A bionic topology and replaceable energy-consuming node uhpc-steel mixed tower drum and its seismic design method and system

Abstract

The application discloses a kind of UHPC-steel mixed tower cylinder of bionic topology and replaceable energy dissipation node and its seismic design method and system, belong to wind power structure seismic and intelligent construction technical field.Tower cylinder includes concrete foundation, bionic topology steel mixed segment, UHPC concrete segment and top steel segment, replaceable energy dissipation node system is arranged in the periphery of tower cylinder bottom, and energy dissipation unit is connected with tower cylinder main body by detachable connecting piece, realize that earthquake damage is concentrated in peripheral component easy to replace;Steel mixed segment adopts three-dimensional space steel framework and UHPC filling layer composite structure obtained by automatic optimization of generative intelligent algorithm, realize smooth transmission of force flow.Seismic design method includes intelligent topological optimization generation steel mixed segment configuration, manufacturing installation replaceable energy dissipation node and the steps of rapid evaluation and repair after earthquake based on unmanned aerial vehicle vision and digital twin.The application realizes the toughness seismic system of "main body elasticity, energy dissipation replaceable", significantly improves structural seismic performance, material utilization efficiency and rapid recovery ability after earthquake.

Description

Technical Field

[0001] This invention belongs to the field of wind power structure seismic resistance and intelligent construction technology, specifically involving an ultra-high performance concrete (UHPC)-steel-concrete tower structure combining generative topology optimization and replaceable energy dissipation nodes and its seismic design method, especially a biomimetic topology and replaceable energy dissipation nodes UHPC-steel-concrete tower and its seismic design method and system. Background Technology

[0002] As wind power technology continues to develop towards larger capacity and taller towers, traditional pure steel or pure concrete towers are no longer sufficient to balance economy and safety. Ultra-high performance concrete (UHPC)-steel composite towers, combining the high strength and durability of UHPC with the excellent ductility of steel, are gradually becoming the mainstream tower type for high-power wind turbines. This composite structure can fully utilize material properties, reducing structural weight while increasing load-bearing capacity, meeting the needs of the wind power industry's expansion into deep-sea and high-wind-shear areas. However, as tower height continues to increase, its response to seismic loads becomes increasingly complex, placing higher demands on its seismic design.

[0003] Despite the numerous advantages of UHPC-concrete composite tower structures, existing technologies still have significant shortcomings in specific construction and design methods. Firstly, in the concrete-steel joint section, traditional designs often employ grouted sleeve connections or flange connections. Due to the significant stiffness difference between the two materials, this connection method results in a substantial stiffness abrupt change at the joint interface. Under cyclic loading such as earthquakes, this area of ​​stiffness abrupt change is prone to stress concentration, leading to peeling failure of the UHPC layer or fatigue cracking in the steel frame joint area. More importantly, these damages typically occur within the structure or at the joint interface, making effective detection and repair difficult once they occur. Secondly, existing tower structures lack consideration for post-earthquake functional recovery in their seismic design concepts. Once a traditional tower structure is damaged in an earthquake, because its components are mostly integral or permanently connected, the entire tower segment or even the entire tower usually needs to be replaced. This results in extremely high repair costs and lengthy downtime, failing to meet the modern requirement for "resilient structures," meaning structures that can quickly restore their functionality after an earthquake. Secondly, the steel-concrete composite section, as the most complex area of ​​stress in the entire tower, has long relied heavily on engineers' personal experience for its structural design. This experience-driven design approach struggles to accurately capture the optimal synergy between the high tensile strength of UHPC and the ductility of steel, often resulting in structures that are either overly conservative, with low material utilization and high structural redundancy, or have localized weak points, failing to achieve truly high-performance design. Finally, current post-earthquake damage assessment of wind turbine towers primarily relies on manual inspection. This method is not only inefficient and poses safety risks associated with working at heights, but also struggles to accurately assess the degree of plastic damage to the internal steel frame, failing to provide precise and quantitative guidance for post-earthquake repair work and severely restricting the structure's rapid recovery capabilities.

[0004] In summary, developing a novel UHPC-steel-concrete tower structure and its seismic design method that enables rapid post-earthquake recovery, high material utilization, and intelligent damage assessment has become an urgent technical problem to be solved in this field. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a UHPC-reinforced concrete tower with biomimetic topology and replaceable energy-dissipating nodes, as well as its seismic design method. By introducing replaceable energy-dissipating components, a resilient seismic-resistant system with "elastic main body and replaceable energy dissipation" is achieved. Generative artificial intelligence technology is used to automatically generate the optimal topological configuration of the reinforced concrete joint section, and machine vision is combined to achieve rapid post-earthquake assessment. This aims to significantly improve the seismic performance, material utilization efficiency, and post-earthquake recovery capability of the tower.

