Magnetic flux leakage detection system with optimized layout of magnetic sensor array

Abstract

The invention relates to the technical field of ferromagnetic material detection, and discloses a magnetic flux leakage detection system with optimized layout of a magnetic sensor array. The system comprises a magnetic sensor array layout feature extraction module, a region detection integrity analysis module, a multistage optimization layout generation module, an array reconstruction interference analysis module and an optimal layout decision updating module. The magnetic sensing array layout feature extraction module acquires surface space magnetic field distribution data of the ferromagnetic material in real time, and constructs a defect and detection environment feature information set; the area detection integrity analysis module analyzes the detection integrity in combination with the initial layout parameters, the detection path and the working state; a multi-level optimization layout generation module generates a multi-level optimization scheme set based on defect features when the integrity is lower than a threshold value; the array reconstruction interference analysis module calculates a reconstruction interference value during layout switching; and the optimal layout decision updating module screens an optimal scheme according to the interference value and updates a detection path. The system can dynamically optimize the array layout and the detection path, and improves the magnetic flux leakage detection performance.

Description

Technical Field

[0001] This invention relates to the field of ferromagnetic material detection technology, specifically to a magnetic flux leakage detection system with optimized magnetic sensor array layout. Background Technology

[0002] In industrial production and equipment maintenance, ferromagnetic materials are widely used in the manufacture of critical equipment such as pipelines, pressure vessels, and mechanical parts due to their high strength and good magnetic permeability. However, during long-term use, these materials are susceptible to surface and internal defects such as cracks, holes, and rust due to environmental corrosion, loads, and fatigue damage. If these defects are not detected and addressed in a timely manner, they will continue to expand over time.

[0003] To ensure the safe and stable operation of ferromagnetic material equipment, magnetic flux leakage (MF) testing technology has become one of the mainstream testing technologies due to its advantages such as no need for coupling agents, wide detection range, and high sensitivity to surface and near-surface defects. In MF testing systems, the layout design of the magnetic sensor array directly determines the accuracy, completeness, and detection efficiency of magnetic field signal acquisition. Traditional magnetic sensor array layouts often adopt a fixed and uniform distribution. This layout approach often only considers the general requirements under ideal testing environments and does not fully take into account the complex factors in actual testing scenarios.

[0004] In actual testing, the testing environment presents numerous uncertainties, such as differences in the surface flatness of the object being tested, interference from environmental magnetic fields, and deviations in the testing path. When using a fixed-layout magnetic sensor array for testing, if the surface of the testing area is uneven, some sensors may not be able to maintain a reasonable detection distance from the object's surface, leading to distortion of the acquired magnetic field signal and affecting the accuracy of defect identification. If there is external magnetic field interference in the testing environment, the fixed-layout sensor array cannot adjust its acquisition position according to the interference, potentially acquiring a large amount of signals containing interference noise, increasing the difficulty of subsequent signal processing. Furthermore, the fixed detection path cannot be dynamically adjusted based on the real-time acquired defect feature information. When a suspected defect area is detected, it is impossible to optimize the sensor layout and detection path to achieve focused retesting of that area, easily leading to missed defects or incomplete detection.

[0005] Traditional magnetic flux leakage (MFL) detection systems lack the flexibility to adjust their layout when dealing with objects of different sizes and defect types. Redesigning the sensor array layout for different objects not only requires significant time and manpower but also leads to a substantial reduction in detection efficiency, making it difficult to meet the demands of industrial-scale batch testing. As industrial equipment becomes larger and more complex, higher demands are placed on the accuracy and efficiency of MFL detection technology. The limitations of traditional magnetic sensor array layouts are becoming increasingly apparent, necessitating a MFL detection system capable of dynamically optimizing the sensor array layout based on the detection environment and defect characteristics to improve detection accuracy, completeness, and adaptability. Summary of the Invention

[0006] The purpose of this invention is to provide a magnetic flux leakage detection system with optimized layout of magnetic sensor array, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a magnetic flux leakage detection system with optimized magnetic sensor array layout, the system comprising: The magnetic sensor array layout feature extraction module is used to acquire spatial magnetic field distribution data of the ferromagnetic material surface in real time through the magnetic sensor array, and to construct a defect feature information set and detection environment feature information. The regional detection integrity analysis module is used to obtain the regional detection integrity corresponding to the defect feature information set based on the detection environment feature information, combined with the initial layout parameters of the magnetic sensor array, the current detection path, and the current working status information of the magnetic sensor array. A multi-level optimized layout generation module is used to generate a multi-level optimized layout scheme set based on the defect feature information set when the detection integrity of the region is lower than a preset abnormal threshold. The array reconstruction interference analysis module is used to calculate the array reconstruction interference value when the current time magnetic sensor array switches to the initial layout parameters, based on the detection environment feature information. The optimal layout decision update module is used to filter the optimal layout scheme of the magnetic sensor array based on the array reconstruction interference value, and update the current detection path of the magnetic sensor array.

[0008] Preferably, in the magnetic sensing array layout feature extraction module, the defect feature information set includes the defect location, size, and magnetic field strength gradient extracted based on spatial magnetic field distribution data; The environmental characteristics information to be detected include the material thickness and the intensity of environmental magnetic noise in different areas of the ferromagnetic material surface.

[0009] Preferably, the area detection integrity analysis module determines the area detection integrity by combining the minimum coverage distance between the current detection path and the defect location, the change in magnetic field strength gradient of the defect size, and the deviation of the environmental magnetic noise intensity from the preset reference noise.

