A method and system for configuring AI-supporting facilities based on urban renewal

By assessing the basic physical parameters and industrial planning parameters of urban renewal areas, classifying business categories and establishing a mapping matrix, hierarchical decomposition of computing power requirements and facility configuration are carried out. This addresses the shortcomings of facility planning in existing technologies, achieves precise adaptation and effective resource utilization, and adapts to the rapid development of artificial intelligence businesses.

CN123117529APending Publication Date: 2026-11-17HUAXIN CONSULTATING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611279387.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing planning of artificial intelligence supporting facilities in urban renewal areas lacks quantitative calculation methods, resulting in over- or under-configuration of facilities. The facilities are not configured in a hierarchical manner according to latency sensitivity, existing resources are not fully utilized, and the planning lacks flexibility and closed-loop feedback, making it difficult to adapt to the explosive growth and technological iteration of artificial intelligence businesses.

Method used

By assessing the basic physical parameters and industrial planning parameters of the target area, classifying business categories and establishing a mapping matrix, hierarchical decomposition of computing power requirements and facility configuration are carried out. Combined with existing resource data, precise configuration is achieved to realize the hierarchical adaptation of facilities and the coordinated utilization of resources.

Benefits of technology

It enables precise allocation of computing power needs, avoids resource waste, ensures end-to-end latency requirements for low-latency services, promotes the reuse of existing resources and the coordinated configuration of facilities, and adapts to the rapid development of artificial intelligence services.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117529A_ABST
    Figure CN123117529A_ABST
Patent Text Reader

Abstract

This application relates to a method and system for configuring artificial intelligence (AI) supporting facilities based on urban renewal. The method includes: assessing the transformation potential of a target area based on its basic physical parameters to obtain existing resource data; classifying AI businesses in the target area into multiple business categories based on industrial planning parameters and latency sensitivity, and establishing a mapping matrix between business categories and computing power requirements; obtaining the total computing power requirements of the target area, and based on the mapping matrix, hierarchically decomposing the total computing power requirements to obtain tiered computing power requirements corresponding to different business categories; and configuring AI supporting facilities based on the tiered computing power requirements and existing resource data to obtain configuration data for guiding the physical AI supporting facilities. This addresses the problem of improving the flexibility of AI supporting facility planning in related technologies.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of artificial intelligence facility planning technology, and in particular to a method and system for configuring artificial intelligence supporting facilities based on urban renewal. Background Technology

[0002] As urban renewal initiatives and the city's overall digital transformation continue to advance, various artificial intelligence (AI) businesses, such as large-scale model development, intelligent manufacturing, low-altitude economy, and vehicle-road collaboration, are gradually being implemented in existing urban renewal areas. The operation of AI applications heavily relies on AI-supporting facilities such as edge computing space, communication equipment rooms, wireless base stations, communication pipelines, and supporting energy channels. Currently, facility planning in urban renewal areas largely follows the traditional approach of control planning and municipal special planning, primarily relying on qualitative control and per capita and per unit area experience indicators to reserve and configure digital infrastructure.

[0003] The existing planning for artificial intelligence supporting facilities in urban renewal areas mainly faces the following technical problems: First, there is a lack of quantitative measurement methods for computing power demand for artificial intelligence businesses, which can easily lead to over-configuration or under-supply of facilities. Second, computing power facilities generally adopt a centralized construction model without distinguishing differentiated needs. Third, the existing resources such as idle properties, power distribution rooms, and weak current rooms in the area are not fully coupled with the computing power demand brought about by industrial introduction. Fourth, it is difficult to adapt to the development trend of explosive growth of artificial intelligence businesses and iterative evolution of network technology.

[0004] Currently, no effective solution has been proposed in the relevant technologies for improving the flexibility of AI-based infrastructure planning in urban renewal. Summary of the Invention

[0005] This application provides a method and system for configuring artificial intelligence supporting facilities based on urban renewal, so as to at least solve the problem of how to improve the flexibility of planning artificial intelligence supporting facilities in related technologies.

[0006] In a first aspect, embodiments of this application provide a method for configuring artificial intelligence-based supporting facilities for urban renewal, the method comprising: Based on the basic physical parameters of the target area, the transformation potential of the target area is assessed to obtain existing resource data; Based on industry planning parameters and latency sensitivity, the artificial intelligence business in the target area is divided into multiple business categories, and a mapping matrix between the business categories and computing power requirements is established. Obtain the total computing power requirement of the target area, and based on the mapping matrix, perform hierarchical decomposition of the total computing power requirement to obtain the hierarchical computing power requirements corresponding to different business categories; Based on the tiered computing power requirements and the existing resource data, artificial intelligence supporting facilities are configured to obtain configuration data for guiding the physical artificial intelligence supporting facilities.

[0007] In one embodiment, the step of assessing the redevelopment potential of the target area based on its basic physical parameters and industrial planning parameters to obtain existing resource data includes: Obtain the basic physical parameters of the candidate existing spaces in the target area. The basic physical parameters include the floor load, net height, ventilation and heat dissipation conditions and power supply conditions of the candidate existing spaces. Based on the aforementioned basic physical parameters, for each of the candidate existing spaces, the transformation and reuse coefficient of the candidate existing space is calculated. The transformation and reuse coefficient is used to characterize the feasibility of transforming the candidate existing space into a space for supporting artificial intelligence facilities. Candidate existing spaces with a reuse coefficient greater than or equal to a preset threshold are identified as reuse and modification spaces. The total amount of reused space is obtained by counting all the reused and modified spaces, and the total amount of reused space is used as the stock resource data.

[0008] In one embodiment, the step of dividing the artificial intelligence business of the target area into multiple business categories based on industry planning parameters and latency sensitivity, and establishing a mapping matrix between the business categories and computing power requirements, includes: Based on the industrial planning parameters, the end-to-end latency requirements for various artificial intelligence application formats in the target area are determined. The end-to-end latency requirements are the time budget for business data to complete the link transmission and processing from the collection end to the processing end. Based on the end-to-end latency requirements, the artificial intelligence services in the target area are divided into multiple service categories according to latency sensitivity; Based on the latency sensitivity of each business category, the edge traffic splitting coefficient and the regional traffic splitting coefficient corresponding to each business category are determined. The edge traffic splitting coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the edge computing power nodes of the region. The regional traffic splitting coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the regional computing power nodes. A mapping matrix is ​​constructed based on the business category, the edge diversion coefficient, and the regional diversion coefficient.

[0009] In one embodiment, obtaining the total computing power requirement of the target area, based on the mapping matrix, involves hierarchically decomposing the total computing power requirement to obtain tiered computing power requirements corresponding to different business categories, including: Based on the scale data of various businesses, unit computing power demand intensity, occupancy rate conversion factor and average annual growth rate of computing power demand in the industrial planning parameters, calculate the total computing power demand of the target area within the planning period. Based on the edge-level and regional-level traffic splitting coefficients in the mapping matrix, the total computing power requirement is decomposed into edge-level computing power requirement, regional-level computing power requirement, and hub-level computing power requirement, which are referred to as the hierarchical computing power requirement.

[0010] In one embodiment, the total computing power requirement is expressed by the following formula:

[0011] In the formula, Q For the area t Total annual computing power requirement, in TOPS; K Number of industry categories; Ik For the first k A collection of business formats related to the same industry; Sk,i For the first k Category of industries i The planned building area or number of enterprises for each business type; ρk,i To determine the unit computing power demand intensity for corresponding business formats, the unit TOPS / 10,000 m² or TOPS / enterprise is calibrated through visits and surveys of typical industry enterprises, questionnaires, and computing power leasing market data. λk,i This refers to the occupancy rate or the conversion factor for reaching full production capacity. gk For the first k Average annual growth rate of computing power demand in related industries; t For the planning period; The edge-level computing power requirement is expressed by the following formula:

[0012] The regional computing power requirement is expressed by the following formula:

[0013] The hub-level computing power requirement is expressed by the following formula:

[0014] In the formula, in the formula: Qk For the first k The computing power needs of related industries; 、 、 These are the computing power requirements at the edge of the region, the regional level, and the city (hub) level, respectively; among them, It will be uniformly supported by a city-level computing power hub, with only high-speed interconnection communication channels reserved in the area; It is undertaken by regional computing power rooms and aggregation communication rooms; It is undertaken by edge computing nodes in the region. K Number of industry categories; The edge split coefficient; This is the regional diversion coefficient.

