Petrochemical storage tank inner wall climbing robot communication relay deployment system and method based on permanent magnetic adsorption

Abstract

The application relates to the field of wall-climbing robot communication relay and discloses a petrochemical storage tank wall-climbing robot communication relay deployment system and method based on permanent magnetic adsorption, which comprises an anchoring calculation module, a magnetic distance calculation module, a distance conversion module, a difference comparison module, an equal magnetic guiding module and a reinforcement deployment module. Based on signal fading amount and multi-dimensional quality indexes, the anchoring fitting calculation of the relay node is carried out to obtain anchoring success signaling. The magnetic force line distortion vector is separated by mutual interference to obtain the natural magnetic distance, and then the scale conversion is carried out to obtain the legal execution distance. The adsorption force and the prepared force are compared by difference to obtain the adsorption force difference. The equal magnetic force line is guided to the path to obtain the relay deployment position and anchor the relay node. The signal-to-noise ratio of the relay node is detected. If the signal-to-noise ratio is low, the control is carried out. If the signal-to-noise ratio does not recover, the robot is put into the reinforcement relay. The application can be used for the automatic deployment of the wall-climbing robot communication relay and the maintenance of the signal quality.

Description

Technical Field

[0001] This invention relates to the field of communication relay for wall-climbing robots, and more particularly to a communication relay deployment system and method for wall-climbing robots in petrochemical storage tanks based on permanent magnet adsorption. Background Technology

[0002] In existing technologies, the deployment of communication relays for wall-climbing robots mainly uses laser ranging, ultrasonic sensors, or visual positioning to measure the distance between nodes. The signal-to-noise ratio of the communication link is used to trigger relay deployment, and the anchoring status is confirmed through the communication link. However, in the oil medium environment inside petrochemical storage tanks, the ranging accuracy of laser and ultrasonic signals is greatly reduced due to medium attenuation and scattering. Visual positioning is difficult to extract effective features due to insufficient light and oil mist obstruction. The communication signal is subject to delays and uncertainties in link quality detection due to fluctuations in the dielectric loss of oil. As a result, relay distance measurement, anchoring status confirmation, and deployment location determination are difficult to perform reliably under the harsh working conditions inside the storage tank. This leads to low relay deployment success rate and long deployment cycle, making it difficult to meet the needs of autonomous deployment of communication relays inside storage tanks.

[0003] In existing technologies, the navigation and positioning of wall-climbing robots rely on GPS, visual SLAM, or inertial navigation systems. GPS signals cannot be received inside storage tanks, visual SLAM faces difficulties in feature matching in low light and oil mist environments, and inertial navigation suffers from cumulative drift. When the communication quality of relay nodes deteriorates, existing solutions employ a reinforcement deployment strategy based on coordinate backtracking, relying on the accuracy of coordinate recording and the continuous effective operation of the positioning system. This approach is unreliable in storage tank environments where positioning signals are missing or drifting. Furthermore, existing permanent magnet adsorption force detection is only used to monitor the robot's adsorption safety and is not associated with relay deployment decisions. The permanent magnet is merely used as an adsorption and fixation element, and its magnetic field response characteristics are not used for sensing or positioning purposes. This results in a disconnect between the robot's perception information and relay deployment requirements, making it difficult to maintain the signal quality of the communication link in environments where positioning signals are missing. Therefore, there is an urgent need to develop a communication relay deployment system and method for wall-climbing robots in petrochemical storage tanks based on permanent magnet adsorption, for automated deployment and signal quality maintenance of communication relays for wall-climbing robots. Summary of the Invention

[0004] This invention provides a communication relay deployment system and method for a robot climbing inside a petrochemical storage tank based on permanent magnet adsorption, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a communication relay deployment system for a climbing robot inside a petrochemical storage tank based on permanent magnet adsorption. The system comprises an anchoring calculation module, a magnetic distance calculation module, a spacing conversion module, a difference comparison module, an isomagnetic guidance module, and a reinforcement deployment module, wherein: The anchoring calculation module is used to perform anchoring and bonding calculation on each relay node based on the fading amount and multi-dimensional quality indicators of the back-transmitted signals of each relay node during the robot's movement process, and to obtain the anchoring success signal of each relay node. The magnetic distance calculation module is used to perform mutual interference separation on the magnetic field line distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic distance of the relay node. The spacing conversion module is used to use the signal fading amount of the successfully anchored relay node measured by the robot as a reference to perform scale conversion on the natural magnetic spacing to obtain the legal execution spacing between adjacent relay nodes. The difference comparison module is used to compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position. The isomagnetic guidance module is used to guide the robot path with isomagnetic lines based on the adsorption force difference and the statutory execution interval, so as to obtain the relay deployment position of the robot and anchor the relay node. The reinforcement deployment module is used to detect the signal-to-noise ratio of the successfully anchored relay node in real time. If the signal-to-noise ratio is too low, the magnetic field strength of the permanent magnet is adjusted until it recovers. If the adjustment does not recover after a preset time, the deployment robot returns to the relay node to deploy the reinforcement relay.

[0006] To address the aforementioned problems, this invention also provides a method for deploying a communication relay for a wall-climbing robot inside a petrochemical storage tank based on permanent magnet adsorption, the method comprising: S1. Based on the signal fading and multi-dimensional quality indicators of each relay node during the robot's movement process, the anchoring and bonding calculation is performed on each relay node to obtain the anchoring success signal of each relay node. S2. Perform mutual interference separation on the magnetic field distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic spacing of the relay node. S3. Using the signal fading amount of the successfully anchored relay node measured by the robot as a reference, the natural magnetic spacing is scaled to obtain the legal execution spacing between adjacent relay nodes. S4. Compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position. S5. Based on the adsorption force difference and the statutory execution interval, perform isomagnetic line guidance on the robot path to obtain the relay deployment position of the robot and anchor the relay node; S6. Detect the signal-to-noise ratio of the successfully anchored relay node in real time. If the signal-to-noise ratio is too low, adjust the magnetic field strength of the permanent magnet until it recovers. If the adjustment does not recover after a preset time, deploy the robot to return to the relay node and deploy a reinforcement relay.

[0007] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention extracts the magnetic field distortion vector of the permanent magnet of the successfully anchored relay node and performs mutual interference separation. The distortion characteristics caused by the mutual interference of magnetic fields between adjacent permanent magnets are used as the source of ranging signal. The natural magnetic spacing between relay nodes is then calculated, and the legal execution spacing is obtained by scaling the measured signal attenuation. Based on this, the robot moves along the magnetic field direction to the target deployment position by comparing the adsorption force difference and guiding it with equal magnetic field lines. The anchoring and bonding calculation is performed by fusing the attenuation of the returned signal and multi-dimensional quality indicators. The above method can complete ranging, navigation and anchoring confirmation using the permanent magnet of the relay node itself, without relying on external sensors or coordinate positioning systems. This enables the automated deployment of communication relays for wall-climbing robots in the oil medium environment inside storage tanks.

[0008] 2. This invention detects the signal-to-noise ratio (SNR) of successfully anchored relay nodes in real time. When the SNR is extremely low, the magnetic field strength of the permanent magnet is first adjusted. By enhancing the magnet's adsorption force, the tightness of the connection between the relay node and the tank wall and the electrical contact are improved. The communication conditions are actively improved through physical layer means until the SNR recovers. Only when the adjustment is ineffective after exceeding the preset time is the robot triggered to return to the relay node to deploy a reinforcement relay. The magnetic field strength adjustment, magnetic field distortion ranging, and isomagnetic field guidance share the same physical field information framework, making the permanent magnet of the relay node a multifunctional core device integrating sensing, ranging, navigation, and self-healing. A complete closed loop based on the same physical field is formed between each step, thereby achieving self-maintaining signal quality of the communication link of the wall-climbing robot inside the storage tank. Attached Figure Description

[0009] Figure 1 This is a system architecture diagram of a communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption, provided in an embodiment of the present invention. Figure 2 This is a flowchart illustrating a method for deploying a communication relay for a wall-climbing robot in a petrochemical storage tank based on permanent magnet adsorption, according to an embodiment of the present invention.

[0010] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 belong to some, but not all, embodiments of the present invention. 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.

[0012] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0013] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0014] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0015] In practice, the server-side equipment deployed in the permanent magnet adsorption-based communication relay deployment system for petrochemical storage tank wall-climbing robots may consist of one or more devices. This system can be implemented as a business instance, a virtual machine, and hardware devices. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node to provide communication relay services to various user terminals. Alternatively, it can also be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage each user terminal. Alternatively, the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption can also be implemented as a server consisting of numerous identical or different types of hardware devices, with one or more hardware devices set up to provide the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption to each user terminal.