[0010] Preferably, when the region detection completeness is lower than a preset anomaly threshold, the multi-level optimized layout generation module acquires the boundary topology nodes of each defect in the defect feature information set. Based on the boundary topology nodes, a safe layout area free from magnetic field interference is defined, and a multi-level optimized layout scheme set for the magnetic sensor array is generated through a multi-level optimization algorithm.

[0011] Preferably, the array reconstruction interference analysis module includes: The state switching cost calculation unit is used to obtain the instruction execution sequence for the magnetic sensor array to switch from the current working state to the initial layout parameters; The dynamic interference quantization unit calculates the array reconstruction interference value based on the number of parameter reconstructions and the magnetic field calibration delay time in the instruction execution sequence.

[0012] Preferably, the optimal layout decision update module performs the following operations for each scheme in the multi-level optimized layout scheme set: Based on the array reconstruction interference value and the material thickness in the detection environment characteristic information, the comprehensive interference impact value of each layout scheme is calculated; The optimal layout scheme is selected by using a gradient optimization algorithm to find the layout scheme with the smallest overall interference impact value.

[0013] Preferably, after updating the current detection path of the magnetic sensor array, the optimal layout decision update module generates a defect analysis report containing a defect feature information set, detection environment feature information, and the optimal layout scheme.

[0014] Preferably, the array reconstruction interference analysis module calculates the array reconstruction interference value based on the offset between the magnetic field calibration delay time and the number of parameter reconstructions.

[0015] Preferably, the safe layout region division process of the multi-level optimized layout generation module includes: Generate a magnetic field attenuation safety boundary based on the defect boundary topology nodes; Connect the current position of the magnetic sensor array with each magnetic field attenuation safety boundary to form a continuous region of defect-free magnetic field intersection.

[0016] Preferably, the optimal layout decision update module determines the comprehensive interference impact value by reconstructing the interference value, material thickness, and the ratio of the path length of the current detection path to the optimal layout scheme using an associated array.

[0017] Compared with the prior art, the beneficial effects of the present invention are: Through the coordinated work of various functional modules, the problems of fixed magnetic sensor array layout, poor adaptability, and insufficient detection integrity in traditional magnetic flux leakage detection systems are effectively solved, demonstrating significant advantages in detection accuracy, environmental adaptability, detection efficiency, and system flexibility.

[0018] In terms of data acquisition and defect information gathering, the system incorporates a magnetic sensor array layout feature extraction module. This module can acquire real-time spatial magnetic field distribution data on the surface of ferromagnetic materials using the magnetic sensor array, and construct a defect feature information set and detection environment feature information. Compared to traditional fixed layouts that can only acquire magnetic field signals from a single fixed area, this module can capture dynamic magnetic field changes during the detection process in real time, comprehensively collect defect-related feature parameters, and record key influencing factors in the detection environment. This provides rich and accurate basic data for subsequent detection analysis and layout optimization, avoiding the problem of misjudgment or omission of defects due to incomplete initial data acquisition.

[0019] Regarding ensuring the integrity of the detection area, the area detection integrity analysis module analyzes the area detection integrity corresponding to the defect feature information set based on the detection environment characteristics, combined with the initial layout parameters of the magnetic sensor array, the current detection path, and the working status information. This process breaks away from the traditional detection model that relies solely on completing the detection along a fixed path before evaluating integrity. It can determine in real time during the detection process whether there are detection blind spots or missing information in the current detection area. By promptly identifying problems with insufficient detection integrity, it provides a clear direction for subsequent layout optimization, effectively reducing the risk of missed defects due to incomplete detection, and ensuring that the detection results accurately reflect the defect distribution on the surface of the ferromagnetic material.

[0020] In terms of dynamic layout optimization and adaptability improvement, the multi-level optimized layout generation module generates a set of multi-level optimized layout schemes based on the defect feature information set when the area detection integrity is lower than the preset anomaly threshold. This multi-level optimization approach is not simply adjusting the sensor position, but rather formulating layout schemes with different priorities and adjustment ranges according to the type, size, distribution density of defects, and the specific conditions of the detection environment, making layout optimization more targeted. Compared to the limitations of traditional fixed layouts that cannot adapt to complex environments and diverse defects, this module allows the magnetic sensor array to flexibly adjust its layout according to the actual detection situation. Whether facing an uneven surface or an environment with external magnetic field interference, it can ensure that the sensor is always in the optimal signal acquisition position through optimized layout, improving the accuracy and reliability of magnetic field signal acquisition.

[0021] Regarding array reconstruction stability and detection continuity, the array reconstruction interference analysis module combines detection environment characteristic information to calculate the array reconstruction interference value corresponding to the switch from the current working state to the initial layout parameters. This design fully considers the impact of interference that may occur during layout switching on the stability of the detection system. By calculating the interference value in advance, the influence of interference factors can be avoided or reduced during layout adjustment, preventing interruption or distortion of the detection signal due to layout switching, and ensuring the continuity and stability of the detection process. Traditional detection systems often ignore interference problems during the switching process when adjusting sensor layout, which can easily lead to gaps in detection data. This system effectively solves this problem through this module, ensuring that the detection work can be carried out smoothly and orderly.