[0015] In one embodiment, the step of configuring AI supporting facilities based on the tiered computing power requirements and the existing resource data to obtain configuration data for guiding the physical AI supporting facilities includes: Based on the edge computing power requirements in the tiered computing power requirements, determine the physical deployment number of edge computing power nodes and the deployment location of each edge computing power node; Based on the aforementioned edge computing power requirements, the parameters of the computer room building area, the mains power supply capacity, and the communication pipeline bore capacity are calculated. Based on the existing resource data, the computer room building area parameters, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters, the supply and demand gap of artificial intelligence supporting facilities is determined; Based on the supply and demand gap, data is generated to guide the configuration of supporting facilities for physical artificial intelligence.

[0016] In one embodiment, determining the physical deployment number of edge computing nodes and the deployment location of each edge computing node based on the edge-level computing power demand in the hierarchical computing power demand includes: The maximum service radius of the edge computing node is determined based on the end-to-end latency requirements of the business corresponding to the edge computing power demand, the edge node processing latency, and the queuing and access latency. The service radius of the edge computing node is taken as a value not exceeding the maximum service radius, and a first number that satisfies the coverage area constraint is determined based on the service radius and the area of ​​the target region. Based on the edge-level computing power requirements and the preset standard computing power configuration for a single node, a second quantity that satisfies the computing power capacity constraint is determined; Select the larger quantity from the first quantity and the second quantity as the physical deployment quantity; A corresponding number of candidate locations are defined in the target area. A site selection suitability index is calculated for each candidate location. Based on the site selection suitability index from high to low, the physical deployment number of locations is selected from the candidate locations as the deployment locations of the edge computing nodes.

[0017] In one embodiment, the step of determining the supply-demand gap for artificial intelligence supporting facilities based on the existing resource data, the building area parameters of the computer room, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters; and generating configuration data to guide the physical artificial intelligence supporting facilities based on the supply-demand gap, including: The final facility requirements are determined based on the building area of ​​the computer room, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters. Based on the existing resource data, the renovation and reuse coefficient, and the final facility demand scale, the supply and demand gap is determined; When the supply-demand gap is less than or equal to zero, a configuration decision is generated to reuse existing space; when the supply-demand gap is greater than zero, a configuration decision is generated to allocate facility resources according to a preset priority order to meet the supply-demand gap.

[0018] In one embodiment, after configuring AI supporting facilities based on the tiered computing power requirements and the existing resource data to obtain configuration data for guiding the physical AI supporting facilities, the method further includes: Configure the artificial intelligence supporting facilities based on the configuration data, and collect the operation status monitoring data of the artificial intelligence supporting facilities; The configuration data is iteratively updated based on the operational status monitoring data.

[0019] Secondly, embodiments of this application provide an AI-based infrastructure configuration system for urban renewal. This system is used to execute the aforementioned method and includes: a stock resource data module, a mapping matrix module, a tiered computing power demand module, and a configuration module; wherein: The existing resource data module is used to assess the transformation potential of the target area based on the basic physical parameters of the target area, and obtain existing resource data. The existing resource data module is used to divide the artificial intelligence business of the target area into multiple business categories according to industry planning parameters and latency sensitivity, and to establish a mapping matrix between the business categories and computing power requirements. The hierarchical computing power requirement module is used to obtain the total computing power requirement of the target area, and based on the mapping matrix, to perform hierarchical decomposition of the total computing power requirement to obtain the hierarchical computing power requirements corresponding to different business categories. The configuration module is used to configure artificial intelligence supporting facilities based on the hierarchical computing power requirements and the existing resource data, and obtain configuration data for guiding the physical artificial intelligence supporting facilities.

[0020] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for configuring artificial intelligence supporting facilities based on urban renewal as described in the first aspect above.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for configuring artificial intelligence supporting facilities based on urban renewal as described in the first aspect above.

[0022] The method and system for configuring AI-supporting facilities based on urban renewal provided in this application embodiment have at least the following technical effects.

[0023] This invention first assesses the transformation potential of the target area based on its basic physical parameters, providing a foundation of existing resource data for subsequent facility configuration. This clarifies the carrying capacity of existing space. Simultaneously, it establishes a mapping matrix between business categories and computing power requirements based on industry planning parameters and latency sensitivity, providing a classification basis for the differentiated decomposition of computing power needs. Then, based on this mapping matrix, it obtains the total computing power demand and performs hierarchical decomposition, enabling the total computing power demand to be accurately distributed to three levels—edge, regional, and hub—according to latency sensitivity. This avoids resource waste or insufficient capacity caused by centralized planning at a single level, while ensuring end-to-end latency requirements for low-latency services. Finally, it combines existing resource data with tiered computing power demands for facility configuration. On the one hand, it eliminates the disconnect between new construction demand and existing idle space by prioritizing the reuse of existing space; on the other hand, it achieves coordinated matching of civil engineering space, energy supply, and communication pipelines by simultaneously inputting multiple configuration parameters. Thus, overall, it achieves the technical effects of data-driven computing power demand, precise hierarchical matching of facilities, economical reuse of existing resources, and coordinated configuration of supporting facilities. This addresses the issue of how to improve the flexibility of planning supporting infrastructure for artificial intelligence in related technologies.

[0024] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0025] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating an artificial intelligence-based infrastructure configuration method for urban renewal, according to an exemplary embodiment. Figure 2This is a structural block diagram illustrating the hierarchical decomposition of computing power requirements according to an exemplary embodiment; Figure 3 This is a schematic diagram illustrating the division and location layout of edge computing service areas according to an exemplary embodiment; Figure 4 This is a structural block diagram of an electronic device according to an exemplary embodiment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0027] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0028] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0029] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0030] The existing planning for artificial intelligence supporting facilities in urban renewal areas mainly faces the following technical problems: (1) Demand assessment lacks quantitative models. Traditional control detailed planning and municipal special planning mainly rely on qualitative control and per capita / land per capita indicators. There are no quantitative assessment methods for computing power demand in the artificial intelligence industry. The reserved scale of computer room space, communication pipeline capacity and power capacity is not based on data, resulting in facilities that are "not enough" or "cannot be used".

[0031] (2) Computing facilities are not configured according to latency sensitivity. Existing computing infrastructure is mostly built in a single-level centralized manner, without distinguishing between non-real-time ultra-large computing power services and millisecond-level real-time services. It cannot meet the ultra-large computing power requirements such as large model training and the real-time service requirements of end-to-end latency within 10ms, such as industrial, low-altitude, and vehicle-road cooperative systems.

[0032] (3) Existing resources are out of sync with industrial demand. Urban renewal areas generally have existing resources such as idle property rooms, power distribution rooms, weak current rooms, remaining pipe holes, and remaining loads of substations. However, existing planning methods lack coupling analysis methods for the baseline data of existing resources and the computing power demand for industrial introduction, resulting in a disconnect between facility configuration and the industrial positioning of the area.

[0033] (4) The planning and construction lacks flexibility and closed loop. Once approved, the existing plans are fixed and do not take into account the explosive growth of artificial intelligence applications, the sinking of edge computing power, the evolution of 10 Gigabit optical networks and 5G-A / 6G networks. They lack a dynamic closed loop mechanism of "demand assessment - facility configuration - construction and acceptance - long-term operation - feedback and correction".

