed source pattern recognition method, obtains the damage location and degree of the spherical tank to be tested, and establishes damage sensitivity indicators and risk warning thresholds.

[0026] Through the above-mentioned spherical tank structure damage identification method based on optical fiber sensing network information, strain and temperature information are collected in real time. The damage source pattern recognition method is constructed by using perception information, filtering and noise reduction processing, structural theory, data processing algorithm and BP neural network analysis data to determine the damage location and extent, thereby realizing automatic identification and precise positioning of the damage.

[0027] A spherical tank structural damage identification method based on fiber optic sensor network information relies on the need to simultaneously monitor longitudinal and transverse strain and temperature changes during operation. A cross-shaped FBG sensor, consisting of multiple FBG strings, is designed to simultaneously acquire structural response information in both directions, providing more comprehensive data support for subsequent damage identification. The FBG sensors are encapsulated in flexible silicone rubber to withstand the harsh environment of spherical tanks, including high temperature, high pressure, and corrosion. The placement of the FBG sensors is determined based on the tank's structural characteristics and stress distribution. Sensors are placed in key locations, such as stress concentration areas, support connections, manholes, and welds, to fully cover the tank's critical areas and capture its structural response under complex operating conditions. To achieve efficient data transmission and centralized processing, the sensors are connected via optical fibers to form a star topology, based on the efficiency and reliability requirements of the fiber optic sensor network. A central data processing unit serves as the core for centralized data collection and processing. A fiber grating interrogator is used to collect the reflected light wave signals from the sensors in real time. In order to improve the accuracy and availability of the data, the collected data was preprocessed by denoising, filtering and normalization to extract effective signal features reflecting strain and temperature changes. Combined with the theoretical analysis results of the spherical tank structure, a damage source pattern recognition method was established using a BP neural network. The preprocessed data was analyzed according to the source pattern recognition method to achieve automatic identification and precise positioning of damage. By comparing the strain and temperature data of different sensors, abnormal strain distribution and temperature change patterns were identified to determine whether the spherical tank is damaged and the location and extent of the damage. According to the needs of monitoring and early warning, a software platform with data storage, management, analysis and display functions was developed. A reasonable monitoring plan and early warning threshold were formulated. When the monitoring data exceeded the preset threshold, the system automatically issued an early warning signal to provide maintenance personnel with detailed information such as the location and extent of the damage to assist in formulating a maintenance plan.

[0028] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A method for identifying structural damage of a spherical tank based on optical fiber sensing network information, characterized by: According to the damage monitoring requirements of the spherical tank (5), a cross-shaped FBG sensor (4) is designed to simultaneously monitor longitudinal and transverse strain and temperature changes; according to the structural characteristics and stress distribution law of the spherical tank, the sensor arrangement position is determined, and the FBG sensors (4) are arranged at the equator (10) of the spherical tank, around the manhole (9), and at the connection between the support and the spherical shell (11), and the FBG sensors (4) are connected in series using a transmission optical fiber (6); according to the specific size of the spherical tank, the number of sensors and the arrangement spacing are determined to form an optical fiber sensing network, and the transmission optical fiber (6) is connected to a distributed fiber Bragg grating demodulator (7) and a data processing center (8) to form a monitoring system, which collects strain and temperature information in real time, uses perception information, structural theory and BP neural network data analysis to identify damage, determine the damage location and degree, and realizes monitoring and early warning of the spherical tank through automatic recognition and precise positioning by algorithm; The cross-shaped FBG sensor (4) is composed of two groups of mutually perpendicular FBGs (2), each unit containing multiple FBGs (2), which are fixed on a base (1), and the strain sensitivity of the sensor is calibrated at a stretching rod (3), and can simultaneously monitor longitudinal and transverse strain and temperature changes; the FBG sensor (4) is encapsulated with flexible silicone rubber that is suitable for high temperature, high pressure, and corrosive environments; The optical fiber sensing network is composed of a plurality of cross-shaped FBG sensors (4) connected to a distributed fiber Bragg grating demodulator (7) and a data processing center (8) via transmission optical fibers (6), thereby realizing centralized data collection and processing; the transmission optical fiber connectors are flange-connected or welded, thereby making the optical fibers easy to replace, maintain, and upgrade; The BP neural network refers to taking the strain and temperature monitored by the FBG sensor as input data into the BP neural network for data processing, and outputting the predicted damage location and damage degree respectively; according to the number of sensors, the number of input layer nodes is set, each sensor collects transverse and longitudinal strain and temperature characteristics as input data of the input layer (12) node, and two hidden layers (13) are designed to process the captured data. Finally, the output layer (14) outputs the damage location and damage degree of the spherical tank by two nodes.

Images