[0016] In terms of implementation, the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption and the user terminal are mutually adaptable. That is, if the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption is implemented as an application installed on a cloud service platform, then the user terminal is a client that establishes a communication connection with the application; or if the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption is implemented as a website, then the user terminal is implemented as a webpage; or if the communication relay deployment system for the petrochemical storage tank wall-climbing robot based on permanent magnet adsorption is implemented as a cloud service platform, then the user terminal is implemented as a mini-program in an instant messaging application.

[0017] like Figure 1 The figure shown is a system architecture diagram of a communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption, provided in an embodiment of the present invention.

[0018] The communication relay deployment system 100 for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption described in this invention can be set up in a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed into a website. Depending on the functions implemented, the communication relay deployment system 100 for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption may include an anchoring calculation module 101, a magnetic distance calculation module 102, a distance conversion module 103, a difference comparison module 104, an isomagnetic guidance module 105, and a reinforcement deployment module 106. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can perform a fixed function, stored in the memory of the electronic device.

[0019] In this embodiment of the invention, in the communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption, each of the above modules can be implemented independently and can call other modules. This calling can be understood as one module connecting to multiple modules of another type and providing corresponding services to those connected modules. In the communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption provided by this embodiment of the invention, the applicable scope of the system architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-based horizontal expansion to quickly and flexibly expand the system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in a virtual device, such as a service instance in a cloud server.

[0020] The following describes, with reference to specific embodiments, each component and its specific workflow of the communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption: The anchoring calculation module 101 is used to perform anchoring and bonding calculations on each relay node based on the fading amount and multi-dimensional quality indicators of the back-transmitted signals of each relay node during the robot's movement process, and to obtain the anchoring success signal of each relay node.

[0021] In this embodiment of the invention, the anchoring calculation module 101 receives the wireless signal fading amount and multi-dimensional quality indicators transmitted back in real time by each relay node during the robot's movement process.

[0022] The signal fading is measured and transmitted back in real time by the wireless communication unit built into the relay node. The fading is expressed as a numerical sequence in decibels, reflecting the degree of signal attenuation in the communication link between the relay node and the robot.

[0023] The multidimensional quality indicators include the contact resistance measurement between the permanent magnet of the relay node and the tank wall, the adsorption preparatory force fluctuation sequence of the permanent magnet, and the mechanical impact acceleration response at the moment of release of the relay node, etc. The parameters of each dimension are synchronously collected by the corresponding sensor at a fixed sampling frequency and associated with the same timestamp.

[0024] For processing signal fading, the anchoring calculation module 101 performs time-domain differential extraction on the fading of the returned signal.

[0025] Specifically, a discrete-time sequence of signal fading is acquired at a fixed sampling interval. The fading rate at each time point is obtained by performing a first-order backward difference operation on adjacent sampling points in the sequence, that is, subtracting the fading sample value of the previous time point from the fading sample value of the current time point.

[0026] The rate of change within a sliding window is statistically analyzed for the differencing sequence, with a window width of 10 sampling points.

[0027] Within a window, for a continuous interval where the absolute value of the rate of change is less than a preset gradual change threshold of 0.5 dB, the starting point of the interval is marked as the decay inflection point, and the average value of the rate of change within the interval is marked as the fading slope.

[0028] During the free movement phase after release, the antenna attitude of the relay node changes continuously, resulting in a large time-domain difference in signal fading with alternating positive and negative values, typically exceeding 2 dB.

[0029] As the relay node gradually adheres to the tank wall and tends to come to a standstill, the time-domain difference value of the signal fading decreases rapidly and tends to stabilize. The inflection point of attenuation corresponds precisely to the moment when the relay node transitions from a moving state to a stationary state.

[0030] Through the above time-domain differential extraction operation, the fading amount of the returned signal is converted into the fading slope and the position of the attenuation inflection point, which reflect its time-domain attenuation trend.

[0031] For the processing of multidimensional quality indicators, the anchoring solution module 101 performs principal quantity extraction on the multidimensional quality indicators.

[0032] The physical quantities included in the multidimensional quality indicators, such as contact resistance measurement, adsorption preparatory force fluctuation sequence and mechanical impact acceleration response, have different dimensions and numerical ranges. Direct integration can lead to the problem of dimensional dominance.

[0033] The anchoring calculation module 101 first performs dimensionless preprocessing on each physical quantity in the multidimensional quality index, that is, subtracting the historical mean from each measured value and dividing it by the historical standard deviation, and then uniformly converting each physical quantity into a dimensionless standard score.

[0034] Typical values ​​for contact resistance measurements range from 0.1 ohms to 10 ohms, typical values ​​for adsorption preparatory force fluctuation sequences range from 0.5 Newtons to 5 Newtons, and typical values ​​for mechanical impact acceleration responses range from 0.1 times the gravitational acceleration to 2 times the gravitational acceleration. After dimensionless conversion, all values ​​are converted to standard scores near the mean.

[0035] Principal component analysis is performed on the standard scores of each dimension to construct the covariance matrix of the standard scores of each dimension, and the eigenvalues ​​and eigenvectors of the covariance matrix are solved.

[0036] The eigenvectors are sorted in descending order of eigenvalues. The top principal components with a cumulative contribution rate of 85% are selected, and the projection values ​​of the original multidimensional quality index onto each principal component are used as the dimensionality-reduced representation feature sequence.

[0037] This characterization sequence retains most of the information of the original multidimensional quality indicators while eliminating the redundancy of correlation between physical quantities in each dimension.

[0038] The anchoring solution module 101 performs step-by-step fitting state determination for each relay node based on the decay gradient sequence and the characterization feature sequence.

[0039] The progressively fitting status adjudication adopts a two-level judgment structure.

[0040] The first level is the initial judgment of the communication dimension: taking the position of the attenuation inflection point as the time reference, the fading slope value within the preset time window before and after that moment, i.e., within 5 sampling intervals, is extracted. If the fading slope is less than the gradual change threshold of 0.5 dB, the communication dimension alignment status of the relay node is determined to be up to standard, and a communication dimension initial judgment standard mark is generated.

[0041] If the fading slope is greater than or equal to the gradual change threshold, it is judged as non-compliant, and a communication dimension initial non-compliant mark is generated.

[0042] The second level is physical dimension re-judgment: taking the value of the corresponding time in the characteristic feature sequence as input, the value is compared with the preset physical fit threshold.

[0043] The physical bonding threshold is set based on the following: when the permanent magnet of the relay node is fully bonded to the tank wall, the contact resistance decreases from more than 10 ohms to less than 0.5 ohms, the adsorption preparatory force fluctuation decreases from more than 1.5 Newtons to less than 0.3 Newtons, and the impact response decays to less than 0.2 times the gravitational acceleration. The above physical effects are concentrated in the numerical changes of the characterization feature sequence after the principal fraction is extracted from the multidimensional quality indicators.

[0044] If the value of the characteristic sequence is lower than the physical fit threshold, the physical dimension fit status of the relay node is determined to be up to standard, and a physical dimension re-judgment standard mark is generated.

[0045] Conversely, if the condition is not met, it is judged as substandard, and a physical dimension re-judgment substandard mark is generated.

[0046] The temporal characteristics of signal fading and the fusion characteristics of multi-dimensional quality indicators independently adjudicate the alignment status of relay nodes in the communication and physical dimensions, respectively. If either dimension fails to meet the standard, a redeployment process is triggered, thereby avoiding misjudgments that may occur in complex oil medium environments due to single-dimensional criteria.

[0047] The anchoring solution module 101 performs time-series cross-verification of the fit confidence sequences of each relay node.

[0048] The bonding confidence sequences of each relay node are merged along the time axis according to the release time of the corresponding relay node. The release time of the relay node with the earliest release time is set as the zero point of a unified time scale, and the release times of the other relay nodes are converted by offset relative to this zero point.

[0049] On a unified time scale, the consistency of the bonding confidence of each relay node within a preset sliding window with a window width preferably of 3 seconds is checked.

[0050] If the bonding confidence of all relay nodes within the window is in a qualified state, then it is determined that the anchoring status of each relay node within the window is synchronously qualified.

[0051] If any relay node within a window has a non-compliant fit confidence level, then the window is considered to have an anchoring anomaly.

[0052] The synchronization compliance status is continuously confirmed for a certain duration. When the synchronization compliance is determined to be met in three consecutive sliding windows, an anchoring qualification mark is generated for each relay node.

[0053] The anchoring qualification flags of each relay node are comprehensively evaluated. If all anchoring qualification flags of all relay nodes are valid, an anchoring success signal is generated and output to the system bus.