[0034] Based on the above, this application provides a quantitative, hierarchical, flexible, and closed-loop feedback-enabled method and system for configuring AI-supporting facilities for urban renewal.

[0035] In this document, it should be understood that the terms used may be technical means used to implement part of the present invention or other summary technical terms. For example, the terms may include: Basic physical parameters: refer to the physical space and facility conditions of the target area, including but not limited to the area area, building area of ​​candidate existing space, floor load, net height, ventilation and heat dissipation conditions, power supply conditions, as well as the remaining pipe capacity and remaining pipe routing directions of existing communication pipelines.

[0036] Existing resource data: refers to the result data obtained after assessing the physical carrying capacity of existing infrastructure in the target area, specifically the total amount of existing reusable space, that is, the sum of the scale of all candidate existing spaces with a reusability coefficient greater than or equal to a preset threshold.

[0037] Industrial planning parameters: These refer to planning data that reflect the direction and scale of industrial development in the target area, including but not limited to the planned building area, number of enterprises, and end-to-end latency requirements of various artificial intelligence application industries in the area's industrial development plan.

[0038] Latency sensitivity: refers to the degree to which artificial intelligence services are sensitive to the end-to-end data transmission and processing completion time, determined by the end-to-end latency requirements of the service. The higher the end-to-end latency requirement (the smaller the value), the higher the latency sensitivity. In this application, services are divided into three levels based on latency sensitivity: high latency sensitivity (end-to-end latency requirement ≤ 10ms), medium latency sensitivity (10ms~50ms), and low latency sensitivity (≥ 50ms).

[0039] Mapping matrix: refers to a data structure that records the mapping relationship between business categories and their corresponding edge diversion coefficients and regional diversion coefficients. Its form can be: "Industry category - edge diversion coefficient - regional diversion coefficient" correspondence.

[0040] Net height: refers to the vertical clearance between the finished floor surface of the existing space and the bottom surface of the upper floor slab or the bottom surface of the lower chord of the roof truss, in meters (m).

[0041] Reuse Coefficient: This refers to the weighted sum of the compliance scores and corresponding weights of four renovation conditions (floor load, net height, ventilation and heat dissipation, and power supply) for each candidate existing space. The value ranges from 0 to 1. This coefficient is used to quantitatively assess the feasibility of renovating candidate existing spaces into spaces that support artificial intelligence facilities.

[0042] End-to-end latency requirement: refers to the maximum allowable time budget for the entire complete link of artificial intelligence business data, from its collection end (such as sensors, cameras, terminal devices), through the communication network to the computing power node for processing, and then back to the execution end or user end, in milliseconds (ms).

[0043] Edge offloading coefficient: This refers to the proportion of the total computing power demand for the k-th type of service that should be handled by edge-level computing nodes in the target area. The coefficient ranges from 0 to 1 and is determined by the end-to-end latency requirements of the service type; the stricter the latency requirements, the higher the edge offloading coefficient. This coefficient is recorded in a mapping matrix and is used to hierarchically decompose the total computing power demand to obtain the edge-level computing power demand.

[0044] Regional load balancing coefficient: This refers to the proportion of the total computing power demand for the k-th type of service that should be handled by regional-level computing power nodes. It is used to hierarchically decompose the total computing power demand to obtain the regional-level computing power demand.

[0045] Unit computing power demand intensity: refers to the amount of computing power resources required per unit scale of the i-th business type in the k-th business category, measured in TOPS / 10,000 m² or TOPS / enterprise.

[0046] Occupancy rate conversion factor: This refers to the conversion ratio of the actual occupancy or production capacity of the i-th business type within the k-th business category to the full occupancy / full production status within the planning period. The value ranges from 0 to 1. This factor is used to reflect the time characteristics of the phased introduction of industries into the area, avoiding overestimation of capacity demand based on a one-time full occupancy status.

[0047] Queuing and access latency: refers to the latency of service data in the communication network caused by data packets queuing and waiting for forwarding, and the sum of channel contention and time-frequency resource scheduling latency experienced by terminal devices when accessing the communication network. It is used to constrain the maximum service radius of edge computing nodes.

[0048] Firstly, embodiments of this application provide a method for configuring AI-supporting facilities based on urban renewal. Figure 1 This is a flowchart of a method for configuring AI-powered supporting facilities based on urban renewal, such as... Figure 1 As shown, the method includes: Step S101: Based on the basic physical parameters of the target area, assess the redevelopment potential of the target area to obtain existing resource data. This specifically includes the following steps: Step S1011: Obtain the basic physical parameters of the candidate existing spaces in the target area. The basic physical parameters include the floor load, net height, ventilation and heat dissipation conditions, and power supply conditions of the candidate existing spaces.

[0049] Optionally, the content collected includes: (1) Existing idle space: survey the area of ​​idle sites that can be transformed, and record the site area, floor load, floor height, load-bearing capacity, ventilation, power supply and other transformation restrictions; (2) Communication facilities: check the remaining pipe hole capacity of the main and branch pipelines, and identify the wireless signal gap and the reserved computing power line; (3) Energy consumption: check the remaining load of substations and power distribution rooms and the remaining indicators of energy consumption control.

[0050] Step S1012: Based on the basic physical parameters, calculate the transformation and reuse coefficient of each candidate existing space. The transformation and reuse coefficient is used to characterize the feasibility of transforming the candidate existing space into a space for supporting artificial intelligence facilities.

[0051] Optionally, for each candidate existing space j, its renovation and reuse coefficient is evaluated using the following formula: (1) In the formula: κ r,j ω is the modifiability and reusability coefficient of candidate stock space j, taking a value from 0 to 1; m y represents the weight of the m-th modification condition; j,m The compliance score for the m-th renovation condition of candidate existing space j is calculated, where m corresponds to four items: floor load, net floor height, ventilation and heat dissipation, and power supply conditions. A score of 1 is assigned to meet the standard, 0.5 is assigned to meet the standard after renovation, and 0 is assigned to fail to meet the standard. r,j Candidate spaces with a value ≥0.6 are identified as reusable and remodelable spaces, and are included in the total stock of reusable spaces S. 存量 .

[0052] Step S1013: Determine the candidate stock space with a renovation and reuse coefficient greater than or equal to a preset threshold as the reuse and renovation space.

[0053] Optionally, when κ r,j Candidate spaces with a threshold value greater than or equal to a preset threshold (which may be 0.6 in this application) are identified as reusable and remodelable spaces and included in the total stock of reusable spaces, S.

[0054] Step S1014: Count all reused and remodeled spaces to obtain the total amount of reused space, and use the total amount of reused space as stock resource data.

[0055] Steps S1011-S1014 transform the potential for renovation of existing space into a quantifiable reuse coefficient, which solves the problem that existing planning methods lack the technical means to quantitatively assess idle existing space. This provides a quantitative basis for whether existing resources are reusable, effectively avoiding the resource mismatch problem of blind new construction or coexistence of idle existing space due to the lack of data on the carrying capacity of existing space.

[0056] Step S102: Based on industry planning parameters and latency sensitivity, divide the artificial intelligence business in the target area into multiple business categories and establish a mapping matrix between business categories and computing power requirements. This specifically includes the following steps: Step S1021: Based on the industrial planning parameters, determine the end-to-end latency requirements for various artificial intelligence application formats in the target area. The end-to-end latency requirements are the time budget for the transmission and processing of business data from the collection end to the processing end.

[0057] Step S1022: Based on end-to-end latency requirements, divide the artificial intelligence business in the target area into multiple business categories according to latency sensitivity.

[0058] Step S1023: Based on the latency sensitivity of each business category, determine the edge diversion coefficient and regional diversion coefficient corresponding to each business category. The edge diversion coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the edge computing power nodes of the region; the regional diversion coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the regional computing power nodes.

