5G base station structure abnormity early warning system

Abstract

The invention discloses a 5G base station structure abnormity early warning system, which relates to the technical field of abnormity early warning and comprises a model construction unit, an amplitude analysis unit, a node preliminary screening unit, a section division unit, a scheme generation unit and a damage simulation unit. According to the method, the communication tower model is established according to an actual structure by using ANSYS software, modal parameters are obtained through a PCG Lanczos method, the inherent frequency of the communication tower in the 5G base station can be accurately calculated through the design, fatigue damage or instantaneous damage caused by resonance can be fundamentally prevented, and by observing node amplitude sets of different orders, the reliability of the communication tower in the 5G base station can be improved. The distribution rationality of structural rigidity and quality of the iron tower can be visually judged, weak links in the design can be identified, so that targeted reinforcement is carried out, a hybrid screening strategy is adopted to search for the optimal distribution point of the sensors, and it is ensured that measuring points can collect signals which can best reflect the overall dynamic characteristics; and data support is provided for subsequent model correction and damage identification.

Description

Technical Field

[0001] This invention relates to the field of anomaly early warning technology, specifically a 5G base station structural anomaly early warning system. Background Technology

[0002] A 5G base station structural anomaly early warning system interacts with a sensor network and a digital model to monitor the tower structure status in real time, intelligently identify damage and anomalies, and achieve accurate early warning and predictive maintenance. Patent application number 202310722225.7 discloses "a 5G base station operation status monitoring system, including a work order generation module, an employee terminal, a drone, a remote controller, and a data analysis module. The work order generation module generates maintenance work orders for monitoring the status of the 5G base station's outer shell and sends these work orders to the employee terminal. The maintenance work order includes the maintenance personnel's personal information and the location of the 5G base station. The employee terminal generates a navigation route based on the location of the 5G base station and the maintenance personnel, guiding the maintenance personnel to the 5G base station. The remote controller receives control commands input by the maintenance personnel and sends these commands to the drone. The drone takes pictures of the 5G base station's outer shell according to the control commands, obtaining an image of the outer shell. The data analysis module performs image recognition on the outer shell image to obtain the monitoring results of the 5G base station's outer shell that requires status monitoring. This invention improves the efficiency of monitoring the outer shell of 5G base stations."

[0003] The aforementioned existing technologies have solved problems such as the inability to monitor the outer shell of 5G base stations in real time. However, during system operation, due to the inability to accurately predict resonance risks, the tower may experience premature fatigue or instantaneous damage due to resonance under wind loads and other excitations. At the same time, the arbitrary placement of sensors makes it impossible to effectively capture key dynamic signals, resulting in distortion or even failure of subsequent health monitoring. Operation and maintenance management has regressed from predictive maintenance to post-event remediation, significantly increasing the total life cycle cost and safety risks, and failing to guarantee the long-term reliable operation of 5G base stations. Summary of the Invention

[0004] The purpose of this invention is to provide a 5G base station structural anomaly early warning system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a 5G base station structural anomaly early warning system, comprising a scheme generation unit and a damage simulation unit; The model building unit is constructed using ANSYS finite element analysis software based on the actual 5G communication tower structure and parameters. A scaled-down three-tube tower model; An amplitude analysis unit calculates the first, second, and third matrices of the three-tube tower model, analyzes the model's attribute characterization formulas according to the first, second, and third matrices, calculates the modal parameters, and constructs performance index evaluation formulas based on the modal parameters. The node preliminary screening unit obtains the number and position coordinates of all nodes in the three-tube tower model, analyzes the total kinetic energy of the nodes, filters the nodes according to the total kinetic energy, calculates the independence matrix based on the filtered nodes, removes redundant nodes using the independence matrix, and outputs the node numbers to be retained. The segmentation unit uses a random sampling strategy to generate an initial set of schemes, calculates the evaluation value of all schemes in the set, and divides all schemes into four segments according to the evaluation value, namely the first segment, the second segment, the third segment, and the fourth segment.

[0006] Preferably, the amplitude analysis unit includes a matrix calculation module and a position transformation combined module; The matrix calculation module divides the structure of the three-tube tower model into a finite number of independent units. , Indicates the number of units. Indicates the first The first independent unit, read the first... Individual units volume Elasticity matrix Strain-displacement matrix Material density And shape function matrix Afterwards, according to Calculate the first matrix ,in , Representing volume, for and Analysis yields the second matrix. ,in Repeat the operation until all individual units have been read; The position transformation combined with the module statistical independent unit The first matrix and the second matrix Then, determine the position transformation matrix for each independent unit. , Indicates the first The position transformation matrix of each independent unit, using the first matrix Second matrix and position transformation matrix The first matrix of the three-tube tower model was calculated. Second matrix ,in , , Indicates the first Position transformation matrix of each independent unit Indicates the unit number.