[0006] To solve the above-mentioned technical problems, the technical solution proposed in this application is as follows:

[0007] This invention provides a UHPC-reinforced concrete tower with biomimetic topology and replaceable energy-consuming nodes, comprising, from bottom to top, a concrete foundation, a biomimetic topology reinforced concrete section, a UHPC concrete section, and a top steel section, and further comprising: A replaceable energy-consuming node system is set at the periphery of the connection between the biomimetic topology steel-concrete composite section and the concrete foundation, and includes multiple energy-consuming units evenly arranged in the circumferential direction. The energy-consuming unit includes an upper connecting plate, a lower connecting plate, and energy-consuming components; The upper connecting plate and the lower connecting plate are respectively anchored to the outer wall of the tower and the foundation through detachable connectors; The energy-consuming component is connected between the upper connecting plate and the lower connecting plate; The energy-consuming unit is connected to the tower body via a detachable connection structure.

[0008] Furthermore, the biomimetic topological steel-concrete composite section includes a three-dimensional spatial steel skeleton and a UHPC filling layer; The three-dimensional spatial steel skeleton is made of high-strength steel and is formed into a mesh structure with spatial curved surfaces and holes through casting or welding. Its geometric shape is the optimal configuration generated by intelligent topology optimization. The UHPC filling layer is cast inside and around the steel frame, forming a combined load-bearing unit with the steel frame.

[0009] Furthermore, the topological configuration of the three-dimensional spatial steel frame has an anisotropic stiffness distribution, which enables the joint section to transmit force uniformly under vertical loads and form multiple plastic hinge zones under horizontal seismic action; the surface of the steel frame is provided with shear connectors.

[0010] Furthermore, the energy dissipation component includes at least one of a buckling restraint support component, a friction damper, a metal yield damper, or a viscoelastic damper; the buckling restraint support component includes an inner core and an outer restraint sleeve, the inner core is made of low yield point steel, the outer restraint sleeve is a UHPC filled steel pipe, and a debonding layer is provided between the inner core and the outer restraint sleeve.

[0011] Furthermore, the number of energy-consuming units is 4 to 8, which are evenly arranged along the circumference of the tower. Each energy-consuming unit is designed to be replaceable independently. The yield bearing capacity of each energy-consuming unit is designed to be 50% to 80% of the elastic bearing capacity of the main structure.

[0012] On the other hand, this application also claims protection for a seismic design method based on any of the foregoing UHPC-reinforced concrete towers, comprising the following steps: Step S1: Intelligent topology optimization generates the steel-concrete composite section configuration. Specifically, the tower design parameters and material performance parameters are obtained. With the optimization objectives of maximizing structural stiffness, minimizing mass, and homogenizing stress distribution, and with process feasibility as the constraint, the optimal topology configuration of the three-dimensional steel skeleton is automatically generated using a data-driven generative intelligent algorithm. Step S2: Fabricate and install replaceable energy dissipation nodes. Specifically, fabricate a steel-concrete composite section based on the optimal topology generated in Step S1, and install multiple energy dissipation units evenly arranged in the circumferential direction on the outer periphery of the bottom of the tower through detachable connectors. The energy dissipation components of each energy dissipation unit are designed to yield and dissipate energy first under the design earthquake, and their yield bearing capacity is lower than the elastic bearing capacity of the main body of the tower. Step S3: Rapid post-earthquake assessment and repair, specifically: After the earthquake, images of the outer surface of the tower are collected by drones, crack features are identified by machine vision, internal damage is inverted by digital twin model and the energy-consuming units that need to be replaced are determined, and a repair plan is generated to guide the replacement work.

[0013] Furthermore, in step S1, the generative intelligent algorithm adopts a framework based on variational autoencoders, generative adversarial networks, or diffusion models, and generates new configurations that meet the constraints by extracting the implicit features of existing high-performance topological configurations; the multi-objective optimization adopts the Pareto front optimization method to generate multiple non-dominated solutions for selection.

[0014] Furthermore, step S1 also includes: performing finite element verification on the generated optimal topology configuration; if the verification result does not meet the preset performance index, it is fed back to the generative intelligent algorithm for iterative optimization until it converges to the optimal solution.

[0015] Furthermore, in step S2, each energy-consuming unit adopts a graded yielding design, including a first yielding stage and a second yielding stage. The yield bearing capacity of the first yielding stage is lower than that of the second yielding stage, forming multiple energy-consuming defense lines.