[0059] Step S1024: Construct a mapping matrix based on business category, edge diversion coefficient, and regional diversion coefficient.

[0060] Optionally, based on the typical business formats of existing enterprises in the target area, the end-to-end latency requirements for various artificial intelligence application formats in the target area are determined. For example, if the typical business format is large-scale model development, the business latency characteristics (end-to-end latency requirements) are determined, which are non-real-time (hourly / daily level). Based on the end-to-end latency requirements, the artificial intelligence businesses in the target area are divided into multiple business categories according to latency sensitivity. This application can be divided into three categories. The edge diversion coefficient and regional diversion coefficient are determined based on the business end-to-end latency requirements. Specifically, for businesses with latency requirements within 10ms, φ is 0.8–1.0; for businesses with latency requirements between 10 and 50ms, φ is 0.3–0.8; and for businesses with latency requirements above 50ms, φ is 0–0.3, and the φ requirement is satisfied. k +ψ k ≤1. Edge split coefficient φ k This represents the proportion of computing power demand from industry category k that should be handled by edge computing nodes within the region; the regional diversion coefficient ψ. kThis indicates the proportion that should be handled by regional-level computing power nodes; the remaining portion is handled by city (hub)-level computing power nodes. A mapping matrix is ​​constructed based on business category, edge diversion coefficient, and regional diversion coefficient. Table 1: Industry Computing Power Demand Classification and Diversion Coefficient Table, as shown in Table 1.

[0061] Table 1: Industry Computing Power Demand Classification and Diversion Coefficient Table

[0062] Steps S1021-S1024 determine the end-to-end latency requirements of various AI application business models based on industry planning parameters, and classify the business into categories with different latency sensitivities based on these latency requirements. Then, for each category, the proportion coefficient of computing power demand allocated to edge nodes and regional nodes is determined, thereby constructing a mapping matrix between business categories and diversion coefficients. This provides a structured classification basis for the subsequent hierarchical decomposition of total computing power demand. This solves the problem of existing computing power infrastructure being centrally planned at a single level and failing to distinguish between non-real-time ultra-large computing power business and millisecond-level real-time business. This allows ultra-large computing power non-real-time business such as large model training to rely on city-level hubs, while real-time business with end-to-end latency requirements of less than 10ms, such as vehicle-road collaboration and low-altitude economy, can be processed by edge nodes nearby. This achieves hierarchical and precise adaptation of computing power resources, ensuring the real-time requirements of low-latency business while avoiding the resource waste caused by sinking all computing power demand to edge nodes.

[0063] Step S103: Obtain the total computing power requirement of the target area. Based on the mapping matrix, decompose the total computing power requirement hierarchically to obtain the hierarchical computing power requirements corresponding to different business categories. This specifically includes the following steps: Step S1031: Based on the scale data of various businesses, unit computing power demand intensity, occupancy rate conversion factor and average annual growth rate of computing power demand in the industrial planning parameters, calculate the total computing power demand of the target area within the planning period. Optionally, the total computing power requirement is expressed by the following formula: (2) In the formula, Q For the area t Total annual computing power requirement, in TOPS; K In this application, K=3, representing the number of industry categories. Ik For the first k A collection of business formats related to the same industry; Sk,i For the first k Category of industries i The planned building area (ten thousand m²) or number of enterprises for each business type; ρk,iTo determine the unit computing power demand intensity for corresponding business formats, the unit TOPS / 10,000 m² or TOPS / enterprise is calibrated through visits and surveys of typical industry enterprises, questionnaires, and computing power leasing market data. λk,i This is the occupancy rate or production capacity conversion factor, with a value ranging from 0 to 1. gk For the first k Average annual growth rate of computing power demand in related industries; t The planning period is the number of years.

[0064] Step S1032: Based on the edge distribution coefficient and regional distribution coefficient in the mapping matrix, the total computing power requirement is decomposed into edge-level computing power requirement, regional-level computing power requirement and hub-level computing power requirement, as hierarchical computing power requirement.

[0065] Optionally, the edge computing power requirement is expressed by the following formula: (3a) Regional computing power requirements are expressed by the following formula: (3b) Hub-level computing power requirements are expressed by the following formula: (3c) In the formula, in the formula: Qk For the first k The computing power requirements of the industry category are obtained by breaking them down by industry category using formula (2); 、 、 These are the computing power requirements at the edge of the region, the regional level, and the city (hub) level, respectively; among them, It will be uniformly supported by a city-level computing power hub, with only high-speed interconnection communication channels reserved in the area; It is undertaken by regional computing power rooms and aggregation communication rooms; It is undertaken by edge computing nodes in the region. K Number of industry categories; The edge split coefficient; This is the regional diversion coefficient.

[0066] Figure 2 This is a structural block diagram illustrating the hierarchical decomposition of computing power requirements according to an exemplary embodiment, such as... Figure 2As shown, steps S1031-S1032 first calculate the total computing power demand of the target area within the planning period based on the scale data of various businesses, unit computing power demand intensity, occupancy rate conversion factor, and average annual growth rate of computing power demand in the industrial planning parameters. This transforms the industrial positioning of the area into quantifiable computing power demand indicators, solving the problem that traditional planning relies mainly on qualitative control and per capita indicators, and there is no basis for reserving computing power scale. Then, based on the edge diversion coefficient and regional diversion coefficient in the mapping matrix, the total computing power demand is decomposed into three levels of computing power demand according to latency sensitivity: edge level, regional level, and hub level. This ensures that non-real-time ultra-large computing power businesses such as large model training are carried by city-level hubs, real-time businesses with end-to-end latency requirements of less than 10ms are processed by edge nodes nearby, and near-real-time businesses are handled by regional-level data centers, thus guaranteeing the end-to-end latency requirements of various businesses.

[0067] Step S104: Based on the hierarchical computing power requirements and existing resource data, configure artificial intelligence supporting facilities to obtain configuration data for guiding the physical artificial intelligence supporting facilities.

[0068] It should be noted that the embodiments of this application are geared towards the actual application scenario of urban renewal areas, in which edge low-latency real-time services (i.e., industry category III in Table 1) such as vehicle-road collaboration, low-altitude economy, urban governance, and intelligent manufacturing are the main focus. Therefore, in the facility configuration stage of step S104, edge-level computing power requirements are used as the main calculation inputs for configuration parameters such as the number of physical deployments of edge computing nodes, the building area of ​​the data center, the capacity of the mains power supply, and the capacity of communication pipelines. For areas where industry categories are dominated by category I or II, the configuration logic can be adjusted adaptively according to the actual scenario. The configuration method driven by edge-level computing power requirements given in the embodiments of this application constitutes the preferred implementation scheme for the specific scenario of urban renewal areas, and does not exclude adaptive configuration based on the three-level distribution system for scenarios dominated by other industry categories.

[0069] Step S104 specifically includes the following steps: Step S1041: Based on the edge-level computing power requirements in the tiered computing power demand, determine the physical deployment quantity of edge computing power nodes and the deployment location of each edge computing power node. Specifically, this includes: Step 1: Determine the maximum service radius of the edge computing nodes based on the end-to-end latency requirements of the business corresponding to the edge computing power demand, the edge node processing latency, and the queuing and access latency.

[0070] Optionally, to ensure low latency, the target area is divided into several edge computing service areas. The service radius *r* of a single service area can range from 500m to 1000m, while simultaneously satisfying coverage and latency constraints. Based on the end-to-end latency requirements of the services corresponding to the edge-level computing power needs, the processing latency of edge nodes, and the queuing and access latency, the maximum service radius of the edge computing nodes is determined. Under latency constraints, the maximum service radius is calculated using the following formula: (4) In the formula: Tb is the end-to-end latency budget for the service, and 10ms is taken for real-time edge services; Tp is the processing latency of the edge node; T q Let be the queuing and access delay; v be the optical signal transmission rate in the optical cable, approximately 5 μs / km. In engineering implementation, r ∈ [500, 1000]m and r ≤ r max The service radius is set at the lower limit in high-density population and industry areas and at the upper limit in low-density areas.