[0007] Preferably, the amplitude analysis unit further includes a force analysis module, a property characterization calculation module, and an amplitude set determination module; The force analysis module obtains the damping coefficient. and Then, using Calculate the third matrix ,in According to the first matrix of the three-tube tower model Second matrix and the third matrix The external force vectors of each node in the model were calculated. ,in , This represents the acceleration vector of a node in the model. This represents the velocity vector of a node in the model. Represents the displacement vector of a node in the model; The attribute representation calculation module when and When this happens, the displacement vector of the node is determined. ,in , Represents the imaginary unit. Indicates the vibration frequency. Indicates time, Represents the set of amplitudes at all nodes. Represents the natural constant, based on the displacement vector. Calculate the corresponding acceleration vector ,in ,use , as well as A property characterization formula for a three-tube tower was constructed. The amplitude set determination module calculates the modal parameters using the PCG Lanczos method according to the characterization formula of a three-tube tower. Based on the amplitude set order of all nodes within the first 20 orders of the modal parameters, a performance index evaluation formula is constructed. The specific performance index evaluation formula is as follows: in, Indicates the performance evaluation value. Indicates the total order. Indicates the first The set of all nodal amplitudes of order 1 Indicates the first The set of all nodal amplitudes of order 1 Indicates a parameter.

[0008] Preferably, the node preliminary screening unit includes a node removal module, a total kinetic energy calculation module, an independent matrix generation module, and a node number output module; The node removal module obtains the number and position coordinates of all nodes in the three-tube tower model, sets the retention number, removes invalid nodes, symmetrical nodes and dense nodes until the number of remaining nodes reaches the retention number, and stores the unremoved nodes in the initial candidate set. The total kinetic energy calculation module reads the amplitude sets of all nodes corresponding to the first 20 orders, and after obtaining the displacement vectors of all nodes in the initial candidate set under each order mode based on the amplitude sets of all nodes, it randomly selects one node. , Represents the node index, and is used for statistics. Displacement vectors at all orders ,in , This represents the position vector at the 20th order. Representing the three translational displacement components at the 20th order, according to Calculate the nodes Total kinetic energy ,in Repeat the operation to calculate the total kinetic energy of all nodes in the initial candidate set; The independent matrix generation module sorts all nodes according to their total kinetic energy from largest to smallest, retains the top 50% of nodes, and removes the remaining nodes. Based on the retained nodes, it constructs the amplitude set corresponding to the first 20 orders. Using amplitude sets Calculate the independent matrix ,in ; The node number output module selects the smallest diagonal element in the independent matrix, removes the node corresponding to that value, constructs a new amplitude set, recalculates the independent matrix based on the new amplitude set, and repeats the operation until the number of nodes in the candidate set equals the threshold, and then outputs all node numbers.

[0009] Preferably, the segmentation unit includes a scheme set generation module, an evaluation value output module, and a sequence division module; The scheme set generation module uses a random sampling strategy to generate an initial scheme set. ,in Indicates the first One option, , This indicates the total number of schemes. Indicates the first Nodes The deployment situation on the above, if This indicates that the node is equipped with sensors. This indicates that no sensors are installed at that node. Indicates the number of nodes in the candidate set; The evaluation value output module randomly selects a scheme. , Indicates the scheme number, according to The sensor positions are determined by the arrangement of each node. The displacement vectors of the corresponding nodes at each order are extracted from the set of all node amplitudes for each order, and these vectors are combined to obtain the set of node amplitudes at different orders. , Representation scheme The Middle The set of nodal amplitudes of order, according to Calculate Evaluation value ,in , Indicates the total order. Representation scheme The Middle The set of nodal amplitudes of order, Representation scheme The Middle The set of nodal amplitudes of order, The parameters are represented, and the operation is repeated until the evaluation values ​​of all solutions are calculated. Then, the current number of executions and the maximum number of executions are determined. The sequence partitioning module ranks all schemes according to their evaluation values. After sorting, the current sequence is divided into four segments according to preset ratio parameters: the first segment, the second segment, the third segment, and the fourth segment.