[0016] Further, in step S3, the use of machine vision to identify crack features specifically includes: automatically identifying the crack width, density, orientation, and distribution area on the UHPC surface using a trained deep convolutional neural network model; the damage inversion specifically involves: inputting the crack features and seismic records into the damage inversion module, and combining the finite element model or physical information neural network model to calculate the cumulative damage of the plastic strain inside the steel frame and the energy dissipation units. The formula for calculating the cumulative damage is:

[0017] Where D is the cumulative damage index, and n is the total number of cycles for the accumulation of plastic strain under seismic loading. Let be the plastic strain increment for the i-th cycle. This represents the ultimate plastic strain of the material.

[0018] On the other hand, this application also claims protection for a system for implementing the aforementioned seismic design method, comprising: The intelligent topology optimization design system is used to perform step S1, including a parametric modeling module, a multi-objective optimization solver, and an iterative verification module; The parametric modeling module is used to establish the initial design domain of the steel-concrete composite section; The multi-objective optimization solver is used to automatically generate the topology of the steel frame with the objectives of maximizing structural stiffness, minimizing mass, and homogenizing stress distribution using a data-driven generative intelligent algorithm. The iterative verification module is used to perform finite element verification on the generated topology and feed the verification results back to the multi-objective optimization solver for iterative optimization. The post-earthquake rapid assessment system, used to perform step S3, includes an unmanned aerial vehicle (UAV)-borne visual acquisition module, a machine vision recognition module, a damage inversion module, and a repair decision module. The UAV-borne visual acquisition module is used for automatic cruise photography of the outer surface of the tower after an earthquake. The machine vision recognition module has a built-in trained crack recognition model for automatically identifying crack features on the UHPC surface. The damage inversion module is used to invert the degree of plastic damage inside the steel frame based on the surface crack characteristics and the digital twin model, and to determine whether each energy-consuming unit needs to be replaced. The repair decision module is used to generate a visual damage report and a replacement list.

[0019] Furthermore, the post-earthquake rapid assessment system also includes a health data fusion module, which is used to compare and analyze the post-earthquake assessment results with the pre-earthquake health monitoring data to generate a report on the trend of structural performance degradation.

[0020] Furthermore, it also includes a cloud-based collaborative platform for receiving and storing monitoring data, topology optimization results, and post-earthquake assessment reports from each tower, and for updating machine learning models based on data from multiple towers to optimize the generation capabilities of generative intelligent algorithms and the recognition accuracy of crack identification models.

[0021] Compared with the prior art, the present invention achieves the following beneficial technical effects: This invention concentrates earthquake damage on easily replaceable peripheral components by using replaceable energy-dissipating nodes, achieving a resilient earthquake-resistant system that is "elastic in its main structure and replaceable in its energy dissipating components." After an earthquake, only the energy-dissipating units need to be replaced to quickly restore functionality. It utilizes generative intelligent algorithms to automatically optimize the topology of the steel-concrete composite section, achieving smooth force flow transmission and efficient material utilization. Combined with UAV vision and digital twin technology, it enables intelligent and rapid post-earthquake assessment and precise repair, significantly improving the structure's seismic toughness, material utilization efficiency, and post-earthquake recovery capability. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart of the seismic design method for UHPC-steel-concrete tower provided in an embodiment of the present invention.

[0024] Figure 2 This is a structural block diagram of a system for implementing a seismic design method for UHPC-reinforced concrete towers, provided as an embodiment of the present invention. Detailed Implementation

[0025] 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 are only some embodiments of the present invention, 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.

[0026] In one embodiment of this application, a UHPC-reinforced concrete tower with biomimetic topology and replaceable energy-dissipating nodes is provided. The tower comprises, from bottom to top, a concrete foundation, a biomimetic topological reinforced concrete section, a UHPC concrete section, and a top steel section. Specifically, this embodiment uses a 160m high wind turbine tower with a rated power of 5MW as an example. The tower is segmented as follows: 0 to 10 meters is the concrete foundation, 10 to 20 meters is the biomimetic topological reinforced concrete section, 20 to 140 meters is the UHPC concrete section, and 140 to 160 meters is the top steel section. The replaceable energy-dissipating node system is located around the connection between the biomimetic topological reinforced concrete section and the concrete foundation, specifically 2 meters from the top surface of the foundation, and includes multiple energy-dissipating units evenly arranged circumferentially. Each energy-dissipating unit includes an upper connecting plate, a lower connecting plate, and an energy-dissipating component. The upper and lower connecting plates are anchored to the outer wall of the tower and the foundation via detachable connectors. The energy dissipation components are connected between the upper and lower connecting plates, thus connecting the entire energy dissipation unit to the main tower body through a detachable connection structure. Six energy dissipation units are evenly distributed around the bottom perimeter of the tower. Each energy dissipation unit is 3 meters high, and its yield bearing capacity is designed to be 60% of the elastic bearing capacity of the main structure. This ensures that the energy dissipation units yield and dissipate energy first under the design earthquake, thereby protecting the main structure and maintaining its elasticity. After a minor or moderate earthquake, the damaged energy dissipation units can be quickly removed from the outer wall of the tower and the foundation, and replaced with new energy dissipation units to quickly restore the tower's seismic resistance.