[0054] If any relay node has an invalid anchoring qualification flag, an anchoring failure flag is generated for that relay node, triggering its re-release and re-anchoring process.

[0055] Through the above-mentioned cross-verification of timing, the alignment status information of each relay node is cross-referenced and verified in the time dimension. Occasional misjudgments caused by instantaneous interference of a single relay node are effectively filtered out by the sliding window mechanism, and the consistency of anchoring timing between nodes is also guaranteed.

[0056] The magnetic distance calculation module 102 is used to perform mutual interference separation on the magnetic field line distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic distance of the relay node.

[0057] In this embodiment of the invention, after the anchored relay node permanent magnet is adsorbed onto the surface of the tank wall steel plate, a static magnetic field generated by the permanent magnet itself exists in the space around it.

[0058] When multiple relay nodes are anchored to different positions on the tank wall, the magnetic fields generated by each permanent magnet are superimposed in space. The magnetic fields between adjacent permanent magnets affect each other, causing the distribution of their respective magnetic field lines to bend and compress relative to the isolated state. This deformation is called magnetic field line distortion.

[0059] The degree of magnetic field line distortion is directly related to the spatial distance between adjacent permanent magnets. The smaller the distance, the more significant the magnetic field superposition effect and the more severe the magnetic field line distortion. The larger the distance, the more independent the magnetic fields of each permanent magnet tend to be, and the degree of distortion is reduced.

[0060] The magnetic distance calculation module 102 obtains the component values ​​of the magnetic induction intensity in three orthogonal directions at each spatial point by performing three-component fluxgate sensing on the space around the successfully anchored relay node permanent magnet.

[0061] The three-component fluxgate sensors are deployed in the space surrounding the anchored relay node permanent magnet. The sensor array is arranged in a three-dimensional grid structure with the relay node permanent magnet as the center and the sensors are equally spaced along three orthogonal directions.

[0062] The sensor synchronously collects three orthogonal components of magnetic induction intensity at each grid node, and organizes the values ​​of each component according to spatial coordinates to form a spatial magnetic field vector sequence.

[0063] Each vector in this spatial magnetic field vector sequence contains three components, which correspond to the intensity and direction of the magnetic field at a certain point in space in three orthogonal directions.

[0064] For the processing of the spatial magnetic field vector sequence, the magnetic moment estimation module 102 performs background magnetic field cancellation on the spatial magnetic field vector sequence.

[0065] The background magnetic field originates from the Earth's magnetic field and interference fields generated by other magnetic objects in the environment surrounding the storage tank. Its intensity is relatively uniform and its direction is consistent at all points in space.

[0066] Background magnetic field cancellation employs a gradient magnetic field sensor configuration. Two spatially adjacent magnetoresistive elements sense the magnetic field value at the same location, and the outputs of the two are differentially calculated to cancel out the relatively uniform background magnetic field component.

[0067] In the canceled output signal, the uniform background magnetic field component is significantly suppressed, while the non-uniform magnetic field component generated by the permanent magnet is retained.

[0068] By performing background magnetic field cancellation processing on the vectors at each point in the spatial magnetic field vector sequence, the contribution of the background field is removed from the magnetic field vectors at each point, resulting in a pure magnetic vector field.

[0069] This pure magnetic vector field contains only the magnetic field components generated by each permanent magnet itself, and its spatial distribution directly reflects the position and magnetization state of the permanent magnet.

[0070] The magnetic moment estimation module 102 performs curvature analysis on the pure magnetic vector field.

[0071] The spatial geometry of magnetic field lines is completely characterized by the curvature vector at each point on the magnetic field line. The magnitude of the curvature vector is equal to the measure of the degree of curvature of the magnetic field line at that point, and its direction points to the center of curvature of the magnetic field line at that point.

[0072] Curvature analysis is achieved by solving the spatial gradient of the unit magnetic field vector at each point in the pure magnetic vector field.

[0073] Taking a point in the pure magnetic vector field as the center, multiple magnetic field vector sampling values ​​are taken in a preset neighborhood around the point. Spatial difference operation is performed on each sampling value to obtain the rate of change of the magnetic field direction at the point with spatial displacement.

[0074] This rate of change is expressed as a vector in three-dimensional space. Its magnitude represents the degree of curvature of the magnetic field lines at that point, and its direction represents the orientation of the curvature.

[0075] By traversing all spatial points in the pure magnetic vector field, the spatial gradient calculation described above is performed on each point in sequence to obtain the curvature vector corresponding to each point. The curvature vectors of all points are then organized according to spatial coordinates to form a magnetic field distortion vector field.

[0076] The curvature vector at each point in the magnetic field distortion vector field reflects the degree of deviation of the magnetic field lines from a straight state at that point. The larger the magnitude of the curvature vector, the more severe the bending of the magnetic field lines at that point, i.e., the more significant the distortion.

[0077] The magnetic moment estimation module 102 performs field source separation on the magnetic field distortion vector field.

[0078] In a scenario where multiple permanent magnets coexist, the magnetic field distortion vector field is a linear superposition of the distortion contributions generated by each permanent magnet. Each permanent magnet generates a specific distortion distribution pattern in the space around it, and the superposition of all patterns forms the observed total distortion field.

[0079] The field source separation adopts a multipole expansion method, which expands the magnetic field distortion vector field into a multipole component sequence according to the spherical harmonic function.

[0080] Multipole expansion decomposes the spatial magnetic field into contribution components of different orders, such as monopoles, dipoles, quadrupoles, and octopuses. Among them, the dipole component corresponds to the dominant contribution of a single permanent magnet in the far field, while the higher-order components correspond to the near-field details.

[0081] The multipole components of each order are orthogonal to each other, and a change in any one order component will not affect the values ​​of other order components.

[0082] Basis mapping is performed on each of the components after multipole expansion, and each component is projected onto a set of pre-constructed orthogonal basis functions. Each basis function in this set corresponds to a standard distortion mode produced by a permanent magnet at a specific spatial location.

[0083] By solving for the projection coefficients of the measured distortion vector field onto each basis function, the contribution of each basis function can be obtained.

[0084] The magnitude of this contribution directly reflects the degree to which the permanent magnet at the corresponding spatial location participates in the total distortion field.

[0085] The contribution of each basis function is filtered by a threshold. Basis functions with contribution values ​​lower than the preset effective threshold are determined to have no corresponding permanent magnet and their contribution values ​​are set to zero. Basis functions with contribution values ​​higher than the preset effective threshold are determined to have a corresponding permanent magnet and their contribution values ​​are retained as independent components of the permanent magnet.

[0086] After the above field source separation operation, the independent components of the contribution of each permanent magnet in the magnetic field distortion vector field are separated and extracted one by one.

[0087] The spatial distribution of each independent component is determined solely by the corresponding single permanent magnet and is unaffected by the magnetic field of adjacent permanent magnets.

[0088] The magnetic distance calculation module 102 performs gradient inversion based on each independent component to obtain the natural magnetic distance of the permanent magnet of the relay node.

[0089] Gradient inversion utilizes the linear relationship between the magnetic field vector and its spatial gradient components to invert the distance of the field source relative to the measurement point by measuring the magnetic field gradient tensor.

[0090] For a permanent magnet corresponding to a certain independent component, the magnetic field vector values ​​of multiple spatial points in a preset neighborhood around the permanent magnet are taken to construct the magnetic field gradient tensor matrix in that region.

[0091] The magnetic field gradient tensor matrix contains the rates of change of the three components of the magnetic field in three spatial directions, and has a total of nine elements.

[0092] According to the magnetic dipole field source model, there is a definite mathematical relationship between the magnetic field generated by the field source and its gradient tensor. Each element of the gradient tensor is inversely proportional to the cube of the distance from the field source to the measurement point.

[0093] The magnetic moment estimation module 102 performs gradient tensor shrinkage calculation on the magnetic field distortion vector field to obtain the estimated distance from the field source to the measurement point.

[0094] The formula for calculating the distance estimate is as follows: ; In the formula, r is the estimated distance from the field source to the measurement point. Let be the element in the i-th row and j-th column of the magnetic field gradient tensor, and k be a known constant combining the magnetic moment of the permanent magnet and the permeability of the vacuum. The magnetic field gradient tensor is composed of the rate of change of the three components of the magnetic induction intensity at each point in the space surrounding the permanent magnet in three spatial directions.

[0095] In this calculation formula, the denominator is the sum of the squares of all nine elements of the magnetic field gradient tensor, which is inversely proportional to the distance from the field source to the eighth power. Thus, the estimated distance from the field source to the measurement point can be directly obtained through closed-loop analytical inversion of the magnetic field gradient tensor data of a single measurement point.

[0096] This distance estimate is the spatial distance from the permanent magnet to the center of the measurement point array.