[0071] Step 2: Use a value that does not exceed the maximum service radius as the service radius of the edge computing node, and determine the first number that satisfies the coverage area constraint based on the service radius and the area of ​​the target region.

[0072] Step 3: Based on the edge-level computing power requirements and the preset standard computing power configuration for a single node, determine the second quantity that meets the computing power capacity constraints.

[0073] Step 4: Select the larger quantity from the first and second quantities as the physical deployment quantity.

[0074] Optionally, the number of edge computing nodes is determined by both coverage and capacity constraints, taking the larger of the two values: (5) In the formula: N is the number of edge computing nodes; A is the area of ​​the region; r is the radius; κ is the coverage overlap and shape reduction coefficient, with a value of 0.6 to 0.8; Q0 is the standard computing power configuration of a single node, which is not less than 1600 TOPS, and can be expanded to have a computing power greater than 5000 TOPS, with an uplink bandwidth of not less than 25Gbps. To meet edge computing power requirements, Step 5: Delineate a corresponding number of candidate locations in the target area, calculate the site suitability index for each candidate location, and select the physical deployment locations from the candidate locations according to the order of the site suitability index from high to low as the deployment locations of the edge computing nodes.

[0075] Optionally, Figure 3 This is a schematic diagram illustrating the division and location layout of edge computing service areas according to an exemplary embodiment, such as... Figure 3As shown, edge computing nodes should preferably be located in the central area of ​​the computing service zone. A site selection suitability index is established for candidate site j: (6) In the formula: L j The site selection suitability index is denoted by ; M is the number of evaluation factors; w m The weight of the m-th evaluation factor is determined using the analytic hierarchy process (AHP) or the entropy weight method; x j,m The normalized score for the m-th factor of candidate point j is given. The evaluation factors and reference weights are shown in Table 2.

[0076] Table 2: Evaluation Factors and Reference Weights for Edge Computing Node Location

[0077] Simultaneously, a set of veto criteria is set: candidate locations located within the influence range of gas stations, concentrated catering areas, chemical storage yards, areas prone to flooding and landslides, strong electromagnetic interference sources, and strong vibration sources are directly vetoed and do not participate in the index ranking. Among the candidate locations that pass the veto criteria screening, they are ranked by L. j The top N points, from largest to smallest, are selected as the locations for edge computing nodes.

[0078] Step S1042: Based on edge computing power requirements, calculate the building area parameters of the data center, the mains power supply capacity parameters, and the communication duct capacity parameters. Specifically, this includes: First, the calculation of the building area of ​​the computer room. Edge computing space (high-performance computing room) includes the main server room area and supporting auxiliary areas, and its total building area is calculated using the following formula: (7) In the formula: The total building area of ​​the edge computing space; μ is the building area index per unit computing power, calibrated to 0.03–0.06 m² / TOPS according to standard computing power configuration; θ is the area ratio of the supporting auxiliary area, ranging from 0.2 to 0.3; η is the unified expansion margin coefficient, taken as 30%, used to adapt to long-term industrial expansion. The area of ​​a single edge computing space should be controlled between 50 m² and 100 m², with a square layout, a net height of not less than 3 m, a floor live load of not less than 6 kN / m², and a dedicated equipment handling passage.

[0079] Second, power capacity calculation The mains power supply capacity for edge computing space and data centers at all levels is calculated using the following formula and implemented in one go according to final requirements: (8) In the formula: P is the mains power supply capacity, in kW; ε is the computing power efficiency, i.e., the computing power that can be provided by a unit of electricity, in TOPS / kW; PUE is the power utilization efficiency, with a value of 1.3 to 1.5 for edge data centers; γ is the power supply redundancy coefficient, with a value of 0.2 to 0.3 for N+1 redundancy configuration. The power supply for a single edge computing space is approximately 20kW to 200kW. The power supply should preferably be configured with dual power sources, or one mains power source and one backup power source (diesel generator, energy storage cabinet, etc.). Initially, it can be connected to the nearest 220V / 380V municipal power supply, and later, 10kV mains power can be introduced as computing power is expanded.

[0080] Third, communication pipeline capacity calculation The capacity of the communication duct orifice is calculated using the following formula and meets the final requirements: (9) In the formula: C is the pipe orifice capacity; The required number of tube holes for the core / convergence layer optical cable; The required number of conduit holes for connecting to the main optical cable; The required number of conduit holes for connecting wiring / terminal optical cables; The number of conduits is determined by the needs of information technology / smart cities; δ is the redundancy coefficient, ranging from 10% to 20%. Standard conduits have an inner diameter of no less than 90mm and an outer diameter of no less than 100mm, and multiple sub-conduits can be installed within them. Communication pipelines are configured according to road grade: backbone pipelines have 8-12 conduits, main pipelines have 18-32 conduits, secondary pipelines have 12-16 conduits, and general pipelines have 6-8 conduits. Outgoing pipelines from regional and area-level equipment rooms should have at least two directions with 12-18 conduits per direction; outgoing pipelines from unit equipment rooms and edge computing nodes should have at least two directions with 6-12 conduits per direction. Backbone and main pipelines should form a ring network and be interconnected. They should be constructed on both sides of roads with a right-of-way width greater than 40m, preferably under sidewalks, and laid within integrated utility tunnels where available.

[0081] Step S1043: Based on existing resource data, data center building area parameters, mains power supply capacity parameters, and communication pipeline capacity parameters, determine the supply and demand gap for artificial intelligence supporting facilities. Specifically, this includes: The final facility requirements are determined based on the building area of ​​the reference computer room, the mains power supply capacity parameters, and the communication pipeline borehole capacity parameters. The supply-demand gap is determined based on existing resource data, renovation and reuse coefficients, and the scale of final facility demand.

[0082] Optionally, based on existing reusable space and considering new industrial demand in the near, medium, and long term, the supply-demand gap can be calculated using the following formula: (10) In the formula: G represents the supply-demand gap; To determine the final facility demand, calculate the capacity using the parameters of the last year of the plan according to formulas (2) to (9) and reserve a 30% capacity expansion margin. κ represents the total amount of existing reusable space (or capacity); r The reusability coefficient of the existing space is determined by equation (1).

[0083] Step S1044: Based on the supply and demand gap, generate configuration data to guide the development of supporting facilities for physical artificial intelligence. Specifically, this includes: When the supply-demand gap is less than or equal to zero, a configuration decision is generated to reuse existing space; when the supply-demand gap is greater than zero, a configuration decision is generated to allocate facility resources according to a preset priority order to meet the supply-demand gap.

[0084] Optionally, the configuration decision rules are as follows: when the supply-demand gap G≤0, priority is given to functional replacement and adaptation of existing space, without new construction; when G>0, the gap is filled according to the preset priority order: "revitalizing idle space → leasing external shared computing power → phased construction", reducing the scale of new construction and lowering one-time investment. During phased construction, civil engineering space, pipeline capacity, and power capacity are constructed in one go according to 100% of the final demand, and the initial computing power equipment is deployed according to 10% to 30% of the final demand. For existing renewal areas, the focus is on filling gaps and upgrading, with communication pipelines laid simultaneously in conjunction with road micro-renovation; for new construction areas, high-standard configuration is carried out in one go according to the final demand, with unified planning and simultaneous construction.