[0027] In one embodiment of this application, the energy dissipation component in the aforementioned energy dissipation unit can be in the form of a buckling restraint support assembly. Specifically, the buckling restraint support assembly includes an inner core and an outer restraint sleeve. The inner core is made of Q235 low yield point steel, and its cross-section can be cross-shaped, I-shaped, or circular. The outer restraint sleeve is a steel pipe filled with C150 grade UHPC. A debonding layer is provided between the inner core and the outer restraint sleeve. This debonding layer can be a silicone layer, a rubber layer, or an air gap layer to ensure that the inner core can freely yield and dissipate energy under tension and compression, while the outer restraint sleeve effectively prevents the inner core from buckling as a whole. In other embodiments, the energy dissipation component can also be one or more combinations of friction dampers, metal yield dampers, or viscoelastic dampers.

[0028] In one embodiment of this application, the tower further includes a biomimetic topological steel-concrete composite section, which is located between the concrete foundation and the UHPC concrete section, with a height of 10 meters. This biomimetic topological steel-concrete composite section includes a three-dimensional spatial steel frame and a UHPC filling layer. The three-dimensional spatial steel frame is made of ZG25Mn cast steel through an integral casting process, forming a mesh structure with spatial curved surfaces and holes. Its geometry is an optimal configuration generated by intelligent topology optimization. The steel frame has a biomimetic skeletal shape with continuously varying internal cavities. Shear connectors are provided on the surface of the steel frame, which can be in the form of studs, perforated steel plates, or corrugated steel plates to enhance the bonding performance and collaborative working ability between the steel frame and the UHPC. The UHPC filling layer is cast inside and around the steel frame, forming a combined load-bearing unit with the steel frame. The UHPC uses C150 grade ultra-high performance concrete.

[0029] In one embodiment of this application, the topological configuration of the aforementioned three-dimensional spatial steel frame has an anisotropic stiffness distribution, with its equivalent stiffness matrix having different stiffness values ​​in three orthogonal directions. This anisotropic stiffness distribution enables the joint section to achieve uniform force transmission when subjected to vertical loads, avoiding local stress concentration; and under horizontal seismic action, it can form plastic hinge zones in multiple predetermined areas of the steel frame, thereby dissipating seismic energy and avoiding stress concentration and failure of a single section. Furthermore, the porosity of the topological configuration of this steel frame can be designed to be between 30% and 70%, and the internal pores are distributed in a gradient, for example, the porosity gradually increases or decreases from the inside to the outside along the radial direction of the tower. This gradient distribution further optimizes the force flow transmission path and material utilization efficiency.

[0030] In one embodiment of this application, the aforementioned three-dimensional spatial steel skeleton adopts a biomimetic structural design, with its topological form derived from biomimetic optimizations of biological skeletons, honeycomb structures, spider web structures, or plant root systems. For example, the hollow porous structure mimicking biological skeletons ensures sufficient strength and stiffness while achieving lightweighting; the hexagonal grid mimicking honeycomb structures can provide uniform load-bearing capacity in all directions. This biomimetic structure has continuously varying cross-sections and spatial curved surfaces, enabling smooth force transmission and avoiding stress concentration caused by abrupt changes in stiffness in traditional connection methods.

[0031] In one embodiment of this application, the replaceable energy dissipation node system further includes a monitoring module. This monitoring module includes strain sensors, displacement sensors, or acceleration sensors installed on each energy dissipation unit, used to collect real-time stress and deformation data of the energy dissipation unit. These sensors can wirelessly transmit the collected data to a cloud server or a central wind farm monitoring system in real time, enabling remote real-time monitoring of the energy dissipation unit's operating status and providing fundamental data support for post-earthquake assessment and predictive maintenance.

[0032] In one embodiment of this application, the aforementioned detachable connector can be in various forms, such as high-strength bolt connectors, prestressed anchor bolt connectors, or clamp-type connectors. The yield bearing capacity of each energy dissipation unit is designed to be between 50% and 80% of the elastic bearing capacity of the main structure. The specific value can be optimized and determined according to the actual height of the tower, load conditions, and seismic fortification level, so as to achieve the predetermined goal of the energy dissipation unit yielding and dissipating energy first under the fortification earthquake.