[0097] Perform the gradient inversion operation on each of the two adjacent permanent magnets that have been anchored to obtain the distance values ​​from each permanent magnet to the same measurement reference point.

[0098] The summation of the two distance values ​​yields the natural magnetic spacing between two adjacent permanent magnet relay nodes.

[0099] Gradient inversion utilizes the linear relationship between the magnetic field vector and its spatial gradient components to invert the distance of the field source relative to the measurement point by measuring the magnetic field gradient tensor.

[0100] For a permanent magnet corresponding to a certain independent component, the magnetic field vector values ​​of multiple spatial points in a preset neighborhood around the permanent magnet are taken to construct the magnetic field gradient tensor matrix in that region.

[0101] The magnetic field gradient tensor matrix contains the rates of change of the three components of the magnetic field in three spatial directions, and has a total of nine elements.

[0102] According to the magnetic dipole field source model, there is a definite mathematical relationship between the magnetic field generated by the field source and its gradient tensor. Each element of the gradient tensor is inversely proportional to the cube of the distance from the field source to the measurement point.

[0103] By substituting the measured gradient tensor matrix into this mathematical relationship and solving in reverse, the spatial distance from the permanent magnet to the center of the measurement point array can be obtained.

[0104] Perform the gradient inversion operation on each of the two adjacent permanent magnets that have been anchored to obtain the distance values ​​from each permanent magnet to the same measurement reference point.

[0105] The summation of the two distance values ​​yields the natural magnetic spacing between two adjacent permanent magnet relay nodes.

[0106] The natural magnetic spacing is expressed in units of length, reflecting the actual spatial interval between adjacent relay nodes on the tank wall surface, and serves as the basic input for subsequent spacing conversion modules to perform scale conversion.

[0107] Through the complete embodiment of the magnetic distance estimation module 102 described above, the magnetic field distortion information of the anchored relay node permanent magnet is effectively extracted and separated into the independent contribution of each permanent magnet. Based on this, the natural magnetic distance between adjacent relay nodes is obtained through gradient inversion.

[0108] This natural magnetic spacing does not rely on external ranging sensors or coordinate positioning systems. It can be measured using only the magnetic field information of the permanent magnet itself, providing a spacing reference value for the subsequent spacing conversion module that is not affected by the oil medium environment.

[0109] The spacing conversion module 103 is used to convert the natural magnetic spacing to a scale using the signal attenuation of the successfully anchored relay node as a reference, and to obtain the legal execution spacing between the adjacent relay nodes. In this embodiment of the invention, the spacing conversion module 103 receives the natural magnetic spacing between adjacent relay nodes output by the magnetic distance calculation module 102, and at the same time receives the signal fading amount of the successfully anchored relay node measured by the robot.

[0110] The signal fading is expressed as a numerical sequence in decibels and is obtained by real-time measurement of the wireless signal transmitted back by the anchored relay node as the robot moves along the tank wall.

[0111] Signal fading includes large-scale fading components and small-scale fading components. The large-scale fading component reflects the slow attenuation trend of the signal as the propagation distance increases, while the small-scale fading component reflects the rapid fluctuations caused by multipath effects.

[0112] The spacing conversion module 103 extracts large-scale fading components from the signal fading of successfully anchored relay nodes.

[0113] Large-scale fading component extraction employs a spatial averaging method, taking a spatial window of a preset length along the robot's motion path. The preferred window length is the motion distance corresponding to 10 signal sampling points.

[0114] The arithmetic mean of the signal attenuation samples within the window is calculated by summing the signal attenuation values ​​of each sample point within the window and dividing by the total number of sample points to obtain the average attenuation within the window.

[0115] Using the center position of the window as the spatial coordinate of the average fading amount, the window is slid along the robot's movement direction, sampling points one by one. The average value within the window is recalculated after each sampling point is slid, thus obtaining the average fading amount sequence corresponding to each spatial point.

[0116] In this average fading sequence, the rapid fluctuations caused by small-scale fading are smoothed and suppressed, while the slow changing trends caused by large-scale fading are preserved, forming a slow-varying fading envelope.

[0117] Through the above large-scale fading component extraction operation, the small-scale fading components are stripped from the original signal fading amount, and only the slowly varying fading envelope reflecting the signal attenuation trend with distance is retained.

[0118] The spacing conversion module 103 fits the distance-loss relationship to the slowly varying fading envelope.

[0119] Using the spatial distance between the robot's current position and the anchored relay node as the independent variable, and the average fading value at the corresponding distance in the slow fading envelope as the dependent variable, a set of distance loss sample points is constructed.

[0120] Each sample point in the sample point set contains a distance value and its corresponding average fading value, and multiple sample points are distributed at different spatial locations on the robot's motion path.

[0121] Linear regression analysis was performed on the sample point set, with the logarithm of the distance as the x-axis and the logarithm of the average fading as the y-axis, to fit the sample points into a straight line.

[0122] The slope of the line represents the rate at which the signal fading changes with the logarithm of the distance, i.e., the path loss exponent; the intercept of the line represents the reference fading at the reference distance.

[0123] The fitting process employs the least squares criterion, determining the optimal slope and intercept of the line by minimizing the sum of squared vertical distances from each sample point to the fitted line.

[0124] The steeper the slope of the line, the faster the signal decays with distance, and the more severe the signal loss caused by the oil medium; the smaller the slope of the line, the slower the signal decays with distance, and the less severe the signal loss caused by the oil medium.

[0125] Through the above distance loss relationship fitting operation, the slowly varying fading envelope is converted into a path loss index, which exists in the form of a dimensionless scalar and quantitatively characterizes the attenuation characteristics of wireless signals in the current oil medium environment.

[0126] The spacing conversion module 103 performs path loss index weighted scaling on the natural magnetic spacing.

[0127] Using the path loss index as a scaling factor for scale conversion, the natural magnetic spacing is divided by the ratio of the path loss index to the preset reference index to obtain the corrected spacing.

[0128] The preset benchmark index is the standard path loss index when the oil medium has no additional attenuation effect on the signal, with a preferred value of 2.0, corresponding to the free space propagation environment.

[0129] When the measured path loss index is greater than the benchmark index, it indicates that the oil medium loss is higher than the benchmark level. The natural magnetic spacing is reduced proportionally, making the relay nodes more densely deployed to compensate for the additional loss.

[0130] When the measured path loss index is less than the benchmark index, it indicates that the oil medium loss is lower than the benchmark level. The natural magnetic spacing is proportionally amplified, making the relay node deployment more sparse to reduce unnecessary deployment costs.

[0131] Through the above path loss exponent weighted scaling operation, the natural magnetic spacing is adaptively amplified or reduced according to the actual signal attenuation characteristics of the current oil medium, thus obtaining the corrected spacing.

[0132] The spacing conversion module 103 performs communication constraint boundary regularization on the corrected spacing.

[0133] The communication constraint boundary includes an upper limit boundary and a lower limit boundary. The upper limit boundary is determined by the maximum coverage radius of the wireless communication module of the relay node, and the lower limit boundary is determined by the effective range of the adsorption force of the permanent magnet of the relay node.

[0134] The corrected spacing is compared with the upper limit boundary. If the corrected spacing is greater than the upper limit boundary, the corrected spacing is assigned the upper limit boundary value.

[0135] The correction spacing is compared with the lower limit boundary. If the correction spacing is less than the lower limit boundary, the correction spacing is assigned the lower limit boundary value.

[0136] If the correction spacing is between the upper and lower limits, the original value of the correction spacing is retained.

[0137] Through the above-mentioned communication constraint boundary regularization operation, the correction spacing is constrained within the feasible range jointly determined by the physical communication capability and permanent magnet adsorption capability of the relay node, thus obtaining the statutory execution spacing.

[0138] The statutory execution interval is expressed in units of length, which reflects the signal attenuation characteristics of the current oil medium environment, and also meets the physical constraints of the relay node's own communication coverage and adsorption retention, serving as the basic input for subsequent isomagnetic guidance modules to perform path planning and position determination.

[0139] Through the complete embodiment of the above-mentioned spacing conversion module 103, the signal fading of the successfully anchored relay node is effectively converted into a path loss index that reflects the attenuation characteristics of the oil medium. The natural magnetic spacing is adaptively scaled according to this index and then normalized by the boundary to obtain the legal execution spacing.

[0140] This statutory execution spacing integrates the physical spacing benchmark obtained from magnetic ranging with the channel attenuation information obtained from actual communication measurements, providing a target deployment spacing for subsequent isomagnetic guidance modules that takes into account both magnetic characteristics and communication quality.

[0141] The difference comparison module 104 is used to compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position.