[0085] It should be noted that the configuration decision generated in step S1044 covers four types of facilities: edge computing space, communication equipment room, wireless base station, and communication pipeline. These four types of facilities constitute a three-level collaborative system of "pipeline-equipment room-base station": the communication pipeline provides a physical channel for fiber optic cable laying, connecting the communication equipment room and the wireless base station; the communication equipment room serves as the core carrier for data aggregation and edge computing; and the wireless base station serves as a key node for data collection, transmission, and edge computing. The configuration standards for each type of facility are shown in Tables 3 to 5.

[0086] Table 3. Standards for Tiered Configuration of Communication Equipment Rooms (Co-construction and Sharing Model)

[0087] Table 4 Classification and Configuration Standards for Wireless Base Stations

[0088] Table 5. Standards for Classified Configuration of Communication Pipelines

[0089] The above-mentioned facility configuration follows the principles of foresight and co-construction and sharing: the channel capacity and space area are 100% reserved for future expansion needs; the communication equipment room, base station site, and communication pipeline are selected and constructed by the overall planning entity in a unified manner, giving priority to the sharing and reuse of multiple operators, and the building macro base station reserves antenna installation positions and feeder channels for no less than 3 operators.

[0090] Steps S1041-S1044 first determine the physical deployment quantity and location of nodes based on edge-level computing power requirements and latency constraints. Then, based on edge-level computing power requirements, they calculate parameters such as the building area of ​​the data center, the mains power supply capacity, and the communication pipeline capacity. Next, they combine existing resource data to determine the supply-demand gap. Finally, based on the supply-demand gap, they generate configuration data for reusing existing space or allocating facility resources. This solves the problem of low flexibility in the planning of AI supporting facilities and improves the flexibility of AI supporting facility planning. The configuration results ensure that the end-to-end latency of edge services is controlled within 10ms through the physical deployment quantity and location. Furthermore, the synchronized calculation of data center area, power capacity, and pipeline capacity, along with the unified reservation of future expansion capacity, achieves coordinated and one-time provisioning of civil engineering space, energy supply, and communication pipelines. Simultaneously, the supply-demand gap analysis prioritizes the reuse of existing space and reduces the scale of new construction.

[0091] Step S105: Configure the AI ​​supporting facilities based on the configuration data, and collect operational status monitoring data of the AI ​​supporting facilities; iteratively update the configuration data based on the operational status monitoring data. Specifically, this includes: Based on the configuration data, determine the computer room space parameters, pipeline routing parameters, and power capacity parameters. During the construction phase, implement pipeline construction and road engineering simultaneously according to the configuration data. Verify the physical indicators of the artificial intelligence supporting facilities item by item based on the configuration data. After completion and acceptance, collect operational status monitoring data of the artificial intelligence supporting facilities; Based on operational status monitoring data, the unit computing power demand intensity, occupancy rate conversion factor, and average annual growth rate of computing power demand for the target area are adjusted. Based on the revised unit computing power demand intensity, occupancy rate conversion factor, and average annual growth rate of computing power demand, the steps for obtaining the total computing power demand of the target area and subsequent steps (steps S101-S105) are re-executed to update the configuration data.

[0092] Optionally, the supporting facilities for artificial intelligence are planned, designed, constructed, and inspected simultaneously with the main project of the urban renewal project: during the planning stage, the facility configuration plan is incorporated into the area planning and implementation and urban health assessment system; during the design stage, the room space, pipeline routing, and power capacity are implemented in the construction drawings; during the construction stage, pipeline construction and road engineering are carried out simultaneously to avoid repeated road excavation; during the acceptance stage, indicators such as room area, net height, load, power supply redundancy, number of pipe holes, and outgoing direction are checked item by item, and the verification basis is the configuration decision generated in step S1044 and the configuration standards in Tables 3 to 5.

[0093] Based on the facility type and service attributes, one or a combination of four operating models should be selected: government-led, public-private partnership, entrusted operation, and joint construction and operation. For communication equipment rooms, base stations, and pipeline facilities shared by multiple operators, the overall coordinating entity will handle unified operation and maintenance, with the operation and maintenance costs shared among the users according to the following formula: (11) In the formula: F i The maintenance cost allocated to the i-th user; F d This represents the total cost of dedicated resource operation and maintenance; r i The resource quantity actually occupied by the i-th user is expressed in terms of rack count, pipe count, or antenna count; n is the number of users; F c The total cost of operation and maintenance for common areas (mains power supply, air conditioning, fire protection, etc.) shall be shared equally among all users.

[0094] At the same time, a dynamic evaluation and feedback mechanism is established: the facility operation status monitoring data, user satisfaction survey results and annual urban health check data are fed back to the stock resource data in step S101, and the unit computing power demand intensity ρ, occupancy rate conversion coefficient λ and annual growth rate g in formula (2) are corrected. Steps S102 to S105 are rolled out to update the facility configuration plan and form a planning closed loop of "evaluation - calculation - configuration - acceptance - operation - feedback".

[0095] The urban health check data feedback and parameter correction mechanism in step S105 enables the configuration scheme to be updated continuously as the industry is implemented and the scenario changes, thus achieving integrated and long-term operation.

[0096] The following examples further illustrate this application. This example takes an urban renewal area as an example. The area has a total area of ​​3.2 km², a planning period of t=5 years, and its leading industries are low-altitude economy (Category III), high-end intelligent manufacturing (Category II), and digital cultural and creative industries (Category I).

[0097] (I) Current Status Survey (Data Acquisition) Step S101 Data Acquisition: There are a total of 3 vacant properties in the area that can be renovated, with a total building area of ​​150m². The renovability and reuse coefficient κ is evaluated according to formula (1).r =0.7; 4 remaining pipe holes in the main pipeline; approximately 300kW of remaining load in the substation of the area.

[0098] (II) Industry Classification and Demand Forecasting The mapping matrix is ​​established in step S102, and the parameters of each industry are shown in Table 6.

[0099] Table 6: Calculation Table of Industry Parameters and Computing Power Requirements for Example 1

[0100] Calculate the flow according to formulas (3a) to (3c): Q 边缘 = 2575×0.9 + 1643×0.2 + 1480×0.05 ≈ 2317 + 329 + 74 = 2720TOPS; Q 区域 = 1643×0.5 + 1480×0.15 ≈ 821 + 222 = 1043 TOPS; Q 枢纽 = 1643×0.3 + 1480×0.8 ≈ 493 + 1184 = 1677 TOPS.

[0101] (III) Facility Configuration Calculation (1) Service area division: The service radius is taken as r = 750m (≤ r max ), the coverage reduction factor κ = 0.7. According to equation (5): N1 under the coverage constraint = 3.2 / (π×0.75²×0.7) = 2.59 =3; N2 = under capacity constraint 2720 / 1600 =2; therefore, N=max{3,2}=3, that is, 3 edge computing nodes are deployed to divide the area into 3 computing service areas.

[0102] (2) Site selection: Calculate the site suitability index for the 6 candidate sites that passed the veto screening according to formula (6), and take L. j The three highest locations include one of which is a renovation project of an idle property (co-located and co-built with the area's computer room, with shared power supply).

[0103] (3) Computer room area: Take μ=0.045m² / TOPS, θ=0.25, η=30%, according to formula (7): A 机房= 2720×0.045×1.25×1.3 ≈ 199m², with an average of about 66m² per node across the three nodes, meeting the configuration standard of 50m² to 100m² per node.

[0104] (4) Power capacity: Take ε=40TOPS / kW, PUE=1.4, γ=0.2, according to formula (8): P = (2720 / 40)×1.4×1.2 ≈ 114kW, the average of each of the 3 nodes is about 38kW, which is in the range of 20kW to 200kW per node. Dual-circuit power supply and N+1 redundancy are adopted. The mains power supply capacity is to be in place all at once in the final stage.

[0105] (5) Communication pipeline: According to formula (9), the main pipeline C=(4+8+6+4)×1.15 ≈ 25.3, take 26 holes, which meets the standard of 18 to 32 holes for the main pipeline; the outgoing pipeline of the edge node is reserved in 2 directions, with 6 to 12 holes in each direction.