[0033] refer to Figure 1 In one embodiment of this application, a seismic design method based on the aforementioned UHPC-reinforced concrete tower is provided. This method includes three main steps: intelligent topology optimization to generate the reinforced concrete composite section configuration, fabrication and installation of replaceable energy-dissipating nodes, and rapid post-earthquake assessment and repair.

[0034] The specific implementation method for generating the steel-concrete composite section configuration through intelligent topology optimization is as follows: First, obtain the tower design parameters and material performance parameters. The tower design parameters include tower height, design load, seismic fortification level, etc., while the material performance parameters include UHPC compressive strength and steel yield strength, etc. Taking this embodiment as an example, the obtained design parameters are the design bending moment of 200MN·m, vertical force of 30MN, and shear force of 8MN at the bottom of the tower, and the material parameters are UHPC compressive strength of 150MPa and steel yield strength of 300MPa. Then, set the optimization objectives, including maximizing structural stiffness, minimizing structural mass, and minimizing stress concentration factor, where the target value of stress concentration factor is set to be less than 1.5. At the same time, set the constraints, including the castability requirements of the steel frame, the requirement for flow pore diameter without air bubble retention during UHPC casting, and process constraints such as minimum component size. Next, using a data-driven generative intelligent algorithm, automatically generate a large number of candidate topology configurations within the design domain. In this embodiment, a generative model based on variational autoencoder is used. Using 2000 existing high-performance topological configurations as a training set, its implicit features are learned, and 50 candidate configurations are automatically generated after inputting design parameters. Finally, finite element analysis is used to perform a detailed performance evaluation of each candidate configuration, including stiffness, mass, stress distribution, and construction feasibility. The optimal steel frame geometry that meets multiple objectives is selected as the final design for the steel-concrete composite section. The optimal configuration selected in this embodiment presents a spatial spiral mesh structure, reducing mass by 18% compared to traditional solid steel cylinders, decreasing the stress concentration factor from 3.2 in traditional structures to 1.3, and exhibiting good casting processability.

[0035] In one embodiment of this application, the aforementioned generative intelligent algorithm can employ a framework based on variational autoencoders, generative adversarial networks (GANs), or diffusion models. Taking a variational autoencoder as an example, it maps the input high-dimensional topological configuration to a low-dimensional latent space through an encoder, extracts latent features, and then reconstructs a new topological configuration from the latent space through a decoder. Generative adversarial networks, through adversarial training between a generator and a discriminator, enable the generator to generate realistic, high-performance topological configurations. Diffusion models generate new configurations by progressively adding and then denoising noise. In the multi-objective optimization process, Pareto front optimization methods can be used to generate multiple non-dominated solutions for designers to select from, thereby obtaining the topological configuration with the best overall performance.

[0036] In one embodiment of this application, the above-mentioned intelligent topology optimization step further includes an iterative optimization process: the generated optimal topology configuration is verified by finite element method. If the verification result does not meet the preset performance index, such as the stress concentration factor being greater than 1.5 or the mass exceeding the target value, the verification result is fed back to the generative intelligent algorithm as a new constraint condition for iterative optimization until it converges to the optimal solution that satisfies all performance indexes.

[0037] The specific implementation method for fabricating and installing replaceable energy dissipation nodes is as follows: First, based on the optimal topology generated in the above steps, a three-dimensional spatial steel frame is fabricated using casting or welding processes, and UHPC is poured on-site to form a steel-concrete composite section. Then, connecting lugs are pre-embedded on the outer periphery of the tower bottom, and multiple energy dissipation units evenly arranged circumferentially are installed through detachable connectors to form a replaceable energy dissipation node system. In this embodiment, six energy dissipation units are evenly arranged circumferentially on the outer periphery of the tower bottom, each energy dissipation unit being 3 meters high and using buckling-restrained bracing. The energy dissipation components of each energy dissipation unit are designed to yield and dissipate energy first under the design earthquake, and their yield bearing capacity is lower than the elastic bearing capacity of the tower body. Specifically, the designed yield bearing capacity is 60% of the elastic bearing capacity of the main structure, thereby ensuring that under the action of the design earthquake, the energy dissipation unit first enters the plastic state to absorb seismic energy, protecting the main structure and maintaining its elasticity.

[0038] In one embodiment of this application, each energy dissipation unit adopts a graded yielding design, including a first yielding stage and a second yielding stage. For example, the energy dissipation unit is designed as a part with two different yield bearing capacities. The yield bearing capacity of the first yielding stage is lower, and it yields and dissipates energy first under the action of small and medium earthquakes. The yield bearing capacity of the second yielding stage is higher, and it plays a role under the action of rare earthquakes, forming multiple energy dissipation defense lines and further enhancing the seismic safety of the structure.