[0010] Preferably, the scheme generation unit includes a first segment analysis module and a second segment analysis module; The first segment analysis module extracts all schemes in the first segment. , Indicates the number of schemes. Indicates the first segment One option, determine the historical best option. Then, select a solution. ,Will and Compare and statistically analyze Sensors are deployed but There are no all the nodes where the sensors are deployed; instead, a node is randomly selected, and... The deployment status of this node was changed from 1 to 0, and statistics were also compiled. There is no sensor deployment but Given all the nodes where sensors are deployed, randomly select one node to perform a test. After changing the node's deployment status from 0 to 1, in the current scheme Two nodes are randomly selected from all nodes, with their configurations set to 1 and 0 respectively. These nodes are then modified to 0 and 1 respectively. The current... The evaluation value, if the evaluation value is less than the previous one The evaluation value will then be the current Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the first segment are selected. The second segment analysis module extracts all schemes in the second segment. , Indicates the number of schemes. Indicates the second segment Choose one of the following options. According to the plan right After adjusting the deployment of all nodes in the current scheme, Four nodes are randomly selected from all nodes, with their initial configurations being 1, 1, 0, and 0 respectively. These nodes are then modified to their corresponding configurations of 0, 0, 1, and 1. The current... The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the second segment are selected.

[0011] Preferably, the scheme generation unit further includes a third segment analysis module, a fourth segment analysis module, and an optimal scheme output module; The third segment analysis module extracts all schemes in the third segment. back, Indicates the number of schemes. Indicates the third segment Choose one of the following options. A new scheme is generated directly using a random sampling strategy, and adjustments are made according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the third segment are selected. The fourth segment analysis module extracts all schemes in the fourth segment. Then, select a solution. Two schemes are randomly selected from the first segment. A portion of each scheme is taken and spliced ​​together to generate a new scheme. The scheme is then adjusted according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the fourth segment are selected. The optimal solution output module selects the solution with the largest evaluation value and compares it with the historical optimal solution. The optimal solution is retained in the historical optimal solution. The current execution count is automatically incremented by one. The module then checks whether the current execution count is equal to the maximum execution count. If it is equal, the current historical optimal solution is directly output. Otherwise, the module enters the next round of the loop.

[0012] Preferably, the damage simulation unit includes a baseline mode calculation module, an early warning module, and a state analysis module; The reference mode calculation module deploys acceleration sensors at each node of the actual communication tower according to the current historical best scheme. When the communication tower is in normal condition, it uses the deployed sensors to collect acceleration data in each time period, calculates the modal parameters in each time period under this condition, averages and summarizes them to obtain the reference mode parameters corresponding to different time periods. The early warning module sets early warning thresholds and alarm thresholds. After updating the three-tube tower model according to the reference modal parameters, a reference model in normal condition is obtained. Real-time acceleration data is collected, and the current modal parameters are calculated based on the real-time acceleration data. The relative rate of change between the current modal parameters and the reference modal parameters is determined. If the relative rate of change exceeds the early warning threshold, the parameter is marked and an early warning message is generated. If any of the following limiting conditions are met, an alarm is issued immediately. The specific limiting conditions are: S101, Key parameters significantly exceed the standard: The rate of change of a single modal parameter continuously and significantly exceeds the alarm threshold; S102, Multi-parameter coordinated anomaly: The rate of change of three or more modal parameters simultaneously exceeds the warning threshold; S103. The amplitude set of all nodes at any order is severely degraded: the MAC value of the amplitude set drops below the alarm threshold; After receiving the alarm, the status analysis module sets the damage parameters in the ANSYS model as design variables. With the goal of minimizing the error between the modal parameters calculated by the model and the current measured modal parameters, it drives multiple simulation iterations through optimization algorithms to continuously adjust the damage location and severity in the model until the error minimization target is achieved. Then, it outputs the damage identified in the model as the best estimate of the actual damage location and severity in the real structure, and analyzes the current health status of the communication tower based on the severity of the damage.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention utilizes ANSYS software to establish a communication tower model based on the actual structure and obtains modal parameters through the PCG Lanczos method. This design can accurately calculate the natural frequency of the communication tower in the 5G base station, fundamentally preventing fatigue damage or instantaneous failure caused by resonance. By observing the set of nodal amplitudes of different orders, the rationality of the distribution of the tower's structural stiffness and mass can be intuitively judged, and weak links in the design can be identified for targeted reinforcement. This invention employs a hybrid screening strategy to find the optimal sensor placement points, ensuring that the measuring points can collect signals that best reflect the overall dynamic characteristics, providing data support for subsequent model correction and damage identification. Through the interaction between the virtual model and the physical monitoring data, the structural health status can be assessed in real time, facilitating timely intervention by maintenance personnel. Attached Figure Description

[0014] Figure 1 A schematic diagram of the overall system flow is provided for embodiments of the present invention; Figure 2 This is an internal module block diagram of the amplitude analysis unit provided in an embodiment of the present invention; Figure 3 This is an internal module block diagram of the node preliminary screening unit provided in an embodiment of the present invention; Figure 4 This is an internal module block diagram of the solution generation unit provided in the embodiments of the present invention; Figure 5 This is an internal module block diagram of the damage simulation unit provided in an embodiment of the present invention.