[0022] In terms of optimal layout decision-making and detection path optimization, the optimal layout decision-making update module filters the optimal layout scheme based on array reconstruction interference values ​​and updates the current detection path. This module does not simply select a single optimization scheme, but comprehensively considers the layout optimization effect and the degree of reconstruction interference to select the optimal scheme that minimizes the impact on system stability while ensuring detection accuracy. Simultaneously, it dynamically updates the detection path based on the optimal layout scheme, allowing for real-time optimization of the detection path according to defect distribution and layout adjustments. This enables focused re-testing of suspected defect areas and allows for reasonable shortening of the detection path in defect-free areas, improving detection accuracy while effectively reducing unnecessary detection processes and increasing overall detection efficiency. Whether dealing with single-defect or multi-defect distribution objects, this system can achieve efficient and accurate detection through optimal layout and path adjustments, greatly enhancing the system's applicability and practicality in different detection scenarios. Attached Figure Description

[0023] Figure 1 This is a timing diagram of the magnetic flux leakage detection system with optimized magnetic sensor array layout described in this invention. Figure 2 A flowchart for region detection integrity analysis; Figure 3 Flowchart generated for multi-level optimized layout; Figure 4 A flowchart for array reconstruction interference analysis; Figure 5 A flowchart for dividing the layout area for safety. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Please see Figure 1 This invention provides a magnetic flux leakage detection system with optimized magnetic sensor array layout, the system comprising: This system acquires real-time spatial magnetic field distribution data on the surface of ferromagnetic materials to construct a defect feature information set and detection environment feature information. Based on the detection environment feature information, combined with the initial layout parameters of the magnetic sensor array, the current detection path, and the current operating status information of the magnetic sensor array, the system obtains the region detection completeness corresponding to the defect feature information set. When the region detection completeness is lower than a preset anomaly threshold, a multi-level optimized layout scheme set is generated based on the defect feature information set. Combining the detection environment feature information, the system calculates the array reconstruction interference value corresponding to the current time when the operating status of the magnetic sensor array switches to the initial layout parameters. Based on the array reconstruction interference value, the optimal layout scheme of the magnetic sensor array is selected, and the current detection path of the magnetic sensor array is updated. This system enables dynamic adjustment of sensor layout in complex detection environments, improving the accuracy and efficiency of defect detection.

[0026] Example 1: See Figure 2 The implementation of a magnetic flux leakage detection system with optimized magnetic sensor array layout involves the specific operational procedures of a magnetic sensor array layout feature extraction module and a region detection integrity analysis module. The system initiates the detection process by real-time acquisition of spatial magnetic field distribution data on the surface of ferromagnetic materials. The magnetic sensor array moves along a preset scanning path on the material surface, with each sensor node capturing magnetic field strength data at a specific sampling frequency. After high-frequency noise is removed by a signal conditioning circuit, this raw data is sent to a data processing unit for spatial interpolation calculations to generate a complete two-dimensional or three-dimensional magnetic field distribution map. During data processing, the system uses a sliding window algorithm to identify areas of magnetic field anomalies, and preliminarily determines the location of potential defects by comparing the differences in magnetic field strength between adjacent areas.

[0027] The defect feature information set is constructed based on the analyzed spatial magnetic field distribution data. The system identifies extreme points in the magnetic field distribution, which typically correspond to the central region of the defect. By calculating the rate of change of magnetic field strength in space, the system obtains the gradient distribution characteristics of each suspected defect region. Defect size is estimated using contour line analysis, with a specific magnetic field strength value as a threshold to delineate the defect boundary. Simultaneously, the system records the peak magnetic field strength and maximum gradient value for each defect region; these data collectively constitute the magnetic field feature fingerprint of the defect. The acquisition of environmental feature information is carried out concurrently. Material thickness data is acquired through an ultrasonic thickness measurement module integrated into a sensor array, which measures the thickness variation of different regions of the material in a non-contact manner. Environmental magnetic noise intensity is monitored by dedicated reference sensors placed around the detection area to record fluctuations in the background magnetic field.

[0028] The regional detection integrity analysis module integrates multi-source information for evaluation and calculation. This module receives defect feature information and detection environment feature information from the feature extraction module, and simultaneously acquires the initial layout parameters of the magnetic sensor array, including the initial position coordinates, spacing configuration, and orientation angle of the sensors. Current detection path information is acquired in real-time from the motion control module, including the motion trajectory coordinate sequence of the sensor array and timestamp data of the detection points. The operating status information of the magnetic sensor array is provided by the status monitoring unit, including the calibration status, sensitivity settings, and current sampling frequency of each sensor. This data is then fed into the integrity assessment algorithm for processing.

[0029] The integrity assessment algorithm calculates the region detection integrity through multi-parameter fusion. The algorithm calculates the minimum coverage distance between the current detection path and the defect location, using a spatial nearest neighbor search algorithm to traverse all points on the path and calculate the geometric distance to each defect location, taking the minimum value as the coverage distance index. This index reflects the degree of coverage of the defect area by the detection path; a larger minimum coverage distance indicates the potential existence of detection blind spots. The algorithm analyzes the change in magnetic field strength gradient of the defect size, evaluating the stability of defect features by comparing the gradient characteristic change amplitude of the same defect in the current detection cycle with that of historical cycles. The deviation of the environmental magnetic noise intensity from the preset reference noise is calculated using standard deviation analysis, comparing the real-time monitored noise data with the reference noise spectrum stored in the database to calculate its statistical deviation value.

[0030] The calculated parameters are used to generate a region detection integrity index through a weighted fusion algorithm. The weighting coefficients are dynamically adjusted according to the characteristics of the detection environment; for example, in a high-noise environment, noise deviation is given a higher weight. The final region detection integrity is a dimensionless numerical index; the lower the value, the worse the detection integrity. The system's preset anomaly threshold is determined based on statistical analysis of a large amount of historical data. When the calculated integrity value is lower than this threshold, the system will trigger an optimization layout process. The entire evaluation process adopts a real-time pipeline architecture to ensure continuous and dynamic evaluation of detection integrity during the detection process. In specific implementation, the magnetic sensor array adopts a matrix arrangement structure, with each sensor node having independent signal acquisition and processing capabilities. The array is connected to the main control unit via a communication bus, uploading the acquired magnetic field data in real time. The data preprocessing stage includes signal filtering, baseline correction, and temperature compensation to eliminate the influence of environmental factors on the measurement results. The defect feature extraction algorithm uses a multi-scale analysis method to identify defect features of different sizes. The detection environment monitoring module uses multi-sensor fusion technology to integrate ultrasonic thickness measurement data and magnetic noise monitoring data to construct a complete environmental feature model.