[0039] The specific implementation method for rapid post-earthquake assessment and repair is as follows: After the earthquake, an unmanned aerial vehicle (UAV) is activated for automatic cruise to collect high-resolution images of the tower's outer surface. In this embodiment, a pre-planned automatic cruise program is initiated, and the UAV spirals upward along the tower, capturing a full-coverage image of the outer surface with a resolution of 0.1 mm / pixel. The entire process takes approximately 25 minutes, and the image data is transmitted to the cloud processing platform in real time. Then, the machine vision recognition module automatically extracts crack features and compares them with a preset damage level database, outputting a crack distribution map. In this embodiment, a convolutional neural network model deployed in the cloud automatically processes the images, identifies micro-cracks near the energy dissipation unit and on the surface of the steel-concrete composite section, and outputs a crack distribution map showing a maximum crack width of 0.15 mm, with no cracks observed in the main structure. Next, the crack features are input into the damage inversion module, which, combined with seismic records and a structural response model, calculates the plastic strain inside the steel frame and the cumulative damage to the energy dissipation unit. In this embodiment, crack characteristics and seismic records are input into the damage inversion module. Combined with the finite element model, the maximum plastic strain inside the steel frame is calculated to be 0.0005, far less than the yield strain. Two of the six energy-dissipating units have accumulated plastic deformation exceeding the design limit and require replacement. Finally, the repair decision module generates a repair plan, specifying the numbers and locations of the energy-dissipating units that need replacement. In this embodiment, a repair report is generated, specifying the energy-dissipating units to be replaced as #2 and #5, and providing replacement steps, including bolt torque requirements and the new unit model. Based on the report, maintenance personnel complete the bolt removal and replacement of the two energy-dissipating units within six hours after the earthquake. The tower resumes operation, and the main structure remains unrepaired.

[0040] In one embodiment of this application, the specific implementation of the above-mentioned machine vision crack feature recognition is as follows: a trained deep convolutional neural network model is used to automatically identify the crack width, density, direction, and distribution area of ​​the UHPC surface, achieving a crack recognition accuracy of 0.1 mm. This model is trained using a large number of labeled UHPC crack images and can accurately distinguish cracks from interfering factors such as surface textures and stains, achieving precise extraction of crack features. The crack features specifically include parameters such as crack length, crack width, crack density, crack direction, and crack distribution area.

[0041] In one embodiment of this application, the specific implementation of the damage inversion is as follows: crack characteristics and seismic records are input into the damage inversion module, and the cumulative damage of the plastic strain and energy dissipation units inside the steel frame is calculated by combining a finite element model or a physical information neural network model. The formula for calculating the cumulative damage is:

[0042] Where D is the cumulative damage index, and n is the total number of cycles for the accumulation of plastic strain under seismic loading. Let be the plastic strain increment for the i-th cycle. This represents the ultimate plastic strain of the material.

[0043] This formula allows for the establishment of a quantitative correlation between surface-observed crack characteristics and internal plastic damage, enabling an accurate assessment of the internal damage state of the structure.

[0044] In one embodiment of this application, the aforementioned rapid post-earthquake assessment step further includes a health data fusion step: comparing and analyzing the post-earthquake assessment results with pre-earthquake health monitoring data to identify structural performance degradation trends and generate predictive maintenance recommendations. For example, if a certain energy-consuming unit accumulates damage after multiple earthquakes, the system can predict that it may fail in a future earthquake, issue an early warning, and recommend replacement, thereby achieving a shift from passive repair to proactive maintenance.

[0045] refer to Figure 2 In one embodiment of this application, a system for implementing the above-described seismic design method is provided. This system includes an intelligent topology optimization design system and a rapid post-earthquake assessment system.

[0046] The intelligent topology optimization design system is used to execute step S1 above, and includes a parametric modeling module, a multi-objective optimization solver, and an iterative verification module. The parametric modeling module establishes the initial design domain for the steel-concrete composite section, which is parametrically described based on the tower's geometric parameters and load conditions. The multi-objective optimization solver automatically generates the topology of the steel frame using a data-driven generative intelligent algorithm, with the objectives of maximizing structural stiffness, minimizing mass, and homogenizing stress distribution, and using UHPC material properties, steel yield strength, and process constraints as constraints. The iterative verification module performs finite element verification on the generated topology and feeds the verification results back to the multi-objective optimization solver for iterative optimization until the optimal configuration satisfying all performance indicators is obtained.