[0015] In the diagram: 1. Model building unit; 2. Amplitude analysis unit; 201. Matrix calculation module; 202. Position transformation and combination module; 203. Force analysis module; 204. Attribute characterization calculation module; 205. Amplitude set determination module; 3. Preliminary node screening unit; 301. Node removal module; 302. Total kinetic energy calculation module; 303. Independent matrix generation module; 304. Node number output module; 4. Segment division unit; 401. Scheme set generation module; 402. Evaluation value output module; 403. Sequence division module; 5. Scheme generation unit; 501. First segment analysis module; 502. Second segment analysis module; 503. Third segment analysis module; 504. Fourth segment analysis module; 505. Optimal scheme output module; 6. Damage simulation unit; 601. Reference mode calculation module; 602. Early warning module; 603. State analysis module. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1 - Figure 5 The present invention provides a technical solution: a 5G base station structural anomaly early warning system, including a solution generation unit 5 and a damage simulation unit 6; Model building unit 1 was constructed using ANSYS finite element analysis software based on the actual 5G communication tower structure and parameters. A scaled-down three-tube tower model; Amplitude analysis unit 2 calculates the first, second, and third matrices of the three-tube tower model, analyzes the attribute characterization formula of the model according to the first, second, and third matrices, calculates the modal parameters, and constructs the performance index evaluation formula based on the modal parameters; Node preliminary screening unit 3 obtains the number and position coordinates of all nodes in the three-tube tower model, analyzes the total kinetic energy of the nodes, filters the nodes according to the total kinetic energy, calculates the independence matrix based on the filtered nodes, removes redundant nodes using the independence matrix, and outputs the number of the retained nodes. Segment division unit 4 uses a random sampling strategy to generate an initial set of schemes, calculates the evaluation value of all schemes in the set, and divides all schemes into four segments according to the evaluation value, namely the first segment, the second segment, the third segment and the fourth segment.