[0031] The regional detection integrity analysis module adopts a distributed computing architecture, allocating computational tasks to multiple processing units for parallel execution. The minimum coverage distance calculation uses a spatial indexing acceleration algorithm to improve search efficiency. The magnetic field gradient change analysis employs a time-series comparison method to establish an evolutionary model of defect features. Noise deviation calculation uses a frequency domain analysis method, converting noise signals to the frequency domain for feature extraction and comparison. The weighted fusion algorithm employs an adaptive weight adjustment mechanism, dynamically optimizing weight allocation based on changes in environmental conditions.

[0032] The system implementation emphasizes real-time data quality and accuracy. Sensor data acquisition employs synchronous sampling technology to ensure temporal consistency of spatial magnetic field data. The defect feature extraction algorithm uses an iterative optimization approach to progressively refine feature parameters. Environmental feature monitoring utilizes continuous sampling to capture instantaneous changes in environmental conditions. The integrity assessment algorithm employs a sliding window update mechanism to continuously update assessment results to reflect the latest detection status. The entire implementation process emphasizes collaborative work between modules, ensuring smooth information transmission through a unified data interface and communication protocol.

[0033] Example 2: See Figure 3 and Figure 4The implementation of the magnetic flux leakage detection system with optimized magnetic sensor array layout involves the specific operation procedures of a multi-level optimized layout generation module and an array reconstruction interference analysis module. When the evaluation value output by the region detection integrity analysis module is lower than a preset anomaly threshold, the multi-level optimized layout generation module begins to execute its function. This module extracts the boundary topological nodes of each defect from the defect feature information set. These nodes are feature point sets obtained by morphological processing of spatial magnetic field distribution data, reflecting the key geometric features of the defect contour. The generation of boundary topological nodes uses the Delaunay triangulation algorithm, discretizing the continuous curve of the defect edge into a series of feature points with topological relationships. These points not only contain spatial coordinate information but also carry the magnetic field gradient vector data at that location.

[0034] Based on these boundary topology nodes, the system begins to delineate safe layout regions free from magnetic field interference. This process employs a region delineation method based on potential field theory, treating each defect as a potential source of magnetic field interference. The safe region is determined by calculating the composite potential field value at each point in space. The system generates magnetic field attenuation safety boundaries, which are composed of points satisfying specific potential field threshold conditions. Their shape and extent are determined by the magnetic field intensity distribution characteristics of the defects. The current position of the magnetic sensor array is spatially connected to each magnetic field attenuation safety boundary. A visibility map algorithm is used to generate a continuous region of intersection of defect-free magnetic fields. This region ensures that the magnetic field interference experienced by the sensor is minimized when operating within this range.

[0035] The multi-level optimization algorithm then generates a set of layout schemes within this safe area. The first level of the algorithm employs a global exploration strategy, using an improved genetic algorithm to perform an extensive search in the solution space. Chromosome encoding uses real numbers, with each gene representing the position coordinates and orientation angle of a sensor. The fitness function comprehensively considers multiple factors such as coverage, signal strength, and interference level. The second level of optimization employs a local fine-tuning search strategy, using a sequential quadratic programming algorithm to refine the initial schemes, adjusting the micro-positions and orientations of the sensor nodes to further improve detection performance. The final output set of multi-level optimized layout schemes contains several Pareto optimal solutions, each corresponding to a complete set of sensor array configuration parameters.

[0036] The array reconstruction interference analysis module begins by evaluating the cost required to switch from the current state to each candidate scheme. The state switching cost calculation unit analyzes the difference between the current operating state and the target state, generating a detailed instruction execution sequence. This sequence includes instructions for moving the sensor's physical location, resetting operating parameters, and initiating the calibration process. Each instruction carries a timestamp and execution priority information, forming an execution plan of a directed acyclic graph. The dynamic interference quantification unit then calculates the degree of interference during the reconstruction process based on this execution plan, focusing primarily on two core indicators: the number of parameter reconstructions and the magnetic field calibration delay time.

[0037] The number of parameter reconstruction attempts reflects the number of sensor parameters that need to be adjusted, including the number of times configurable parameters such as position coordinates, sampling frequency, and sensitivity are changed. The magnetic field calibration delay includes hardware response delay and software processing delay; the hardware delay depends on the mechanical motion characteristics of the sensor, while the software delay is related to the complexity of the parameter configuration algorithm. These metrics are transformed into an array reconstruction interference value through a comprehensive mathematical model, which quantifies the impact on the system's detection performance during state switching. The array reconstruction interference value is calculated using the following formula:

[0038] in: Indicates the array reconstruction interference value. This represents the total number of parameters that need to be refactored. It is the first The reconstruction weight coefficients of each parameter, Indicates the first The difficulty factor of reconstructing each parameter. It is the calibration time weighting coefficient. Indicates the first Calibration delay time for each sensor. Parameter reconstruction weighting coefficients. The weight of location parameters is dynamically adjusted based on parameter type and importance, with location parameters typically having a higher weight than other configuration parameters. (Reconstruction difficulty factor) The difficulty of reconstructing position parameters depends on the physical complexity of parameter adjustment; for example, the distance the sensor moves affects the difficulty. Calibration time weighting coefficient. Configuration is based on system real-time requirements, with higher requirements resulting in greater weight. In implementation, the multi-level optimization layout generation module employs a distributed computing architecture, decomposing optimization tasks into multiple computing nodes for parallel processing. Each computing node is responsible for generating a scheme for a sub-region, with scheme fusion achieved through coordination nodes. Boundary topology node processing utilizes point cloud data processing technology, employing a KD-tree data structure for fast spatial retrieval and nearest neighbor lookup. The division of safe zones utilizes computational geometry methods, determining the optimal sensor placement area by constructing convex hulls and Voronoi diagrams.