[0142] In this embodiment of the invention, the difference comparison module 104 receives the measured value of the permanent magnet adsorption force at the current position of the robot and the preliminary adsorption force reference value of the permanent magnet of the relay node.

[0143] The permanent magnet attraction force at the robot's current position is sensed in real time by a force sensing unit located at the connection between the robot's permanent magnet attraction device and the vehicle body.

[0144] The force sensing unit uses a piezoresistive or strain gauge force sensor. Its sensitive element is connected in series with the force transmission path of the permanent magnet adsorption device. The adsorption force between the permanent magnet and the tank wall is transmitted to the force sensing unit through the force transmission path. The force sensing unit outputs an electrical signal that is proportional to the pressure it bears.

[0145] The electrical signal is amplified and filtered by the signal conditioning circuit, and then converted into a digital quantity by the analog-to-digital converter, which serves as the real-time suction force amplitude at the robot's current position.

[0146] For processing the real-time adsorption force amplitude at the robot's current position, the difference comparison module 104 performs real-time adsorption force sensing at each spatial point on the robot's motion path.

[0147] Specifically, during the robot's movement along the tank wall, the force sensing unit's data latching operation is triggered at a fixed spatial sampling interval. The preferred sampling interval is once every 10 millimeters the robot advances.

[0148] Each time data is latched, the current electrical signal output by the force sensing unit is held and sent to the analog-to-digital conversion channel. After analog-to-digital conversion, the digital quantity of the adsorption force corresponding to that spatial point is obtained.

[0149] The digital values ​​of the adsorption force at each spatial point are arranged sequentially according to the robot's movement direction to form a dynamic adsorption force amplitude sequence. Each value in this sequence corresponds to a specific spatial point on the robot's movement path.

[0150] The dynamic adsorption force amplitude sequence includes both the steady-state adsorption force component when the robot moves along a flat wall surface and the adsorption force fluctuation component caused by local unevenness of the wall surface, weld protrusions, or curvature changes.

[0151] The difference comparison module 104 performs trend term stripping on the dynamic adsorption force amplitude of each spatial point of the robot.

[0152] The dynamic adsorption force amplitude sequence is composed of two superimposed parts: a trend term and a fluctuation term. The trend term reflects the slow drift of the adsorption force as the robot's spatial position changes, while the fluctuation term reflects the rapid fluctuations caused by local features of the wall.

[0153] Trend term stripping uses a sliding window averaging method, taking the adsorption force values ​​of the five adjacent sampling points before and after the current spatial point as the center to form a local window.

[0154] The arithmetic mean of the adsorption force values ​​within the window is used as the estimated value of the trend term at the center point of the window.

[0155] The window is slid along the direction of robot movement, sampling points one by one. The average value within the window is recalculated after each sampling point is slid, resulting in a sequence of estimated trend terms for each spatial point.

[0156] The residual fluctuation components of each spatial point are obtained by subtracting the corresponding trend term estimate from the original dynamic adsorption force amplitude sequence point by point.

[0157] The sequence of estimated trend terms is used as the steady-state adsorption force amplitude at each spatial point. This steady-state adsorption force amplitude reflects the baseline level of the robot's permanent magnet adsorption force under the current wall conditions.

[0158] Through the trend stripping operation described above, the rapidly fluctuating components in the dynamic adsorption force amplitude are filtered out, leaving only the steady-state adsorption force amplitude that reflects the baseline level of adsorption force.

[0159] The difference comparison module 104 synchronously tracks and locks the pre-adsorption force of the permanent magnet of the relay node.

[0160] The initial adsorption force refers to the initial adsorption force generated between the permanent magnet and the tank wall of the relay node before release. This value is continuously measured by the force-sensitive unit set at the mechanical locking part of the relay node during the movement of the robot carrying the relay node.

[0161] When the relay node is carried by the robot along the tank wall, the air gap between its permanent magnet and the tank wall remains constant, and the pre-adsorption force is maintained near a relatively stable reference value.

[0162] As the relay node moves between different positions on the tank wall, the changes in the local magnetic properties of the tank wall cause slight fluctuations in the pre-adsorption force.

[0163] The difference comparison module 104 samples the output signal of the force-sensitive unit of the relay node at fixed time intervals, and the sampling frequency is preferably 100 Hz.

[0164] The median filter within a sliding window is applied to the sampled pre-adsorption force time series. The window width is preferably 5 sampling points. The median value of each sampled value in the window is taken as the filtered output of the window after being sorted in ascending order of amplitude.

[0165] The latest valid value in the filtered output sequence is used as the reserve force amplitude at the current moment, and the update of this value as the spatial location of the relay node changes is continuously tracked.

[0166] Through the above synchronous tracking and locking operation, the preparatory force amplitude of the relay node permanent magnet is acquired in real time and kept updated synchronously with the robot's motion process.

[0167] The difference comparison module 104 calculates the difference between the steady-state adsorption force amplitude and the preparatory force amplitude after registering them at the same point.

[0168] Same-point registration refers to aligning the steady-state adsorption force amplitude measured by the robot at a certain spatial point with the preparatory force amplitude of the relay node at the same spatial point.

[0169] Because the robot moves synchronously with the relay node, the robot's spatial position and the relay node's spatial position completely overlap during the movement.

[0170] Using the robot's current spatial position as an index, retrieve the steady-state adsorption force amplitude corresponding to that position and the preparatory force amplitude output by the force-sensitive unit of the relay node at the same moment.

[0171] Subtracting the preparatory force amplitude from the steady-state adsorption force amplitude yields the instantaneous adsorption force deviation at that spatial location.

[0172] The above registration and difference operations are performed on multiple consecutive spatial points along the robot's motion path to obtain the instantaneous adsorption force deviation sequence corresponding to each point.

[0173] A positive instantaneous adsorption force deviation indicates that the robot's current adsorption force is greater than the relay node's preparatory force, and the robot can overcome the adsorption resistance when releasing the relay node and detach smoothly. A negative instantaneous adsorption force deviation indicates that the robot's current adsorption force is insufficient to overcome the relay node's preparatory force, and the robot may not be able to release the relay node from its carrying position.

[0174] The mean value of the instantaneous adsorption force deviation sequence is calculated within a sliding window. The window width is preferably 10 spatial points. The sum of the instantaneous adsorption force deviations of each point within the window is divided by the total number of points within the window to obtain the average adsorption force deviation corresponding to the window.

[0175] The average adsorption force deviation is output as the adsorption force difference at the robot's current position to the isomagnetic guidance module.

[0176] The adsorption force difference, expressed in units of force, reflects the surplus or deficiency of the robot's current adsorption capacity relative to the prepared adsorption force of the relay node. It provides input basis for the isomagnetic guidance module to determine whether the robot has the mechanical conditions to release the relay node at the current position and to determine the direction of the isomagnetic field lines.

[0177] Through the complete embodiment of the above-mentioned difference comparison module 104, the real-time adsorption force of the robot's current position and the preparatory adsorption force of the relay node are respectively stripped by trend term and locked by synchronous tracking. They are then registered and the difference is calculated at the same spatial point to obtain the adsorption force difference that reflects the degree of surplus of the robot's adsorption capacity relative to the preparatory force of the relay node.

[0178] The difference in adsorption force provides a mechanical criterion for the subsequent isomagnetic guidance module to plan the path of isomagnetic lines and determine the location of relay deployment.

[0179] The isomagnetic guidance module 105 is used to guide the robot path with isomagnetic lines based on the adsorption force difference and the statutory execution interval, so as to obtain the relay deployment position of the robot and anchor the relay node.

[0180] In this embodiment of the invention, the equal magnetic guidance module 105 receives the adsorption force difference output by the difference comparison module 104 and the legal execution distance output by the distance conversion module 103.

[0181] The adsorption force difference is expressed in units of force and reflects the surplus of the permanent magnet adsorption force at the robot's current position relative to the prepared adsorption force of the relay node. The spatial distribution of this surplus corresponds to the density and direction of the magnetic field lines on the surface of the tank wall.

[0182] The statutory execution interval is expressed in units of length and reflects the target deployment interval between adjacent relay nodes after correction for the attenuation characteristics of oil media.

[0183] The isomagnetic guidance module 105 calculates the spatial gradient direction of the adsorption force difference.

[0184] Specifically, taking the robot's current position as the center, four adjacent points are selected within the robot's motion plane: a forward point, a backward point, and a left and right point perpendicular to the direction of travel. The spatial interval between each adjacent point and the center point is equal, preferably 10 millimeters.

[0185] The adsorption force difference samples at the four adjacent points are obtained respectively, and each sample value is associated with its corresponding spatial coordinates.