[0047] The post-earthquake rapid assessment system is used to perform step S3 above, including an UAV-borne visual acquisition module, a machine vision recognition module, a damage inversion module, and a repair decision module. The UAV-borne visual acquisition module is used for automatic cruise imaging of the tower's outer surface after the earthquake; its flight path can be pre-planned to achieve full coverage imaging of the tower surface. The machine vision recognition module has a built-in trained crack recognition model for automatically identifying crack features on the UHPC surface. This model uses a deep convolutional neural network structure and is trained using a large number of labeled UHPC crack images. The damage inversion module is used to invert the degree of plastic damage inside the steel frame based on surface crack features and a digital twin model, and to determine whether each energy-consuming unit needs to be replaced. The repair decision module generates a visualized damage report and replacement list, clearly identifying the numbers and locations of the energy-consuming units that need to be replaced, and providing instructions on the replacement steps.

[0048] In one embodiment of this application, the aforementioned post-earthquake rapid assessment system further includes a health data fusion module, used to compare and analyze the post-earthquake assessment results with pre-earthquake health monitoring data to generate a structural performance degradation trend report. This module can integrate historical monitoring data, analyze the performance degradation patterns of energy-consuming units and the main structure, and provide a scientific basis for operation and maintenance decisions.

[0049] In one embodiment of this application, the system further includes a cloud-based collaborative platform for receiving and storing monitoring data, topology optimization results, and post-earthquake assessment reports from each wind farm. Based on data from multiple wind farms, the platform updates the machine learning model, optimizing the generative intelligent algorithm's generation capability and the crack identification model's accuracy. This cloud-based collaborative platform employs a federated learning framework, training the model locally at each wind farm and only uploading model parameters to the cloud for aggregation and updates. This allows for continuous model optimization and iterative upgrades while protecting data privacy.

[0050] In summary, the UHPC-concrete composite tower and its seismic design method proposed in this invention, by organically combining intelligent topology optimization, replaceable energy-consuming nodes and intelligent evaluation technology, successfully constructs a resilient seismic-resistant system with "elastic main body and replaceable energy consumption", which significantly improves the seismic safety, economy and rapid post-earthquake recovery capability of wind turbine towers, and has extremely high practical value and broad application prospects.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A UHPC-reinforced concrete tower with biomimetic topology and replaceable energy-consuming nodes, characterized in that, It includes, from bottom to top, a concrete foundation, a biomimetic topological steel-concrete composite section, a UHPC concrete section, and a top steel section, and also includes: A replaceable energy-consuming node system is set at the periphery of the connection between the biomimetic topology steel-concrete composite section and the concrete foundation, and includes multiple energy-consuming units evenly arranged in the circumferential direction. The energy-consuming unit includes an upper connecting plate, a lower connecting plate, and energy-consuming components; The upper connecting plate and the lower connecting plate are respectively anchored to the outer wall of the tower and the foundation through detachable connectors; The energy-consuming component is connected between the upper connecting plate and the lower connecting plate; The energy-consuming unit is connected to the tower body via a detachable connection structure.

2. The UHPC-reinforced concrete tower according to claim 1, characterized in that, The biomimetic topological steel-concrete composite section includes a three-dimensional spatial steel skeleton and a UHPC filling layer; The three-dimensional spatial steel skeleton is made of high-strength steel and is formed into a mesh structure with spatial curved surfaces and holes through casting or welding. Its geometric shape is the optimal configuration generated by intelligent topology optimization. The UHPC filling layer is cast inside and around the steel frame, forming a combined load-bearing unit with the steel frame.

3. The UHPC-reinforced concrete tower according to claim 2, characterized in that, The topological configuration of the three-dimensional spatial steel frame has an anisotropic stiffness distribution, which enables the joint section to transmit force uniformly under vertical loads and form multiple plastic hinge zones under horizontal seismic action; the surface of the steel frame is provided with shear connectors.

4. The UHPC-reinforced concrete tower according to claim 1, characterized in that, The energy dissipation component includes at least one of a buckling restraint support component, a friction damper, a metal yield damper, or a viscoelastic damper; the buckling restraint support component includes an inner core and an outer restraint sleeve, the inner core is made of low yield point steel, the outer restraint sleeve is a UHPC filled steel pipe, and a debonding layer is provided between the inner core and the outer restraint sleeve.

5. The UHPC-reinforced concrete tower according to claim 1, characterized in that, The number of energy-consuming units is 4 to 8, which are evenly arranged along the circumference of the tower. Each energy-consuming unit is designed to be replaceable independently. The yield bearing capacity of each energy-consuming unit is designed to be 50% to 80% of the elastic bearing capacity of the main structure.