[0039] The array reconfiguration interference analysis module employs an event-driven architecture to perform evaluation tasks. When a new candidate layout scheme is generated, an evaluation event is triggered, initiating the cost calculation process. The generation of the instruction execution sequence utilizes a rule-based expert system, automatically generating the optimal execution steps based on factors such as sensor type, movement distance, and parameter adjustment magnitude. The dynamic interference quantification unit uses time series analysis methods to predict the time consumption and resource usage of each execution step, thereby accurately estimating the overall interference level.

[0040] This implementation emphasizes the real-time performance and accuracy of each processing step. The multi-level optimization algorithm employs an adaptive convergence strategy, dynamically adjusting the number of iterations and the search range based on the complexity of the detection environment. Boundary topology processing introduces multi-resolution analysis technology to extract defect features at different scales, ensuring that the generated safe area avoids interference without sacrificing detection coverage. The interference analysis module uses an incremental update mechanism, enabling rapid reassessment of interference values ​​when environmental conditions change, without requiring complete recalculation. The entire implementation emphasizes collaborative work between subsystems. The multi-level optimization layout generation module and the feature extraction module maintain data synchronization, ensuring the use of up-to-date defect information. The array reconstruction interference analysis module interacts closely with the system status monitoring module, acquiring the sensor's current operating status and performance parameters in real time. All calculations employ algorithms with good numerical stability, avoiding rounding errors and numerical overflow, ensuring the reliability of the calculation results. Data storage utilizes time-series database technology, comprehensively recording all intermediate results and final solutions during the optimization process, providing a data foundation for subsequent analysis.

[0041] Example 3: Implementation of a magnetic flux leakage detection system with optimized magnetic sensor array layout involves the complete operation flow of the optimal layout decision update module. This module receives a scheme set from the multi-level optimized layout generation module and evaluation data from the array reconstruction interference analysis module, and begins to perform comprehensive decision analysis. For each candidate scheme in the multi-level optimized layout scheme set, the system calculates its comprehensive interference impact value, which reflects the overall interference level experienced by the system when adopting that layout scheme. The calculation process comprehensively considers the array reconstruction interference value, the material thickness parameter in the detection environment feature information, and the path matching degree between the current detection path and the optimal layout scheme. Material thickness data affects magnetic field propagation characteristics; thicker materials require higher sensor sensitivity, which may amplify the interference effect generated during layout switching. The calculation of the comprehensive interference impact value adopts a multi-factor weighted model and is quantified by the following formula:

[0042] in: This represents the overall interference impact value. It is the array reconstruction interference value. For material thickness, It is a thickness reference constant. This represents the time cost of path adjustment. and These are the weighting coefficients for the interference term and the path term, respectively. Thickness reference constant. Based on material type and testing standards, the time cost of path adjustment is... The weighting is calculated by subtracting the execution time of the new path from the execution time of the current path. The weighting coefficients are dynamically adjusted based on the priority of the detection tasks; tasks with high real-time requirements will be assigned higher weights to path items.

[0043] The gradient optimization algorithm seeks the optimal solution that minimizes the overall disturbance impact among all candidate solutions. The algorithm employs the conjugate gradient method for iterative optimization, with the initial point set to the parameter values ​​corresponding to the current layout scheme. In each iteration, the algorithm calculates the gradient direction of the objective function at the current point and performs a one-dimensional search along this direction to find the optimal solution for the step size. To avoid getting trapped in local optima, the algorithm introduces a momentum term and an adaptive learning rate mechanism. When the change in the objective function value after multiple consecutive iterations is less than a set threshold, the algorithm terminates and outputs the current optimal solution. The final selected layout scheme must not only minimize the overall disturbance impact but also pass feasibility verification, including sensor movement range limitations, energy consumption constraints, and real-time requirements.

[0044] After selecting the optimal layout, the system updates the current detection path of the magnetic sensor array. The path update process employs a smooth transition strategy, calculating the difference between the current path and the new path to generate a progressive path adjustment sequence. The new detection path needs to avoid known defect areas while ensuring complete coverage of undetected areas. The path planning algorithm uses an improved A* algorithm, incorporating magnetic field strength distribution as part of a heuristic function, prioritizing paths with less magnetic field interference. The updated path information is sent to the actuator via the motion control interface, simultaneously updating the system's internal path database.

[0045] The generation of the defect analysis report is a crucial step in the implementation process. The report integrates all relevant data from the start of detection to the current moment, including the original defect feature information set, environmental monitoring data, layout optimization process records, and the final optimal layout scheme. The defect feature information set contains detailed parameters for each defect, such as spatial coordinates, dimensional measurements, extreme magnetic field strength values, and gradient distribution characteristics. The environmental feature information record includes material thickness distribution maps, environmental magnetic noise spectrum characteristics, and auxiliary parameters such as temperature and humidity. The optimal layout scheme section details the final configuration of the sensor array, including the position coordinates, orientation angle, operating parameter settings, and expected detection performance indicators for each sensor.