[0018] Amplitude analysis unit 2 includes a matrix calculation module 201 and a position transformation combination module 202; Matrix calculation module 201 divides the structure of the three-tube tower model into a finite number of independent units. , Indicates the number of units. Indicates the first The first independent unit, read the first... Individual units volume Elasticity matrix Strain-displacement matrix Material density And shape function matrix Afterwards, according to Calculate the first matrix ,in , Representing volume, for and Analysis yields the second matrix. ,in Repeat the operation until all individual units have been read; Position transformation combined module 202 statistical independent unit The first matrix and the second matrix Then, determine the position transformation matrix for each independent unit. , Indicates the first The position transformation matrix of each independent unit, using the first matrix Second matrix and position transformation matrix The first matrix of the three-tube tower model was calculated. Second matrix ,in , , Indicates the first Position transformation matrix of each independent unit Indicates the unit number; Amplitude analysis unit 2 also includes a force analysis module 203, an attribute characterization calculation module 204, and an amplitude set determination module 205; Force analysis module 203 obtains damping coefficient and Then, using Calculate the third matrix ,in According to the first matrix of the three-tube tower model Second matrix and the third matrix The external force vectors of each node in the model were calculated. ,in , This represents the acceleration vector of a node in the model. This represents the velocity vector of a node in the model. Represents the displacement vector of a node in the model; Attribute representation calculation module 204 and When this happens, the displacement vector of the node is determined. ,in , Represents the imaginary unit. Indicates the vibration frequency. Indicates time, Represents the set of amplitudes at all nodes. Represents the natural constant, based on the displacement vector. Calculate the corresponding acceleration vector ,in ,use , as well as The property characterization formula for the three-tube tower is constructed as follows: in, Describes the first matrix. Represents the second matrix, Represents the imaginary unit. Indicates the vibration frequency. Indicates time, Represents the set of amplitudes at all nodes. Represents the natural constant; The amplitude set determination module 205 calculates the modal parameters using the PCG Lanczos method according to the characterization formula of a three-tube tower. Based on the amplitude set order of all nodes within the first 20 orders of the modal parameters, a performance index evaluation formula is constructed. The specific performance index evaluation formula is as follows: in, Indicates the performance evaluation value. Indicates the total order. Indicates the first The set of all nodal amplitudes of order 1 Indicates the first The set of all nodal amplitudes of order 1 Indicates parameters; The node preliminary screening unit 3 includes a node removal module 301, a total kinetic energy calculation module 302, an independent matrix generation module 303, and a node number output module 304; The node removal module 301 obtains the number and position coordinates of all nodes in the three-tube tower model. After setting the number to be retained, it removes invalid nodes, symmetrical nodes, and dense nodes. Invalid nodes are specifically those located inside the structure, those obscured by other components, and those where sensors cannot be installed in practice. Symmetrical nodes are specifically those that can be obtained through symmetry operations and correspond to the nodes that have not been removed. Dense nodes are specifically those whose distance from the nodes that have not been removed is less than a preset value. This process continues until the number of remaining nodes reaches the number to be retained. The nodes that have not been removed are then stored in the initial candidate set. The total kinetic energy calculation module 302 reads the amplitude sets of all nodes corresponding to the first 20 orders, and after obtaining the displacement vectors of all nodes in the initial candidate set under each order mode based on the amplitude sets of all nodes, it randomly selects one node. , Represents the node index, and is used for statistics. Displacement vectors at all orders ,in , This represents the position vector at the 20th order. Representing the three translational displacement components at the 20th order, according to Calculate the nodes Total kinetic energy ,in Repeat the operation to calculate the total kinetic energy of all nodes in the initial candidate set; The independent matrix generation module 303 sorts all nodes in descending order of total kinetic energy, retains the top 50% of nodes, and removes the remaining nodes. Based on the retained nodes, it constructs the amplitude set corresponding to the first 20 orders. Using amplitude sets Calculate the independent matrix ,in ; The node number output module 304 selects the smallest value of the diagonal element in the independent matrix, removes the node corresponding to the value, constructs a new amplitude set, recalculates the independent matrix based on the new amplitude set, repeats the operation until the number of nodes in the candidate set equals the threshold, and then outputs all node numbers. The segment division unit 4 includes a scheme set generation module 401, an evaluation value output module 402, and a sequence division module 403; The scheme set generation module 401 uses a random sampling strategy to generate an initial scheme set. ,in Indicates the first One option, , This indicates the total number of schemes. Indicates the first Nodes The deployment situation on the above, if This indicates that the node is equipped with sensors. This indicates that no sensors are installed at that node. Indicates the number of nodes in the candidate set; Evaluation value output module 402 randomly selects a scheme , Indicates the scheme number, according to The sensor positions are determined by the arrangement of each node. The displacement vectors of the corresponding nodes at each order are extracted from the set of all node amplitudes for each order, and these vectors are combined to obtain the set of node amplitudes at different orders. , Representation scheme The Middle The set of nodal amplitudes of order, according to Calculate Evaluation value ,in , Indicates the total order. Representation scheme The Middle The set of nodal amplitudes of order, Representation scheme The Middle The set of nodal