Multi-source government affair data-driven urban intelligent center dynamic management system

Abstract

The invention discloses a multi-source government affair data-driven urban intelligent center dynamic government system, relates to the technical field of urban government, and solves the problems of data fusion staticization, disjunction between a government strategy and a scene and lack of full-link dynamic feedback in the prior art. The system comprises a multi-source data access layer, a dynamic fusion engine, a scenarized governance center, a feedback optimization module and double databases. The multi-source data access layer accesses and preprocesses multi-source government affair data; the dynamic fusion engine realizes data real-time fusion based on the space-time label and the service label; the scenarized governance center generates a dynamic governance instruction according to the fused data; the feedback optimization module collects execution effect data, generates optimization parameters and feeds back the optimization parameters to the preorder module to realize closed-loop optimization; the double databases are used for storing total data and governance strategies; according to the method, dynamic closed-loop treatment of data-strategy-execution-optimization is realized, the data fusion efficiency, the scene adaptability and the treatment effect are improved, and the method is suitable for various urban treatment scenes.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of urban governance and government data processing, and relates to a system for dynamic governance of an urban intelligent hub driven by multi-source government data. BACKGROUND

[0002] With the advancement of digital government construction, multi-source government data (such as government service data, urban operation data, emergency data, etc.) has become the core support for urban governance. In the prior art, although a “one-network handling” and “one-network unified management” linkage system (such as CN202310909165.X) and a public data-led urban data space system (such as CN202411919366.9) have appeared, there are still three key defects: Data fusion is static: existing systems mostly fuse data through preset fixed rules (such as CN202410876543.2), which cannot dynamically adjust the fusion logic according to data characteristics (such as the timeliness of emergency data and the relevance of government service data), resulting in low cross-department data correlation efficiency and even correlation failure.

[0003] Governance strategy and scene are disconnected: existing intelligent hubs mostly output static reports or basic strategies (such as CN202310567890.1), without generating dynamic instructions adapted to the scene in combination with real-time data characteristics, for example, in a heavy rain emergency scene, it is impossible to adjust the drainage and traffic diversion strategies in real time according to the rainfall.

[0004] Lack of full-link dynamic feedback: existing systems do not build a closed loop of “data-strategy-execution-optimization” (such as the three-stage framework in the “Urban Government Data Governance Technology White Paper (2024)”), and the governance effect data cannot feed back to data fusion and strategy generation, resulting in a decline in governance efficiency over time. Therefore, in order to solve the above technical problems, the technical solution of the present application is set. SUMMARY

[0005] The purpose of the present application is to provide a system for dynamic governance of an urban intelligent hub driven by multi-source government data, which aims to solve the technical problems of “data fusion static, governance strategy and scene disconnected, lack of full-link dynamic feedback” in existing multi-source government data-driven urban governance systems, and to realize dynamic adaptation and continuous optimization of urban governance.

[0006] The technical solution adopted by the present application is as follows: A system for dynamic governance of an urban intelligent hub driven by multi-source government data, comprising a multi-source data access layer, a dynamic fusion engine, a scenario-based governance hub, a feedback optimization module, and a dual database, the connection relationship and functions of each module being as follows: Multi-source data access layer: access full government data through four types of interfaces (government service, city operation, emergency event, third-party cooperation), and transmit to dynamic fusion engine after preprocessing (format standardization, outlier removal), while ensuring data access security through OAuth2.0 identity verification.

[0007] Dynamic fusion engine: based on the space-time label (timestamp + administrative division code) and business label (government field + business type) of data, call the mapping rule library (including self-learning unit) to realize real-time classification and fusion of data, and synchronize the fused data to the full data pool.

[0008] Scenario-based governance hub: identify the current governance scenario (daily monitoring / emergency disposal / service optimization), call the basic strategy of the governance strategy library and adjust the parameters, generate dynamic execution instructions, and issue them to the end execution unit after compliance verification.

[0009] Feedback optimization module: collect execution effect data (progress / achievement rate / public feedback), compare it with the expected effect after encryption, generate optimization parameters through gradient descent algorithm, and feedback to dynamic fusion engine and scenario-based governance hub, while updating the governance strategy library.

[0010] Dual database: full data pool uses distributed storage to support fast data retrieval; governance strategy library stores basic strategies and updates them according to feedback optimization period, and is linked with scenario-based governance hub.

[0011] Working principle of the application Data access stage: multi-source data access layer accesses data through government service data interface (such as "one network handling" platform interface), city operation data interface (such as traffic monitoring, water quality monitoring interface), emergency event data interface (such as fire alarm, weather warning interface), third-party cooperation data interface (such as enterprise credit data interface); identity verification unit accesses visitor identity through OAuth2.0 verification interface, after verification, data preprocessing unit performs format standardization (unified to JSON format), outlier removal (based on 3σ principle), missing value completion (based on the same field data mean), and the processed data is transmitted to dynamic fusion engine.