[0046] The report generation module employs a structured data storage format, with all data organized chronologically. Report output supports multiple formats, including readable text summaries, detailed data tables, and intuitive graphical displays. The graphical displays include magnetic field distribution cloud maps, defect location marker maps, sensor layout diagrams, and detection path trajectory maps. These visualizations help users intuitively understand the detection results and system optimization process. The report also includes data quality assessment metrics, illustrating the reliability of each data point and the potential error range. In practical implementation, the calculation of the comprehensive interference impact value utilizes a distributed parallel computing architecture. The computational tasks for each candidate solution are assigned to different processing units, achieving efficient large-scale data processing through the MapReduce framework. The gradient optimization algorithm employs an incremental computation mode, utilizing historical optimization data to accelerate the convergence process. The path update algorithm introduces a real-time obstacle avoidance mechanism, enabling dynamic path adjustments to cope with sudden environmental changes.

[0047] The defect analysis report generation adopts a template-based design to ensure the standardization and consistency of the output format. The data verification process ensures that all data included in the report undergoes integrity checks and logical consistency verification. Report storage employs a version control mechanism, retaining historical versions for traceability and comparison. The entire implementation process emphasizes data traceability; each calculation step and decision result is timestamped and includes execution context information, forming a complete audit trail.

[0048] During the implementation of this system, special attention was paid to the real-time performance of the algorithms. The gradient optimization algorithm employs sparse matrix optimization and parallel computing techniques to reduce computation time. The path planning algorithm introduces preprocessing and caching mechanisms to improve path update efficiency. The report generation module adopts a streaming processing architecture to achieve real-time data collection and processing. All modules are fault-tolerant, maintaining normal operation of basic functions even in the event of partial data loss or anomalies. Data exchange and communication use a unified protocol standard to ensure seamless collaboration between modules. The comprehensive interference impact value calculation module maintains real-time synchronization with the optimization scheme database to ensure that the latest scheme data is used for calculation. The gradient optimization algorithm interacts with the parameter constraint management system to ensure that the generated solution meets all physical constraints. The path update module is tightly integrated with the real-time positioning system to ensure the accuracy and executability of path planning. The report generation module extracts the necessary information from the data warehouse and generates a comprehensive analysis report through data fusion technology.

[0049] Example 4: See Figure 5The implementation method of the magnetic flux leakage detection system with optimized magnetic sensor array layout involves a detailed operational process of the offset calculation method in the array reconstruction interference analysis module and the safe area division process of the multi-level optimized layout generation module. When performing interference assessment, the array reconstruction interference analysis module calculates the final array reconstruction interference value based on the offset between the magnetic field calibration delay time and the number of parameter reconstructions. The offset calculation requires establishing historical benchmark data. The system maintains a database containing historical calibration delay times and the number of parameter reconstructions, recording typical values ​​under different detection environments and operating conditions. This historical data is processed using time series analysis to extract representative benchmark values. For example, a moving average method is used to calculate the benchmark level of the delay time, and statistical distribution analysis is used to determine the normal range of the number of parameter reconstructions.

[0050] The magnetic field calibration delay time in the current operating state is obtained through real-time monitoring. The system records a timestamp at the start of each calibration operation and again after calibration is completed, calculating the difference between the two timestamps to obtain the actual delay time. The number of parameter reconstructions is statistically analyzed through instruction execution sequence analysis. The system parses all parameter adjustment instructions required to switch from the current state to the target state and calculates the number of parameters that need to be modified. After obtaining the current values, the system calculates their relative offsets from the corresponding reference values. The delay time offset is expressed as a percentage change rate, while the parameter reconstruction number offset is standardized using absolute difference. These two offsets are combined in a weighted manner to form the final array reconstruction interference value. The weighting coefficient is dynamically adjusted according to the stability of the detection environment, with higher weight given to the delay time offset in environments with large fluctuations.

[0051] The process of dividing the safe layout area begins with the processing of the topological nodes of the defect boundaries. The system performs spatial clustering analysis on the boundary node set of each defect to identify key nodes representing the geometric characteristics of the defect. These key nodes are used to generate magnetic field attenuation safety boundaries, forming closed boundaries by connecting nodes with similar magnetic field attenuation characteristics. The boundary generation algorithm considers the spatial distribution of magnetic field strength and uses contour tracing technology to determine the boundary lines of the safe area. The connection operation between the current position of the magnetic sensor array and each magnetic field attenuation safety boundary uses a visibility graph algorithm to calculate whether the direct lines from the current position to each safety boundary point intersect with other defect areas, and filters out the set of collision-free connections.

[0052] Based on these collision-free connections, the system constructs a continuous region of defect-free magnetic field intersections. This region is calculated using polygon Boolean operations, performing a union operation on the safe regions corresponding to each safe boundary and subtracting the influence range of all defective regions. The resulting continuous region is one or more polygonal areas, representing the spatial range within which the sensor can be safely deployed. Within this region, the magnetic field interference level is below a preset threshold, ensuring the reliability of the detection data. The system also performs a quality assessment on this continuous region, calculating its area, connectivity, and boundary smoothness to ensure the generated region is suitable for sensor placement. In the specific implementation, offset calculation uses a sliding window mechanism to update historical baseline data. The system maintains a fixed-size historical data window; after each new detection task is completed, old data is removed, new data is added to the window, and the baseline value is updated accordingly. This mechanism ensures that the baseline data reflects the most recent operating state and adapts to slow changes in system performance. The safe region partitioning algorithm employs a multi-resolution processing strategy, quickly determining the approximate safe range at a coarse-grained level and performing fine-grained, precise calculations in the region of interest.

[0053] The weight adjustment during offset calculation employs a fuzzy logic control method. The system dynamically determines weight coefficients based on multiple factors, including environmental noise levels, material uniformity, and detection speed requirements, using fuzzy inference rules. This method's advantage lies in its ability to handle uncertainty and nonlinear relationships, resulting in a more rational weight allocation. Machine learning techniques are introduced into the safe zone division process. Historical successful deployment data is used to train a classification model, assisting in determining which areas are more suitable for sensor placement and improving the practicality of the division results.