amplitudes of order, The parameters are represented, and the operation is repeated until the evaluation values ​​of all solutions are calculated. Then, the current number of executions and the maximum number of executions are determined. Sequence partitioning module 403 partitions all schemes according to the evaluation value. After sorting, the current sequence is divided into four segments according to a preset ratio parameter: the first segment, the second segment, the third segment, and the fourth segment. The ratio parameter for the first segment is... , , Indicates the maximum proportion of the first segment. Indicates the current execution count. Indicates the maximum number of executions. The constant representing the rate of decay is used to control the rate of decay. The scaling parameter for the second segment is... , , This indicates the minimum proportion of the second segment. This indicates the maximum proportion of the second segment, and the proportion parameter for the third segment is... , , This indicates the maximum proportion of the third segment, and the proportion parameter for the fourth segment is... , ; The scheme generation unit 5 includes a first segment analysis module 501 and a second segment analysis module 502; The first segment analysis module 501 extracts all schemes in the first segment. , Indicates the number of schemes. Indicates the first segment One option, determine the historical best option. Then, select a solution. ,Will and Compare and statistically analyze Sensors are deployed but There are no all the nodes where the sensors are deployed; instead, a node is randomly selected, and... The deployment status of this node was changed from 1 to 0, and statistics were also compiled. There is no sensor deployment but Given all the nodes where sensors are deployed, randomly select one node to perform a test. After changing the node's deployment status from 0 to 1, in the current scheme Two nodes are randomly selected from all nodes, with their configurations set to 1 and 0 respectively. These nodes are then modified to 0 and 1 respectively. The current... The evaluation value, if the evaluation value is less than the previous one The evaluation value will then be the current Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the first segment are selected. The second segment analysis module 502 extracts all schemes in the second segment. , Indicates the number of schemes. Indicates the second segment Choose one of the following options. According to the plan right After adjusting the deployment of all nodes in the current scheme, Four nodes are randomly selected from all nodes, with their initial configurations being 1, 1, 0, and 0 respectively. These nodes are then modified to their corresponding configurations of 0, 0, 1, and 1. The current... The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the second segment are selected. The scheme generation unit 5 also includes a third segment analysis module 503, a fourth segment analysis module 504, and an optimal scheme output module 505; The third segment analysis module 503 extracts all schemes in the third segment. back, Indicates the number of schemes. Indicates the third segment Choose one of the following options. A new scheme is generated directly using a random sampling strategy, and adjustments are made according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the third segment are selected. The fourth segment analysis module 504 extracts all schemes in the fourth segment. Then, select a solution. Two schemes are randomly selected from the first segment. A portion of each scheme is taken and spliced ​​together to generate a new scheme. The scheme is then adjusted according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the fourth segment are selected. The optimal solution output module 505 selects the solution with the largest evaluation value and compares it with the historical optimal solution. The optimal solution is retained in the historical optimal solution. The current execution count is automatically incremented by one. The module then checks whether the current execution count is equal to the maximum execution count. If it is equal, the current historical optimal solution is directly output. Otherwise, the module enters the next loop. Damage simulation unit 6 includes a baseline mode calculation module 601, an early warning module 602, and a state analysis module 603; The reference mode calculation module 601 deploys acceleration sensors at each node of the actual communication tower according to the current historical best scheme. When the communication tower is in normal condition, it uses the deployed sensors to collect acceleration data in each time period, calculates the modal parameters in each time period under this condition, averages and summarizes them to obtain the reference mode parameters corresponding to different time periods. The early warning module 602 sets early warning thresholds and alarm thresholds. After updating the three-tube tower model according to the reference modal parameters, a reference model in normal condition is obtained. Real-time acceleration data is collected, and the current modal parameters are calculated based on the real-time acceleration data. The relative rate of change between the current modal parameters and the reference modal parameters is determined. If the relative rate of change exceeds the early warning threshold, the parameter is marked and an early warning message is generated. If any of the following constraints are met, an alarm is issued immediately. The specific constraints are: S101, Key parameters significantly exceed the standard: The rate of change of a single modal parameter continuously and significantly exceeds the alarm threshold; S102, Multi-parameter coordinated anomaly: The rate of change of three or more modal parameters simultaneously exceeds the warning threshold; S103. The amplitude set of all nodes at any order is severely degraded: the MAC value of the amplitude set drops below the alarm threshold; After receiving the alarm, the status analysis module 603 sets the damage parameters in the ANSYS model as design variables. With the goal of minimizing the error between the modal parameters calculated by the model and the current measured modal parameters, it drives multiple simulation iterations through optimization algorithms to continuously adjust the damage location and severity in the model until the error minimization target is achieved. Then, it outputs the damage identified in the model as the best estimate of the actual damage location and severity in the real structure, and analyzes the current health status of the communication tower based on the severity of the damage.