[0186] Perform a difference operation on the adsorption force difference between the forward and backward points, subtract the adsorption force difference between the forward and backward points from the adsorption force difference at the forward point, and divide the result by the spatial distance between the two points to obtain the directional derivative of the adsorption force difference in the direction of travel.

[0187] Perform a difference operation on the adsorption force difference between the left and right points. Subtract the adsorption force difference of the right point from the adsorption force difference of the left point, and divide the result by the spatial distance between the two points to obtain the directional derivative of the adsorption force difference perpendicular to the direction of travel.

[0188] The directional derivatives in the direction of travel and the directional derivatives in the vertical direction together form the spatial gradient vector of the adsorption force difference. The direction of this vector represents the direction in which the adsorption force difference changes the fastest in space.

[0189] In the magnetic field distribution on the surface of the tank wall, the contour lines of the adsorption force difference are orthogonal to the tangential direction of the magnetic field lines, and the direction in which the adsorption force difference changes the fastest is the normal direction of the magnetic field lines.

[0190] Rotate the spatial gradient vector 90 degrees around the normal direction of the robot's current position to obtain a direction orthogonal to the gradient direction. This direction is the tangential direction of the isomagnetic field lines at the robot's current position.

[0191] The tangential direction of the magnetic field lines is expressed as a unit vector, and the direction of this unit vector indicates the instantaneous orientation that the robot should follow when moving along the magnetic field lines.

[0192] Through the above spatial gradient direction calculation operation, the spatial distribution information of the adsorption force difference is converted into the tangential direction of the isomagnetic field lines at the robot's current position.

[0193] The isomagnetic guidance module 105 performs path recursion for the robot based on the tangential direction of the isomagnetic lines and the statutory execution interval.

[0194] The path recursion adopts a step-by-step iterative method, which divides the statutory execution interval into several recursion steps of equal length, with the preferred step length being 5 mm.

[0195] Starting from the robot's current position, move the robot one step along the tangential direction of the isomagnetic field lines to obtain the first recursive position.

[0196] At this recursive position, the spatial gradient direction is recalculated to obtain the tangential direction of the magnetic field lines at this position. Then, the position is moved by one recursive step along the new tangential direction to obtain the second recursive position.

[0197] Repeat the above step movement and direction update operations, accumulating the displacement after each step, and summing the step size of each movement until the accumulated displacement reaches the value of the statutory execution interval.

[0198] Connect all the recursive positions passed through in the above stepping process in the order of movement to form a path of equal magnetic field lines from the robot's current position to the target deployment position.

[0199] Each point on the path of these magnetic field lines is located on the same magnetic field line, and when the robot moves along the path, its direction of travel is always consistent with the tangent of the magnetic field line at its location.

[0200] Through the above path recursion operation, the statutory execution spacing and the tangential direction of the isomagnetic field lines are jointly converted into an isomagnetic field line path that the robot can execute.

[0201] The isomagnetic guidance module 105 performs magnetic tangential tracking of the robot's real-time position on the isomagnetic field path.

[0202] During the robot's movement along the path of isomagnetic field lines, the robot's current spatial coordinates are acquired at fixed time intervals, with the sampling interval preferably being 50 milliseconds.

[0203] Compare the current spatial coordinates with the coordinates of the corresponding reference point on the path of the isomagnetic field lines, and calculate the lateral offset of the current spatial coordinates relative to the reference point coordinates in the direction perpendicular to the path.

[0204] Based on this lateral offset, the angle between the robot's current actual direction of motion and the path reference direction is calculated.

[0205] This included angle, measured in degrees, represents the degree of deviation of the robot's actual heading from the desired tangential direction of the magnetic field lines, i.e., the magnetic field line tracking yaw angle.

[0206] A positive yaw angle for magnetic field line tracking indicates that the robot's heading is biased to the right of the path, a negative value indicates that the robot's heading is biased to the left of the path, and zero indicates that the robot's heading is completely consistent with the tangential direction of the path.

[0207] Through the above magnetic field line tangential tracking operation, the robot's real-time pose deviation on the path of equal magnetic field lines is quantified as the magnetic field line tracking yaw angle.

[0208] The isomagnetic guidance module 105 tracks the yaw angle based on the magnetic field lines, drives the robot to move along the isomagnetic field line path to the spatial point corresponding to the statutory execution interval, and uses the spatial point as the robot's relay deployment position and triggers the relay node anchoring.

[0209] Specifically, when the yaw angle of magnetic field tracking is not zero, differential speed correction is performed on the rotational speeds of the left and right drive wheels of the robot based on the sign and magnitude of the yaw angle.

[0210] When the yaw angle is positive when the magnetic field line tracking is positive, increase the speed of the right drive wheel and decrease the speed of the left drive wheel to make the robot's heading yaw to the right to reduce the yaw angle.

[0211] When the yaw angle is negative when tracking magnetic field lines, increase the speed of the left drive wheel and decrease the speed of the right drive wheel to make the robot's heading yaw to the left to reduce the yaw angle.

[0212] The magnitude of differential correction is proportional to the absolute value of the yaw angle tracked by the magnetic field lines—the larger the absolute value of the yaw angle, the greater the speed difference between the two drive wheels, and the stronger the corrective steering torque.

[0213] When the yaw angle of the magnetic field line tracking is zero, keep the speed of the left and right drive wheels equal, so that the robot can move in a straight line in the current direction.

[0214] During the robot's movement, the above-mentioned tangential tracking and differential correction operations are continuously performed, enabling the robot to follow the path of the isomagnetic field lines in real time until the cumulative displacement reaches the legally defined execution interval.

[0215] When the robot's cumulative displacement reaches the legally defined execution interval, the robot's current spatial location is determined as the relay deployment location.

[0216] At the relay deployment location, the magnetic guidance module 105 sends a release command to the mechanical locking component of the relay node, triggering the permanent magnet of the relay node to adhere to the surface of the tank wall under the action of the pre-adsorption force, thus completing the anchoring operation of the relay node.

[0217] Through the complete embodiment of the above-described magnetic field guidance module 105, the magnetic field line spatial distribution information provided by the adsorption force difference and the target deployment distance provided by the statutory execution interval are used together to generate a magnetic field line path. The robot autonomously moves to the target position along the path through real-time tangential tracking and differential speed correction and completes the anchoring of the relay node.

[0218] This guidance method does not rely on GPS signals, visual features, or inertial coordinate accumulation, and can still achieve accurate positioning and reliable deployment in environments with insufficient light, oil mist, and missing positioning signals inside the storage tank.

[0219] The reinforcement deployment module 106 is used to detect the signal-to-noise ratio of the successfully anchored relay node in real time. If the signal-to-noise ratio is too low, the magnetic field strength of the permanent magnet is adjusted until it recovers. If the adjustment does not recover after a preset time, the deployment robot returns to the relay node to deploy the reinforcement relay.

[0220] In this embodiment of the invention, the reinforcement deployment module 106 receives the real-time signal-to-noise ratio measurement sequence transmitted back by the successfully anchored relay node.

[0221] The signal-to-noise ratio (SNR) measurement is continuously measured by the wireless communication unit of the relay node during communication and transmitted back at fixed time intervals. The sampling interval is preferably 100 milliseconds. Each measurement is in decibels and reflects the signal quality level of the current communication link.

[0222] The signal-to-noise ratio sequence is arranged continuously on the time axis, forming a real-time monitoring data stream that reflects the evolution of communication link quality over time.

[0223] The reinforcement deployment module 106 performs sliding window mean filtering on the real-time back transmission signal-to-noise ratio of successfully anchored relay nodes.

[0224] Specifically, taking the current moment as the end of the window, a preset number of continuous signal-to-noise ratio samples are taken to form a filtering window, and the window width is preferably 10 sampling points.

[0225] The arithmetic mean of all signal-to-noise ratio (SNR) samples within the window is calculated. The sum of the SNR values ​​of each sample point within the window is then divided by the total number of sample points to obtain the filtered output value corresponding to that window.

[0226] The window is slid forward along the time axis, sampling point by sampling point. At each new sampling time, the end of the window is updated to the current time, the earliest sampled value at the beginning of the window is discarded, the latest sampled value is included, and the average value within the window is recalculated.

[0227] After sliding window mean filtering, random spike fluctuations caused by instantaneous interference in the signal-to-noise ratio sequence are smoothed and suppressed, resulting in a signal-to-noise ratio filtered sequence.

[0228] This signal-to-noise ratio (SNR) filtering sequence retains the trend information of SNR changes with the working status of relay nodes and environmental conditions, while filtering out high-frequency random noise.

[0229] The reinforcement deployment module 106 performs low-threshold cross-pass inspection on the signal-to-noise ratio filtered sequence of successfully anchored relay nodes.