[0054] The system implementation prioritizes a balance between computational efficiency and data accuracy. Offset calculations employ an incremental update algorithm to avoid recalculating all historical data each time. Safe zone partitioning utilizes a spatial index structure to accelerate geometric calculations, and R-trees are used to organize defect regions and boundary nodes, improving spatial query efficiency. All calculation processes employ error control mechanisms to ensure the numerical stability of the final results. Data recording and tracking are crucial components of the implementation process. The system meticulously records the basic data, intermediate results, and final values ​​for each offset calculation, forming a complete data chain. Process data is also retained for each step of safe zone partitioning, including initial boundary nodes, intermediately generated boundary lines, and Boolean operation results. This data is used not only for real-time decision-making but also to support subsequent analysis and algorithm improvement.

[0055] Table 1: Calculation Table of Magnetic Field Calibration Delay Time and Parameter Reconstruction Number Offset

[0056] Table 1 illustrates offset calculation examples for five detection tasks, including delay time offset, parameter reconstruction count offset, and the final calculated comprehensive interference value. The historical average delay column displays the baseline delay time calculated from historical data, while the current delay column records the actual measured value. The delay offset is expressed as a percentage relative to the historical baseline. The baseline reconstruction count column displays the baseline level of parameter reconstruction, while the current reconstruction count column records the actual number of parameters requiring reconstruction. The count offset uses standardized values, and the comprehensive interference value is the final result obtained by weighted combination of the two offsets. These data demonstrate the variation in interference levels across different detection tasks, providing a quantitative basis for layout decisions. During implementation, the system employs a real-time monitoring mechanism to track offset trends. When the offsets of multiple consecutive detection tasks exceed the normal range, the system triggers an early warning mechanism, indicating potential equipment performance degradation or environmental anomalies. The safe zone division results are visualized to help operators intuitively understand the spatial distribution characteristics of deployable areas. All calculation processes are repeatable; the same input data always produces the same output results, ensuring the consistency and predictability of system behavior.

[0057] Example 5: Implementation of a Magnetic Leakage Detection System with Optimized Layout of Magnetic Sensor Array. This example focuses on the method for determining the comprehensive interference impact value in the optimal layout decision update module. This module performs decision analysis by associating multiple key parameters, including array reconstruction interference value, material thickness characteristics, and the path length ratio between the current detection path and the optimal layout scheme. The array reconstruction interference value comes from previous evaluation processes and reflects the degree of system interference caused by state switching. Material thickness data is extracted from environmental characteristic information; different thickness regions have different effects on magnetic field propagation, requiring differentiated evaluation standards. The path length ratio is obtained by calculating the ratio of the path length required for the new scheme to the current path length; this parameter directly affects the execution efficiency and time cost of the detection task.

[0058] The determination of the comprehensive interference impact value employs a multi-parameter fusion algorithm, integrating the three core parameters mentioned above into a comprehensive index through weighted combination. The algorithm first standardizes each parameter to eliminate the influence of dimensional differences. The array reconstruction interference value uses a maximum-minimum normalization method, mapping it to a numerical range between zero and one. The material thickness parameter is segmented according to material type and detection standards, with different normalization coefficients assigned to different thickness segments. The path length ratio is directly used as a scaling factor in the calculation, reflecting the relative impact of path changes. These processed parameters are fused using a linear weighted model, with the weight coefficients dynamically configured according to the specific requirements of the detection task.

[0059] The determination of weighting coefficients considers multiple factors, including detection accuracy requirements, task urgency, and equipment status. For detection tasks requiring high accuracy, the weight of array reconstruction interference values ​​is appropriately increased because system stability has a significant impact on detection quality. In time-sensitive tasks, the weight of the path length ratio is increased accordingly, prioritizing detection efficiency. Equipment status monitoring data is used to fine-tune the weighting allocation; when sensor performance fluctuates, the system reduces the weighting requirement for path efficiency and increases its focus on system stability. This dynamic weighting mechanism allows the system to adapt to different working scenarios and changing requirements.

[0060] The calculation of the comprehensive interference impact value employs an iterative optimization approach. The system generates multiple weight combination schemes, calculates the corresponding comprehensive values ​​for each, and determines the optimal value through comparative analysis. Each weight combination scheme is generated based on decision preferences, covering evaluation perspectives with different emphases. Constraints are introduced during the calculation process to ensure that the comprehensive value accurately reflects the overall interference situation of the system. The final determined comprehensive interference impact value serves as a dimensionless index for scheme comparison and selection.

[0061] During the scheme comparison phase, the system sorts the comprehensive interference impact values ​​of each candidate layout scheme and selects the scheme with the smallest value as the optimal solution. The comparison process considers not only the numerical magnitude but also the contribution of each parameter, ensuring that the selected scheme achieves a good balance in all aspects. For schemes with similar values, the system performs sensitivity analysis to assess the impact of changes in each parameter on the comprehensive value, selecting the scheme with better stability. The final selected optimal scheme must satisfy all physical constraints, including sensor movement range limitations, energy consumption constraints, and real-time requirements. The implementation process after the scheme is determined adopts a gradual adjustment strategy to avoid excessive interference caused by sudden state transitions. The system generates a transition plan, adjusting the sensor array layout parameters in stages. After each stage, a performance evaluation is performed to ensure a smooth system transition. During the transition, changes in the comprehensive interference impact value are continuously monitored. If abnormal fluctuations are detected, adjustments are paused and the scheme is re-evaluated. This cautious implementation method minimizes the impact of the layout optimization process on the detection task.