[0019] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0020] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A 5G base station structural anomaly early warning system, characterized in that, It includes a scheme generation unit (5) and a damage simulation unit (6); Model building unit (1), the model building unit (1) is built using ANSYS finite element analysis software according to the actual 5G communication tower structure and parameters. A scaled-down three-tube tower model; Amplitude analysis unit (2) calculates the first matrix, second matrix and third matrix of the three-tube tower model, analyzes the attribute characterization formula of the model according to the first matrix, second matrix and third matrix, calculates the modal parameters, and constructs the performance index evaluation formula based on the modal parameters; The node preliminary screening unit (3) obtains the number and position coordinates of all nodes in the three-tube tower model, analyzes the total kinetic energy of the nodes, filters the nodes according to the total kinetic energy, calculates the independent matrix based on the filtered nodes, removes the redundant nodes using the independent matrix, and outputs the number of the retained nodes. The segment division unit (4) uses a random sampling strategy to generate an initial set of schemes, calculates the evaluation value of all schemes in the set, and divides all schemes into four segments according to the evaluation value, namely the first segment, the second segment, the third segment and the fourth segment.

2. The 5G base station structural anomaly early warning system according to claim 1, characterized in that, The amplitude analysis unit (2) includes a matrix calculation module (201) and a position transformation combination module (202). The matrix calculation module (201) divides the structure of the three-tube tower model into a finite number of independent units. , Indicates the number of units. Indicates the first The first independent unit, read the first... Individual units volume Elasticity matrix Strain-displacement matrix Material density And shape function matrix Afterwards, according to Calculate the first matrix ,in , Representing volume, for and Analysis yields the second matrix. ,in Repeat the operation until all individual units have been read; The position transformation combination module (202) is a statistical independent unit. The first matrix and the second matrix Then, determine the position transformation matrix for each independent unit. , Indicates the first The position transformation matrix of each independent unit, using the first matrix Second matrix and position transformation matrix The first matrix of the three-tube tower model was calculated. Second matrix ,in , , Indicates the first Position transformation matrix of each independent unit Indicates the unit number.

3. The 5G base station structural anomaly early warning system according to claim 2, characterized in that, The amplitude analysis unit (2) further includes a force analysis module (203), an attribute characterization calculation module (204), and an amplitude set determination module (205). The force analysis module (203) obtains the damping coefficient. and Then, using Calculate the third matrix ,in According to the first matrix of the three-tube tower model Second matrix and the third matrix The external force vectors of each node in the model were calculated. ,in , This represents the acceleration vector of a node in the model. This represents the velocity vector of a node in the model. Represents the displacement vector of a node in the model; The attribute representation calculation module (204) when and When this happens, the displacement vector of the node is determined. ,in , Represents the imaginary unit. Indicates the vibration frequency. Indicates time, Represents the set of amplitudes at all nodes. Represents the natural constant, based on the displacement vector. Calculate the corresponding acceleration vector ,in ,use , as well as A property characterization formula for a three-tube tower was constructed. The amplitude set determination module (205) calculates the modal parameters according to the PCG Lanczos method and the characterization formula of the three-tube tower, and constructs the performance index evaluation formula based on the amplitude set order of all nodes in the first 20 orders of the modal parameters.

4. The 5G base station structural anomaly early warning system according to claim 1, characterized in that, The node preliminary screening unit (3) includes a node removal module (301), a total kinetic energy calculation module (302), an independent matrix generation module (303), and a node number output module (304). The node removal module (301) obtains the number and position coordinates of all nodes in the three-tube tower model, sets the retention number, removes invalid nodes, symmetrical nodes and dense nodes until the number of remaining nodes reaches the retention number, and stores the unremoved nodes in the initial candidate set. The total kinetic energy calculation module (302) reads the set of amplitudes of all nodes corresponding to the first 20 orders, and after obtaining the displacement vectors of all nodes in the initial candidate set under each order mode based on the set of amplitudes of all nodes, it randomly selects a node. , Represents the node sequence number, and is used for statistics. Displacement vectors at all orders ,in , This represents the position vector at the 20th order. Representing the three translational displacement components at the 20th order, according to Calculate the nodes Total kinetic energy ,in Repeat the operation to calculate the total kinetic energy of all nodes in the initial candidate set; The independent matrix generation module (303) sorts all nodes in descending order of total kinetic energy, retains the top 50% of nodes, and removes the remaining nodes. Based on the retained nodes, it constructs the amplitude set corresponding to the first 20 orders. Using amplitude sets Calculate the independent matrix ,in ; The node number output module (304) selects the smallest value of the diagonal element in the independent matrix, removes the node corresponding to the value, constructs a new amplitude set, recalculates the independent matrix based on the new amplitude set, repeats the operation until the number of nodes in the candidate set is equal to the threshold, and then outputs all node numbers.

5. A 5G base station structural anomaly early warning system according to claim 1, characterized in that, The segment division unit (4) includes a scheme set generation module (401), an evaluation value output module (402), and a sequence division module (403). The scheme set generation module (401) uses a random sampling strategy to generate an initial scheme set. ,in Indicates the first One option, , This indicates the total number of schemes. Indicates the first Nodes The deployment situation on the above, if This indicates that the node is equipped with sensors. This indicates that no sensors are installed at that node. Indicates the number of nodes in the candidate set; The evaluation value output module (402) randomly selects a scheme. , Indicates the scheme number, according to The sensor positions are determined by the arrangement of each node. The displacement vectors of the corresponding nodes at each order are extracted from the set of all node amplitudes for each order, and these vectors are combined to obtain the set of node amplitudes at different orders. , Representation scheme The Middle The set of nodal amplitudes of order, according to Calculate Evaluation value ,in , Indicates the total order. Representation scheme The Middle The set of nodal amplitudes of order, Representation scheme The Middle The set of nodal amplitudes of order, The parameters are represented, and the operation is repeated until the evaluation values ​​of all solutions are calculated. Then, the current number of executions and the maximum number of executions are determined. The sequence partitioning module (403) divides all schemes according to their evaluation values. After sorting, the current sequence is divided into four segments according to preset ratio parameters: the first segment, the second segment, the third segment, and the fourth segment.