[0106] (6) Gap decision: The final data center requirement is 199m², and the existing reusable space is κ. r ·S 存量 = 0.7×150 = 105m², according to formula (10): G = 199 Since 105 = 94m² > 0, an idle room (approximately 66m²) will be renovated for suitability, and a new edge computing space of approximately 100m² will be built in phases. The civil engineering of the new building will be completed in one go, with 100% of the final construction completed. The initial computing equipment will be deployed at 20% of the final capacity (approximately 550 TOPS).

[0107] (iv) Configuration data implementation and update optimization Complete the configuration data implementation and adopt the co-construction and co-operation model in step S105. The three operators share the data center and pipeline, and share the operation and maintenance costs according to formula (11). After the annual urban health check data is returned, if the low-altitude economic computing power demand intensity ρ is revised from 80 to 95 TOPS / 10,000 m², then execute steps S101 to S105 on a rolling basis. 边缘 The update was increased to approximately 3200 TOPS, triggering the third edge node to transition from a reserved state to a device expansion deployment.

[0108] In summary, this application provides a method for configuring AI-supporting facilities based on urban renewal. First, it assesses the transformation potential of the target area based on its basic physical parameters, providing a data foundation of existing resources for subsequent facility configuration. This clarifies the carrying capacity of the existing space. Simultaneously, it establishes a mapping matrix between business categories and computing power requirements based on industrial planning parameters and latency sensitivity, providing a classification basis for the differentiated decomposition of computing power needs. Then, based on this mapping matrix, it obtains the total computing power requirement and performs hierarchical decomposition, enabling the total computing power requirement to be accurately allocated to the edge level and district level according to latency sensitivity. The three-tiered system—domain-level, hub-level, and central-level—avoids resource waste or insufficient capacity caused by centralized planning at a single level, while ensuring end-to-end latency requirements for low-latency services. Finally, facility configuration is based on existing resource data and tiered computing power needs. This approach eliminates the disconnect between new construction needs and idle existing resources by prioritizing the reuse of existing space, and simultaneously inputs multiple configuration parameters to achieve coordinated support for civil engineering space, energy supply, and communication pipelines. Overall, this achieves the technical effects of data-driven computing power needs, precise tiered facility adaptation, economical reuse of existing resources, and coordinated configuration of supporting facilities. It addresses the challenge of improving the flexibility of AI infrastructure planning in related technologies.

[0109] Secondly, embodiments of this application provide an AI-based infrastructure configuration system for urban renewal. The system includes a stock resource data module, a mapping matrix module, a tiered computing power demand module, and a configuration module; wherein: The existing resource data module is used to assess the transformation potential of the target area based on its basic physical parameters, and to obtain existing resource data. The existing resource data module is used to divide the artificial intelligence business in the target area into multiple business categories based on industry planning parameters and latency sensitivity, and to establish a mapping matrix between business categories and computing power requirements. The tiered computing power requirement module is used to obtain the total computing power requirement of the target area. Based on the mapping matrix, the total computing power requirement is decomposed into tiers to obtain the tiered computing power requirements corresponding to different business categories. The configuration module is used to configure artificial intelligence supporting facilities based on hierarchical computing power requirements and existing resource data, and obtain configuration data to guide the physical artificial intelligence supporting facilities.

[0110] In summary, this application provides an AI-based infrastructure configuration system for urban renewal, which addresses the challenge of improving the flexibility of AI infrastructure planning in related technologies through the synergistic effect of existing resource data modules, mapping matrix modules, hierarchical computing power demand modules, and configuration modules. Specifically, the process begins by assessing the transformation potential of the target area based on its fundamental physical parameters, providing a data foundation of existing resources for subsequent facility configuration. This clarifies the carrying capacity of existing space. Simultaneously, a mapping matrix between business categories and computing power requirements is established based on industry planning parameters and latency sensitivity, providing a classification basis for the differentiated decomposition of computing power demands. Then, based on this mapping matrix, the total computing power demand is obtained and decomposed hierarchically. This allows the total computing power demand to be precisely allocated to three levels—edge, regional, and hub—according to latency sensitivity. This avoids resource waste or insufficient capacity caused by centralized planning at a single level, while ensuring end-to-end latency requirements for low-latency services. Finally, facility configuration is carried out by combining existing resource data with hierarchical computing power demands. On the one hand, prioritizing the reuse of existing space eliminates the disconnect between new construction demand and existing idle space. On the other hand, simultaneously inputting multiple configuration parameters achieves coordinated matching of civil engineering space, energy supply, and communication pipelines. This results in the overall technical effects of data-driven computing power demand, precise hierarchical matching of facilities, economical reuse of existing resources, and coordinated configuration of supporting facilities. This addresses the issue of how to improve the flexibility of planning supporting infrastructure for artificial intelligence in related technologies.

[0111] It should be noted that the AI-based infrastructure configuration system for urban renewal provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, terms such as "module," "unit," and "subunit" can refer to combinations of software and / or hardware that perform predetermined functions. Although the apparatus described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0112] Thirdly, embodiments of this application provide an electronic device, Figure 4 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 4 As shown, the electronic device may include a processor 41 and a memory 42 storing computer program instructions.

[0113] Specifically, the processor 41 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0114] The memory 42 may include a large-capacity memory for data or instructions. For example, and not limitingly, the memory 42 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 42 may include removable or non-removable (or fixed) media. Where appropriate, the memory 42 may be internal or external to a data processing device. In a particular embodiment, the memory 42 is non-volatile memory. In a particular embodiment, the memory 42 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0115] The memory 42 can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 41.

[0116] The processor 41 reads and executes computer program instructions stored in the memory 42 to implement any of the above embodiments of the method for configuring artificial intelligence supporting facilities based on urban renewal.

[0117] In one embodiment, an AI-supporting facility configuration device for urban renewal may further include a communication interface 43 and a bus 40. Wherein, as... Figure 4 As shown, the processor 41, memory 42, and communication interface 43 are connected through bus 40 and complete communication with each other.

[0118] The communication interface 43 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication interface 43 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0119] Bus 40 includes hardware, software, or both, that couples together components of an AI-based urban renewal infrastructure configuration device. Bus 40 includes, but is not limited to, at least one of the following: Data Bus, Address Bus, Control Bus, Expansion Bus, and Local Bus. For example, and not as a limitation, bus 40 may include Accelerated GraphicsPort (AGP) or other graphics buses, Extended Industry Standard Architecture (EISA) buses, Front Side Bus (FSB), HyperTransport (HT) interconnects, Industry Standard Architecture (ISA) buses, InfiniBand interconnects, Low Pin Count (LPC) buses, memory buses, Micro Channel Architecture (MCA) buses, Peripheral Component Interconnect (PCI) buses, PCI-Express (PCI-X) buses, Serial Advanced Technology Attachment (SATA) buses, Video Electronics Standards Association Local Bus (VLB) buses, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 40 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0120] Fourthly, embodiments of this application provide a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the method for configuring artificial intelligence supporting facilities based on urban renewal provided in the first aspect.

[0121] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0122] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform steps implementing the method for configuring artificial intelligence-based supporting facilities for urban renewal provided in the first aspect.

[0123] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0124] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for configuring AI-supporting facilities based on urban renewal, characterized in that, The method includes: Based on the basic physical parameters of the target area, the transformation potential of the target area is assessed to obtain existing resource data; Based on industry planning parameters and latency sensitivity, the artificial intelligence business in the target area is divided into multiple business categories, and a mapping matrix between the business categories and computing power requirements is established. Obtain the total computing power requirement of the target area, and based on the mapping matrix, perform hierarchical decomposition of the total computing power requirement to obtain the hierarchical computing power requirements corresponding to different business categories; Based on the tiered computing power requirements and the existing resource data, artificial intelligence supporting facilities are configured to obtain configuration data for guiding the physical artificial intelligence supporting facilities.