[0230] The preset low quality threshold is in decibels and is determined by those skilled in the art based on the receiving sensitivity of the relay node's wireless communication module and the required communication reliability of the system. The preferred value is 10 decibels.

[0231] The value of each sampling point in the signal-to-noise ratio filtering sequence is compared with a preset low-quality threshold. If the sampled value is lower than the low-quality threshold, the sampling point is marked as a low signal-to-noise ratio state.

[0232] The duration of the low signal-to-noise ratio (SNR) state on the time axis is accumulated and counted, starting from the moment when the SNR first falls below the threshold and continuing until the SNR rises back above the threshold.

[0233] If the duration of the low signal-to-noise ratio state exceeds the preset trigger confirmation duration, preferably 2 seconds, then it is determined that the signal-to-noise ratio is in a continuous ultra-low state, and an ultra-low signal-to-noise ratio trigger flag is generated.

[0234] The ultra-low trigger flag serves as the activation signal for initiating magnetic field strength control operations. The trigger confirmation duration is used to filter out transient drops in the signal-to-noise ratio, avoiding unnecessary control operations due to brief interference.

[0235] Through the aforementioned low-threshold cross-pass inspection operation, time intervals in the signal-to-noise ratio (SNR) filtered sequence that are continuously in an ultra-low state are identified, and corresponding ultra-low SNR trigger markers are generated.

[0236] The reinforcement deployment module 106 drives the excitation current of the permanent magnet of the relay node in a step-by-step manner based on the ultra-low signal-to-noise ratio trigger flag of the successfully anchored relay node.

[0237] An excitation coil is wound on the permanent magnet of the relay node. The coil is connected to a controllable current source built into the relay node. The output current of the controllable current source is controlled by the current setting value sent by the reinforcement deployment module 106.

[0238] Upon receiving the low signal-to-noise ratio trigger flag, the reinforcement deployment module 106 uses the current excitation current value as a reference and gradually increases the current setting value according to the preset step increment, preferably 5% of the current current value.

[0239] After each current boost, wait for a preset stabilization time, preferably 500 milliseconds, to allow the permanent magnet's magnetic field strength to be fully established and stabilized.

[0240] With each step increase in current, the magnetic field strength of the permanent magnet increases accordingly, the attraction between the permanent magnet and the tank wall increases, the tightness of the connection between the relay node and the tank wall is improved, the electrical conductivity of the contact interface is improved, and the signal-to-noise ratio of the communication link recovers.

[0241] At the end of the stable duration after each incremental step, the reinforcement deployment module 106 re-acquires the signal-to-noise ratio measurement value at the current moment and compares the recovered signal-to-noise ratio value with the preset hysteresis threshold.

[0242] The hysteresis threshold is higher than the low quality threshold. The reason for setting it is to prevent the signal-to-noise ratio from repeatedly crossing the threshold, which would cause frequent switching of the control state. The preferred hysteresis threshold is 15 dB.

[0243] If the signal-to-noise ratio after recovery reaches or exceeds the hysteresis threshold, it is determined that the signal-to-noise ratio has recovered to the standard state, and the reinforcement deployment module 106 maintains the current excitation current value unchanged and exits the control process.

[0244] If the signal-to-noise ratio after the rebound is still lower than the hysteresis threshold, the next step of incremental drive will continue until the signal-to-noise ratio reaches the target or the duration of the control exceeds the preset window.

[0245] Through the above step-by-step incremental drive operation, the excitation current of the permanent magnet of the relay node is increased step by step, the magnetic field strength of the permanent magnet is gradually increased, and the signal-to-noise ratio recovery is tracked and determined in real time after each step.

[0246] The reinforcement deployment module 106 determines the reinforcement requirements based on the increase in signal-to-noise ratio and the duration of regulation after the magnetic field strength of the permanent magnet of the relay node is raised.

[0247] The duration of the control is counted from the moment the signal-to-noise ratio ultra-low trigger flag is first detected until the current moment.

[0248] The preset timeout duration is determined by those skilled in the art based on the system's response speed requirements for link recovery and the physical response time of the permanent magnet's magnetic field strength adjustment, and is preferably 30 seconds.

[0249] If the signal-to-noise ratio rises above the hysteresis threshold within the duration of regulation, it is determined that no reinforcement is needed, a regulation exit state is generated, the current excitation current is maintained, and the regulation process is exited.

[0250] If the signal-to-noise ratio fails to recover to above the hysteresis threshold after the control duration reaches the preset timeout period, it is determined that the oil medium loss in the area where the current relay node is located has exceeded the range that can be compensated by magnetic field strength control alone, and a reinforcement relay needs to be deployed.

[0251] At this point, a reinforcement deployment command is generated, which includes the spatial coordinates of the relay node that needs to be reinforced and the spatial coordinates of the robot's current location.

[0252] The reinforcement deployment module 106 performs a backtracking search on the coordinate points recorded in the robot's historical path according to the reinforcement deployment instruction. It uses the spatial coordinates of the relay node that needs reinforcement as the search target and finds the path point closest to the relay node in the robot's historical path coordinate sequence.

[0253] Use this waypoint as the target waypoint for the robot to return to the relay node, and plan the return path from the robot's current position to this target waypoint.

[0254] The return path is planned by tracing back along the robot's historical path, that is, by passing through each historical path point in the reverse order in which the robot entered the current area, ensuring that the robot returns along a known safe path.

[0255] The reinforcement deployment module 106 sends the return path to the robot's motion control system, driving the robot to move along the return path to the neighboring area of ​​the relay node.

[0256] When the robot arrives in the neighborhood of the relay node, it triggers the release and anchoring process of the reinforcement relay, and completes the deployment and anchoring of the reinforcement relay according to the relay node anchoring method according to any one of claims 1 to 9.

[0257] Through the complete embodiment of the above-mentioned reinforcement deployment module 106, the signal-to-noise ratio of the successfully anchored relay node is continuously detected and filtered in real time. When the signal-to-noise ratio is continuously too low, the magnetic field strength of the permanent magnet is actively adjusted by stepping up the excitation current to attempt physical layer self-healing. Only after the adjustment timeout is ineffective will the reinforcement relay be deployed.

[0258] This two-level response mechanism provides a physical layer self-healing method with zero additional hardware cost before triggering reinforcement deployment, which can avoid unnecessary reinforcement deployment; the magnetic field strength regulation shares the same physical field information framework with the aforementioned magnetic field distortion ranging and isomagnetic field guidance, making the relay node permanent magnet a multifunctional core device integrating sensing, ranging, navigation and self-healing, thereby realizing the signal quality self-maintenance of the communication link of the wall-climbing robot in the storage tank.

[0259] Reference Figure 2 The diagram shown is a flowchart illustrating a method for deploying a communication relay for a wall-climbing robot inside a petrochemical storage tank based on permanent magnet adsorption, according to an embodiment of the present invention. In this embodiment, the method includes: S1. Based on the signal fading and multi-dimensional quality indicators of each relay node during the robot's movement process, the anchoring and bonding calculation is performed on each relay node to obtain the anchoring success signal of each relay node. S2. Perform mutual interference separation on the magnetic field distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic spacing of the relay node. S3. Using the signal fading amount of the successfully anchored relay node measured by the robot as a reference, the natural magnetic spacing is scaled to obtain the legal execution spacing between adjacent relay nodes. S4. Compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position. S5. Based on the adsorption force difference and the statutory execution interval, perform isomagnetic line guidance on the robot path to obtain the relay deployment position of the robot and anchor the relay node; S6. Detect the signal-to-noise ratio of the relay node that has been successfully anchored in real time. If the signal-to-noise ratio is too low, adjust the magnetic field strength of the permanent magnet until it recovers. If the adjustment does not recover after a preset time, deploy the robot to return to the relay node and deploy a reinforcement relay.

[0260] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0261] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0262] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A communication relay deployment system and method for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption, characterized in that, The system includes an anchoring calculation module, a magnetic distance calculation module, a spacing conversion module, a difference comparison module, an isomagnetic guidance module, and a reinforcement deployment module, wherein: The anchoring calculation module is used to perform anchoring and bonding calculation on each relay node based on the fading amount and multi-dimensional quality indicators of the back-transmitted signals of each relay node during the robot's movement process, and to obtain the anchoring success signal of each relay node. The magnetic distance calculation module is used to separate the mutual interference of the magnetic field line distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic distance of the relay node. The spacing conversion module is used to use the signal fading amount of the successfully anchored relay node measured by the robot as a reference to perform scale conversion on the natural magnetic spacing to obtain the legal execution spacing between adjacent relay nodes. The difference comparison module is used to compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position. The isomagnetic guidance module is used to guide the robot path with isomagnetic lines based on the adsorption force difference and the statutory execution interval, so as to obtain the relay deployment position of the robot and anchor the relay node. The reinforcement deployment module is used to detect the signal-to-noise ratio of the successfully anchored relay node in real time. If the signal-to-noise ratio is too low, the magnetic field strength of the permanent magnet is adjusted until it recovers. If the adjustment does not recover after a preset time, the deployment robot returns to the relay node to deploy the reinforcement relay.

2. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 1, characterized in that, When the anchoring calculation module performs anchoring and bonding calculations on each relay node based on the fading amount and multi-dimensional quality indicators of the signals transmitted back from each relay node during the robot's movement process, and obtains the successful anchoring signaling for each relay node, it is specifically used for: The attenuation gradient sequence of the attenuation of the back-transmission signal is obtained by performing time-domain difference extraction on the attenuation amount of the back-transmission signal. Principal quantification is performed on the multidimensional quality indicators to obtain the characterization sequence of the multidimensional quality indicators. Based on the decay gradient sequence and the characterization sequence, the fitting state of each relay node is determined step by step to obtain the fitting confidence sequence of each relay node. The anchoring success signaling of each relay node is obtained by performing time-series cross-verification on the fitting confidence sequence of each relay node.

3. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 1, characterized in that, When the magnetic distance estimation module performs mutual interference separation on the magnetic field line distortion vector of the successfully anchored relay node permanent magnet to obtain the natural magnetic distance of the relay node, it is specifically used for: Three-component sensing is performed on the space surrounding the successfully anchored permanent magnet of the relay node to obtain the spatial magnetic field vector sequence of the permanent magnet of the relay node; Background magnetic field cancellation is performed on the spatial magnetic field vector sequence to obtain the pure magnetic vector field of the relay node permanent magnet; Curvature analysis is performed on the pure magnetic vector field to obtain the magnetic field distortion vector field of the relay node permanent magnet; The source of the magnetic field distortion vector field is separated, and the natural magnetic spacing of the relay node permanent magnet is obtained by gradient inversion based on the separated distortion vector.

4. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 3, characterized in that, When the magnetic moment estimation module performs field source separation on the magnetic field distortion vector field and performs gradient inversion based on the separated distortion vector to obtain the natural magnetic spacing of the relay node permanent magnet, it is specifically used for: The magnetic field line distortion vector field is expanded using a multipole orthogonal basis to obtain the independent components of the contribution of each permanent magnet in the magnetic field line distortion vector field. The independent components of the contribution of each permanent magnet are inverted to obtain the field source spatial coordinates of the independent components of the contribution of each permanent magnet. The distance between adjacent permanent magnets is calculated based on the spatial coordinates of the field source to obtain the natural magnetic distance between the permanent magnets of the relay node.

5. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 4, characterized in that, The spacing conversion module, when using the signal fading amount of the successfully anchored relay node as a reference, performs a scale conversion on the natural magnetic spacing to obtain the legal execution spacing between adjacent relay nodes, specifically for: Large-scale fading components are extracted from the signal fading of the successfully anchored relay node to obtain the slowly varying fading envelope of the successfully anchored relay node. By fitting the distance loss relationship to the slowly varying fading envelope, the path loss index of the successfully anchored relay node is obtained. The path loss exponent is used to weight and scale the natural magnetic spacing to obtain the corrected spacing of the successfully anchored relay nodes. The communication constraint boundary is normalized on the modified spacing to obtain the statutory execution spacing of the successfully anchored relay node.

6. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 1, characterized in that, When the difference comparison module performs a difference comparison between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position, it is specifically used for: The adsorption force at each spatial point along the robot's movement path is sensed in real time to obtain the dynamic adsorption force amplitude at each spatial point of the robot. By stripping the trend term from the dynamic adsorption force amplitude at each spatial point of the robot, the steady-state adsorption force amplitude at each spatial point of the robot is obtained. The pre-adsorption force of the permanent magnet of the relay node is synchronously tracked and locked to obtain the pre-adsorption force amplitude of the permanent magnet of the relay node; The difference between the steady-state adsorption force amplitude and the preparatory force amplitude is obtained by performing on-site registration at the same point, thus obtaining the adsorption force difference at the current position of the robot.

7. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 5, characterized in that, When the isomagnetic guidance module performs isomagnetic field line guidance on the robot path based on the adsorption force difference and the statutory execution interval, and obtains the robot's relay deployment position and anchors the relay node, it is specifically used for: The spatial gradient direction of the adsorption force difference is calculated to obtain the tangential direction of the isomagnetic field lines at the current position of the robot; Based on the tangential direction of the isomagnetic field lines and the statutory execution distance, the path of the robot is recursively calculated to obtain the isomagnetic field line path from the current position of the robot to the target deployment position; The robot's real-time position on the path of the equal magnetic field lines is tracked tangentially by magnetic field lines to obtain the robot's magnetic field line tracking yaw angle. Based on the yaw angle of the magnetic field line tracking, the robot is driven to move along the path of the equal magnetic field line to the spatial point corresponding to the statutory execution interval, and the spatial point is used as the relay deployment position of the robot and the relay node anchoring is triggered.

8. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 7, characterized in that, When the isomagnetic guidance module performs path recursion on the robot based on the tangential direction of the isomagnetic field lines and the statutory execution interval to obtain the isomagnetic field line path from the robot's current position to the target deployment position, it is specifically used for: By projecting the statutory execution interval into the tangential direction of the magnetic field lines, the recursive step size of the statutory execution interval in the direction of the magnetic field lines is obtained; Based on the recursive step size, the single-step displacement of the robot's current position is calculated to obtain the robot's single-step recursive position; The tangential direction of the magnetic field lines at the single-step recursion position is updated to obtain the new tangential direction at the single-step recursion position; Based on the new tangential direction, the recursive step size is iterated step by step until the displacement reaches the statutory execution distance, thus obtaining the isomagnetic field path from the robot's current position to the target deployment position.

9. The communication relay deployment system for a petrochemical storage tank wall-climbing robot based on permanent magnet adsorption as described in claim 1, characterized in that, The reinforcement deployment module detects the signal-to-noise ratio (SNR) of successfully anchored relay nodes in real time. If the SNR is extremely low, it adjusts the magnetic field strength of the permanent magnet until it recovers. If the SNR does not recover after a preset time, the deployment robot returns to the relay node to deploy the reinforcement relay. Specifically, it is used for: The signal-to-noise ratio of the successfully anchored relay node is subjected to sliding window mean filtering to obtain the signal-to-noise ratio filtering sequence of the successfully anchored relay node. A low-threshold cross-pass inspection is performed on the signal-to-noise ratio filtered sequence of the successfully anchored relay node to obtain the ultra-low signal-to-noise ratio trigger flag of the successfully anchored relay node. Based on the ultra-low signal-to-noise ratio trigger flag of the successfully anchored relay node, the excitation current of the permanent magnet of the relay node is driven by step increment to obtain the magnetic field strength rise state of the permanent magnet of the relay node. The signal-to-noise ratio increase after the magnetic field strength of the permanent magnet of the relay node is raised and the duration of the regulation are used to determine the reinforcement requirement, so as to obtain the regulation exit state or the reinforcement deployment command. Based on the reinforcement deployment command, the robot is driven to return to the neighborhood of the relay node to deploy the reinforcement relay.

10. A method for deploying communication relays for a wall-climbing robot inside a petrochemical storage tank based on permanent magnet adsorption, characterized in that, The method for deploying a communication relay system for a climbing robot inside a petrochemical storage tank based on permanent magnet adsorption as described in claim 1, wherein the method is as follows: S1. Based on the signal fading and multi-dimensional quality indicators of each relay node during the robot's movement process, the anchoring and bonding calculation is performed on each relay node to obtain the anchoring success signal of each relay node. S2. Perform mutual interference separation on the magnetic field distortion vector of the anchored relay node permanent magnet to obtain the natural magnetic spacing of the relay node. S3. Using the signal fading amount of the successfully anchored relay node measured by the robot as a reference, the natural magnetic spacing is scaled to obtain the legal execution spacing between adjacent relay nodes. S4. Compare the difference between the adsorption force at the current position of the robot and the preparatory force of the relay node to obtain the adsorption force difference at the current position. S5. Based on the adsorption force difference and the statutory execution interval, perform isomagnetic line guidance on the robot path to obtain the relay deployment position of the robot and anchor the relay node; S6. Detect the signal-to-noise ratio of the relay node that has been successfully anchored in real time. If the signal-to-noise ratio is too low, adjust the magnetic field strength of the permanent magnet until it recovers. If the adjustment does not recover after a preset time, deploy the robot to return to the relay node and deploy a reinforcement relay.

Images