6. A 5G base station structural anomaly early warning system according to claim 1, characterized in that, The scheme generation unit (5) includes a first segment analysis module (501) and a second segment analysis module (502); The first segment analysis module (501) extracts all schemes in the first segment. , Indicates the number of schemes. Indicates the first segment One option, determine the historical best option. Then, select a solution. ,Will and Compare and statistically analyze Sensors are deployed but There are no all the nodes where the sensors are deployed; instead, a node is randomly selected, and... The deployment status of this node was changed from 1 to 0, and statistics were also compiled. There is no sensor deployment but Given all the nodes where sensors are deployed, randomly select one node to perform a test. After changing the node's deployment status from 0 to 1, in the current scheme Two nodes are randomly selected from all nodes, with their configurations set to 1 and 0 respectively. These nodes are then modified to 0 and 1 respectively. The current... The evaluation value, if the evaluation value is less than the previous one The evaluation value will then be the current Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the first segment are selected. The second segment analysis module (502) extracts all schemes in the second segment. , Indicates the number of schemes. Indicates the second segment Choose one of the following options. According to the plan right After adjusting the deployment of all nodes in the current scheme, Four nodes are randomly selected from all nodes, with their initial configurations being 1, 1, 0, and 0 respectively. These nodes are then modified to their corresponding configurations of 0, 0, 1, and 1. The current... The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the second segment are selected.

7. A 5G base station structural anomaly early warning system according to claim 6, characterized in that, The scheme generation unit (5) further includes a third segment analysis module (503), a fourth segment analysis module (504), and an optimal scheme output module (505); The third segment analysis module (503) extracts all schemes in the third segment. back, Indicates the number of schemes. Indicates the third segment Choose one of the following options. A new scheme is generated directly using a random sampling strategy, and adjustments are made according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the third segment are selected. The fourth segment analysis module (504) extracts all schemes in the fourth segment. Then, select a solution. Two schemes are randomly selected from the first segment. A portion of each scheme is taken and spliced ​​together to generate a new scheme. The scheme is then adjusted according to the new scheme. Calculate the current The evaluation value, if the evaluation value is less than the previous one The current evaluation value is... Replace with the one before selection Conversely, no operation is performed, and the operation is repeated until all schemes in the fourth segment are selected. The optimal solution output module (505) selects the solution with the largest evaluation value and compares it with the historical optimal solution. The optimal solution is retained in the historical optimal solution. The current execution count is automatically incremented by one. The module determines whether the current execution count is equal to the maximum execution count. If it is equal, the current historical optimal solution is directly output. Otherwise, the module enters the next round of loop.

8. A 5G base station structural anomaly early warning system according to claim 1, characterized in that, The damage simulation unit (6) includes a baseline mode calculation module (601), an early warning module (602), and a state analysis module (603). The reference mode calculation module (601) deploys acceleration sensors at each node of the actual communication tower according to the current historical best scheme. When the communication tower is in normal condition, it uses the deployed sensors to collect acceleration data in each time period, calculates the modal parameters in each time period under this condition, averages and summarizes them to obtain the reference mode parameters corresponding to different time periods. The early warning module (602) sets early warning thresholds and alarm thresholds. After updating the three-tube tower model according to the reference modal parameters, it obtains the reference model in normal state, collects real-time acceleration data, calculates the current modal parameters according to the real-time acceleration data, determines the relative rate of change between the current modal parameters and the reference modal parameters, and marks the parameter if the relative rate of change exceeds the early warning threshold and generates early warning information. If any of the limiting conditions are met, an alarm is issued immediately. After receiving the alarm, the status analysis module (603) sets the damage parameters in the ANSYS model as design variables. With the goal of minimizing the error between the modal parameters calculated by the model and the current measured modal parameters, it drives multiple simulation iterations through optimization algorithms to continuously adjust the damage location and severity in the model until the error minimization target is achieved. Then, it outputs the damage identified in the model as the best estimate of the actual damage location and severity in the real structure, and analyzes the current health status of the communication tower based on the severity of the damage.

Images