2. The method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 1, characterized in that, The assessment of the redevelopment potential of the target area, based on its basic physical parameters and industrial planning parameters, yields existing resource data, including: Obtain the basic physical parameters of the candidate existing spaces in the target area. The basic physical parameters include the floor load, net height, ventilation and heat dissipation conditions and power supply conditions of the candidate existing spaces. Based on the aforementioned basic physical parameters, for each of the candidate existing spaces, the transformation and reuse coefficient of the candidate existing space is calculated. The transformation and reuse coefficient is used to characterize the feasibility of transforming the candidate existing space into a space for supporting artificial intelligence facilities. Candidate existing spaces with a reuse coefficient greater than or equal to a preset threshold are identified as reuse and modification spaces. The total amount of reused space is obtained by counting all the reused and modified spaces, and the total amount of reused space is used as the stock resource data.

3. The method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 1, characterized in that, The process involves dividing the artificial intelligence business in the target area into multiple business categories based on industry planning parameters and latency sensitivity, and establishing a mapping matrix between the business categories and computing power requirements, including: Based on the industrial planning parameters, the end-to-end latency requirements for various artificial intelligence application formats in the target area are determined. The end-to-end latency requirements are the time budget for business data to complete the link transmission and processing from the collection end to the processing end. Based on the end-to-end latency requirements, the artificial intelligence services in the target area are divided into multiple service categories according to latency sensitivity; Based on the latency sensitivity of each business category, the edge traffic splitting coefficient and the regional traffic splitting coefficient corresponding to each business category are determined. The edge traffic splitting coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the edge computing power nodes of the region. The regional traffic splitting coefficient is the proportion of the total computing power demand of the corresponding business category allocated to the regional computing power nodes. A mapping matrix is ​​constructed based on the business category, the edge diversion coefficient, and the regional diversion coefficient.

4. The method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 1, characterized in that, The process of obtaining the total computing power requirement of the target region involves, based on the mapping matrix, performing a hierarchical decomposition of the total computing power requirement to obtain tiered computing power requirements corresponding to different business categories, including: Based on the scale data of various businesses, unit computing power demand intensity, occupancy rate conversion factor and average annual growth rate of computing power demand in the industrial planning parameters, calculate the total computing power demand of the target area within the planning period. Based on the edge-level and regional-level traffic splitting coefficients in the mapping matrix, the total computing power requirement is decomposed into edge-level computing power requirement, regional-level computing power requirement, and hub-level computing power requirement, which are referred to as the hierarchical computing power requirement.

5. A method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 4, characterized in that, The total computing power requirement is expressed by the following formula: In the formula, Q For the area t Total annual computing power requirement, in TOPS; K Number of industry categories; Ik For the first k A collection of business formats related to the same industry; Sk,i For the first k Category of industries i The planned building area or number of enterprises for each business type; ρk,i To determine the unit computing power demand intensity for corresponding business formats, the unit TOPS / 10,000 m² or TOPS / enterprise is calibrated through visits and surveys of typical industry enterprises, questionnaires, and computing power leasing market data. λk,i This refers to the occupancy rate or the conversion factor for reaching full production capacity. gk For the first k Average annual growth rate of computing power demand in related industries; t For the planning period; The edge-level computing power requirement is expressed by the following formula: The regional computing power requirement is expressed by the following formula: The hub-level computing power requirement is expressed by the following formula: In the formula, in the formula: Qk For the first k The computing power needs of related industries; 、 、 These are the computing power requirements at the edge of the region, the regional level, and the city (hub) level, respectively; among them, It will be uniformly supported by a city-level computing power hub, with only high-speed interconnection communication channels reserved in the area; It is undertaken by regional computing power rooms and aggregation communication rooms; It is undertaken by edge computing nodes in the region. K Number of industry categories; The edge split coefficient; This is the regional diversion coefficient.

6. A method for configuring AI-supporting facilities based on urban renewal according to claim 1, characterized in that, The process of configuring AI supporting facilities based on the tiered computing power requirements and the existing resource data to obtain configuration data for guiding the physical AI supporting facilities includes: Based on the edge computing power requirements in the tiered computing power requirements, determine the physical deployment number of edge computing power nodes and the deployment location of each edge computing power node; Based on the aforementioned edge computing power requirements, the parameters of the computer room building area, the mains power supply capacity, and the communication pipeline bore capacity are calculated. Based on the existing resource data, the computer room building area parameters, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters, the supply and demand gap of artificial intelligence supporting facilities is determined; Based on the supply and demand gap, data is generated to guide the configuration of supporting facilities for physical artificial intelligence.

7. A method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 6, characterized in that, The step of determining the physical deployment quantity of edge computing nodes and the deployment location of each edge computing node based on the edge-level computing power demand in the hierarchical computing power demand includes: The maximum service radius of the edge computing node is determined based on the end-to-end latency requirements of the business corresponding to the edge computing power demand, the edge node processing latency, and the queuing and access latency. The service radius of the edge computing node is taken as a value not exceeding the maximum service radius, and a first number that satisfies the coverage area constraint is determined based on the service radius and the area of ​​the target region. Based on the edge-level computing power requirements and the preset standard computing power configuration for a single node, a second quantity that satisfies the computing power capacity constraint is determined; Select the larger quantity from the first quantity and the second quantity as the physical deployment quantity; A corresponding number of candidate locations are defined in the target area. A site selection suitability index is calculated for each candidate location. Based on the site selection suitability index from high to low, the physical deployment number of locations is selected from the candidate locations as the deployment locations of the edge computing nodes.

8. A method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 6, characterized in that, The supply and demand gap of artificial intelligence supporting facilities is determined based on the existing resource data, the computer room building area parameters, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters. Based on the aforementioned supply-demand gap, data is generated to guide the configuration of supporting facilities for physical artificial intelligence, including: The final facility requirements are determined based on the building area of ​​the computer room, the mains power supply capacity parameters, and the communication pipeline bore capacity parameters. Based on the existing resource data, the renovation and reuse coefficient, and the final facility demand scale, the supply and demand gap is determined; When the supply-demand gap is less than or equal to zero, a configuration decision is generated to reuse existing space; when the supply-demand gap is greater than zero, a configuration decision is generated to allocate facility resources according to a preset priority order to meet the supply-demand gap.

9. A method for configuring artificial intelligence supporting facilities based on urban renewal according to claim 1, characterized in that, After configuring AI supporting facilities based on the tiered computing power requirements and the existing resource data to obtain configuration data for guiding the physical AI supporting facilities, the method further includes: Configure the artificial intelligence supporting facilities based on the configuration data, and collect the operation status monitoring data of the artificial intelligence supporting facilities; The configuration data is iteratively updated based on the operational status monitoring data.

10. A system for configuring AI-powered supporting facilities based on urban renewal, characterized in that: The system is used to execute the method according to any one of claims 1 to 9, and the system includes: a stock resource data module, a mapping matrix module, a hierarchical computing power demand module, and a configuration module; wherein: The existing resource data module is used to assess the transformation potential of the target area based on the basic physical parameters of the target area, and obtain existing resource data. The existing resource data module is used to divide the artificial intelligence business of the target area into multiple business categories according to industry planning parameters and latency sensitivity, and to establish a mapping matrix between the business categories and computing power requirements. The hierarchical computing power requirement module is used to obtain the total computing power requirement of the target area, and based on the mapping matrix, to perform hierarchical decomposition of the total computing power requirement to obtain the hierarchical computing power requirements corresponding to different business categories. The configuration module is used to configure artificial intelligence supporting facilities based on the hierarchical computing power requirements and the existing resource data, and obtain configuration data for guiding the physical artificial intelligence supporting facilities.