6. A seismic design method for a UHPC-reinforced concrete tower according to any one of claims 1 to 5, characterized in that, Includes the following steps: Step S1: Intelligent topology optimization generates the steel-concrete composite section configuration. Specifically, the tower design parameters and material performance parameters are obtained. With the optimization objectives of maximizing structural stiffness, minimizing mass, and homogenizing stress distribution, and with process feasibility as the constraint, the optimal topology configuration of the three-dimensional steel skeleton is automatically generated using a data-driven generative intelligent algorithm. Step S2: Fabricate and install replaceable energy dissipation nodes. Specifically, fabricate a steel-concrete composite section based on the optimal topology generated in Step S1, and install multiple energy dissipation units evenly arranged in the circumferential direction on the outer periphery of the bottom of the tower through detachable connectors. The energy dissipation components of each energy dissipation unit are designed to yield and dissipate energy first under the design earthquake, and their yield bearing capacity is lower than the elastic bearing capacity of the main body of the tower. Step S3: Rapid post-earthquake assessment and repair, specifically: After the earthquake, images of the outer surface of the tower are collected by drones, crack features are identified by machine vision, internal damage is inverted by digital twin model and the energy-consuming units that need to be replaced are determined, and a repair plan is generated to guide the replacement work.

7. The seismic design method according to claim 6, characterized in that, In step S1, the generative intelligent algorithm adopts a framework based on variational autoencoders, generative adversarial networks, or diffusion models. It generates new configurations that meet the constraints by extracting the latent features of existing high-performance topological configurations. The multi-objective optimization adopts the Pareto front optimization method to generate multiple non-dominated solutions for selection.

8. The seismic design method according to claim 6, characterized in that, Step S1 further includes: performing finite element verification on the generated optimal topology configuration; if the verification result does not meet the preset performance index, it is fed back to the generative intelligent algorithm for iterative optimization until it converges to the optimal solution.

9. The seismic design method according to claim 6, characterized in that, In step S2, each energy-consuming unit adopts a graded yielding design, including a first yielding stage and a second yielding stage. The yield bearing capacity of the first yielding stage is lower than that of the second yielding stage, forming multiple energy-consuming defense lines.

10. The seismic design method according to claim 6, characterized in that, In step S3, the use of machine vision to identify crack features specifically includes: automatically identifying the crack width, density, orientation, and distribution area on the UHPC surface using a trained deep convolutional neural network model; the damage inversion specifically involves: inputting the crack features and seismic records into the damage inversion module, and combining the finite element model or physical information neural network model to calculate the cumulative damage of the plastic strain inside the steel frame and the energy dissipation units. The formula for calculating the cumulative damage is: Where D is the cumulative damage index, and n is the total number of cycles for the accumulation of plastic strain under seismic loading. Let be the plastic strain increment for the i-th cycle. This represents the ultimate plastic strain of the material.

11. A system for implementing the seismic design method of claim 6, characterized in that, include: The intelligent topology optimization design system is used to perform step S1, including a parametric modeling module, a multi-objective optimization solver, and an iterative verification module; The parametric modeling module is used to establish the initial design domain of the steel-concrete composite section; The multi-objective optimization solver is used to automatically generate the topology of the steel frame with the objectives of maximizing structural stiffness, minimizing mass, and homogenizing stress distribution using a data-driven generative intelligent algorithm. The iterative verification module is used to perform finite element verification on the generated topology and feed the verification results back to the multi-objective optimization solver for iterative optimization. The post-earthquake rapid assessment system, used to perform step S3, includes an unmanned aerial vehicle (UAV)-borne visual acquisition module, a machine vision recognition module, a damage inversion module, and a repair decision module. The UAV-borne visual acquisition module is used for automatic cruise photography of the outer surface of the tower after an earthquake. The machine vision recognition module has a built-in trained crack recognition model for automatically identifying crack features on the UHPC surface. The damage inversion module is used to invert the degree of plastic damage inside the steel frame based on the surface crack characteristics and the digital twin model, and to determine whether each energy-consuming unit needs to be replaced. The repair decision module is used to generate a visual damage report and a replacement list.

12. The system according to claim 11, characterized in that, The post-earthquake rapid assessment system also includes a health data fusion module, which is used to compare and analyze the post-earthquake assessment results with the pre-earthquake health monitoring data to generate a report on the trend of structural performance degradation.

13. The system according to claim 11, characterized in that, It also includes a cloud-based collaborative platform, which receives and stores monitoring data, topology optimization results, and post-earthquake assessment reports from each tower, and updates machine learning models based on data from multiple towers to optimize the generation capabilities of generative intelligent algorithms and the recognition accuracy of crack identification models.

Images