[0062] Data recording and analysis are integrated throughout the entire implementation process. The system meticulously records the comprehensive interference impact value and the values ​​of its constituent parameters for each candidate solution. This data is used for subsequent analysis and improvement, employing machine learning methods to uncover correlations between parameters and optimize weight allocation strategies. The accumulation of historical data helps the system build a more accurate evaluation model, improving the accuracy and reliability of decision-making. A complete audit trail is maintained for all decision-making processes, including parameter values, calculation procedures, and final results, ensuring traceability and transparency.

[0063] During system implementation, real-time performance was given special attention. The calculation of the comprehensive interference impact value employs a highly efficient algorithm, capable of completing within milliseconds. Parameter standardization uses pre-calculated conversion coefficients to avoid delays caused by real-time calculations. The weight allocation strategy uses a lookup table method, pre-configuring weight combinations based on common scenarios and selecting the most suitable configuration in real-time based on task characteristics. These optimization measures ensure the system can make rapid layout decisions during real-time detection. Collaboration with other modules is also a crucial aspect of the implementation. The comprehensive interference impact value calculation module maintains real-time data synchronization with the status monitoring module, using the latest equipment status information. Interaction with the path planning module ensures accurate calculation of path length ratios, using actual executable path data rather than theoretical values. Cooperation with the environmental perception module ensures real-time updates of material thickness data, reflecting environmental changes during detection. This cross-module collaborative mechanism ensures decisions are based on comprehensive and accurate information. Error handling and anomaly management mechanisms guarantee the reliability of the implementation. When input data is abnormal or errors occur during calculation, the system can automatically switch to a backup decision mode, using conservative weight configurations for calculation. All anomalies are recorded and analyzed to improve the system's robustness. The implementation process also includes a self-checking function to periodically verify the rationality and consistency of the calculation results, ensuring the long-term stable operation of the system.

[0064] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0065] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A magnetic flux leakage detection system with optimized magnetic sensor array layout, characterized in that, The system includes: The magnetic sensor array layout feature extraction module is used to acquire spatial magnetic field distribution data of the ferromagnetic material surface in real time through the magnetic sensor array, and to construct a defect feature information set and detection environment feature information. The regional detection integrity analysis module is used to obtain the regional detection integrity corresponding to the defect feature information set based on the detection environment feature information, combined with the initial layout parameters of the magnetic sensor array, the current detection path, and the current working status information of the magnetic sensor array. A multi-level optimized layout generation module is used to generate a multi-level optimized layout scheme set based on the defect feature information set when the detection integrity of the region is lower than a preset abnormal threshold. The array reconstruction interference analysis module is used to calculate the array reconstruction interference value when the current time magnetic sensor array switches to the initial layout parameters, based on the detection environment feature information. The optimal layout decision update module is used to filter the optimal layout scheme of the magnetic sensor array based on the array reconstruction interference value, and update the current detection path of the magnetic sensor array.

2. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 1, characterized in that, In the magnetic sensing array layout feature extraction module, the defect feature information set includes the defect location, size, and magnetic field strength gradient extracted based on spatial magnetic field distribution data. The environmental characteristics information to be detected include the material thickness and the intensity of environmental magnetic noise in different areas of the ferromagnetic material surface.

3. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 2, characterized in that, The region detection integrity analysis module determines the region detection integrity by taking into account the minimum coverage distance between the current detection path and the defect location, the change in magnetic field strength gradient of the defect size, and the deviation of the environmental magnetic noise intensity from the preset reference noise.

4. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 3, characterized in that, When the region detection integrity is lower than a preset anomaly threshold, the multi-level optimized layout generation module obtains the boundary topology nodes of each defect in the defect feature information set. Based on the boundary topology nodes, a safe layout area free from magnetic field interference is defined, and a multi-level optimized layout scheme set for the magnetic sensor array is generated through a multi-level optimization algorithm.

5. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 4, characterized in that, The array reconstruction interference analysis module includes: The state switching cost calculation unit is used to obtain the instruction execution sequence for the magnetic sensor array to switch from the current working state to the initial layout parameters; The dynamic interference quantization unit calculates the array reconstruction interference value based on the number of parameter reconstructions and the magnetic field calibration delay time in the instruction execution sequence.

6. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 5, characterized in that, The optimal layout decision update module performs the following operations for each scheme in the multi-level optimized layout scheme set: Based on the array reconstruction interference value and the material thickness in the detection environment characteristic information, the comprehensive interference impact value of each layout scheme is calculated; The optimal layout scheme is selected by using a gradient optimization algorithm to find the layout scheme with the smallest overall interference impact value.

7. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 6, characterized in that, After updating the current detection path of the magnetic sensor array, the optimal layout decision update module generates a defect analysis report containing a set of defect feature information, detection environment feature information, and the optimal layout scheme.

8. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 7, characterized in that, The array reconstruction interference analysis module calculates the array reconstruction interference value based on the offset between the magnetic field calibration delay time and the number of parameter reconstructions.

9. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 8, characterized in that, The safe layout region division process of the multi-level optimized layout generation module includes: Generate a magnetic field attenuation safety boundary based on the defect boundary topology nodes; Connect the current position of the magnetic sensor array with each magnetic field attenuation safety boundary to form a continuous region of defect-free magnetic field intersection.

10. The magnetic flux leakage detection system with optimized magnetic sensor array layout according to claim 9, characterized in that, The optimal layout decision update module determines the comprehensive interference impact value by reconstructing the interference value, material thickness, and the ratio of the path length of the current detection path to the optimal layout scheme using the associated array.

Images