[0012] Data fusion stage: the label generation unit of the dynamic fusion engine generates spatio-temporal labels (such as 2024-10-1509:45_510104, where 510104 is the administrative division code of the Jinjiang District of Chengdu) and business labels (such as "government service _ social security payment") for the access data; the real-time matching unit calls the rules in the mapping rule library (such as "social security payment data and population information data are associated according to the ID number"), and completes the real-time matching and fusion of data and governance scenarios; the fused data is transmitted to the scenario-based governance hub on the one hand, and stored in the full data pool of the dual database (using the Hadoop distributed storage architecture, supporting retrieval according to spatio-temporal labels and business labels) on the other hand.

[0013] Policy generation and instruction issuing stage: the scenario recognition unit of the scenario-based governance hub identifies the governance scenario (such as "rainstorm emergency response scenario") according to the fusion data characteristics (such as "rainstorm warning data + road water accumulation data"); the policy generation unit calls the basic policy in the governance policy library (such as "rainstorm emergency basic policy: close the road section when the water accumulation is over 30 cm"), adjusts the parameters combined with real-time data (such as "current water accumulation is 45 cm"), and generates a dynamic governance policy (such as "close XX road section and dispatch 3 water trucks to the area"); the instruction verification unit of the instruction issuing unit calls the compliance rules in the governance policy library (such as "road closure needs to notify the traffic department in advance") to verify the instruction, and after verification, the instruction is converted into a standardized execution instruction (such as a JSON format instruction) and transmitted to the city governance end execution unit (such as the traffic management department terminal and the emergency rescue team terminal).

[0014] Feedback optimization stage: the effect data collection unit of the feedback optimization module collects the execution progress data (such as "water truck has arrived at XX road section, water accumulation has decreased to 20 cm"), target achievement rate data (such as "no vehicle is stranded in the closed road section, target achievement rate is 100%"), and public feedback data (such as "citizens feedback that the water accumulation disposal is timely through the government APP"); the data encryption unit encrypts the collected data using the SM4 algorithm, and the encrypted data is transmitted to the optimization analysis unit; the optimization analysis unit compares the collected data with the expected effect data (such as "water accumulation is reduced to less than 10 cm within 2 hours") in the governance policy library, and calculates the optimization parameters (such as "next time when the rainstorm water accumulation is over 40 cm, dispatch the water truck in advance") through the gradient descent algorithm; the parameter output unit transmits the optimization parameters to the dynamic fusion engine (updates the mapping rules) and the scenario-based governance hub (updates the policy parameters), and updates the governance policy library of the dual database; the rule self-learning unit of the dynamic fusion engine iteratively trains the mapping rules according to the historical data and optimization records of the full data pool every quarter, and updates the mapping rule library.

[0015] In summary, due to the adoption of the above technical solutions, the present application has the following advantages: 1. A multi-source government data-driven urban intelligent hub dynamic governance system, which has the advantages of data fusion and dynamicity, solves the problem of cross-department data correlation failure through dynamic mapping of space-time tags and business tags combined with rule self-learning, and improves the data fusion efficiency by more than 40% compared with existing systems (based on comparative tests of the same data set).

[0016] 2. The application has the advantages of governance strategy scenario, realizes the automatic generation of "data features-scene recognition-dynamic instructions", and shortens the emergency event disposal response time to within 5 minutes (the existing system is 15 minutes on average), and improves the satisfaction degree of government service by 25%.

[0017] 3. The application has the advantages of governance link closed loop, and builds a closed loop of "data-strategy-execution-optimization" through a feedback optimization module, and the governance strategy adaptation accuracy gradually improves with running time, and stabilizes at more than 92% after 6 months (the existing system is 75% on average). BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments, experimental examples and comparative examples, and it should be understood that the following drawings only show some embodiments of the application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor, wherein: Figure 1 is the architecture diagram of the application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the application more clear and explicit, the following will further illustrate the application in combination with the drawings, embodiments, experimental examples and comparative examples. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application, that is, the described embodiments are only a part of the embodiments of the application, not all the embodiments. The components of the embodiments of the application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the application provided in the drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0021] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0022] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0023] Example 1 This invention discloses a system for dynamic governance of a city's intelligent central hub driven by multi-source government data, such as... Figure 1 As shown, the specific implementation method of this embodiment is as follows: Routine governance scenarios in small and medium-sized cities (taking Mianyang City, Sichuan Province as an example) System configuration: The multi-source data access layer accesses government service data interfaces (Mianyang City's "One-Stop Government Service" platform, including business interfaces such as social security payment and housing provident fund withdrawal) and urban operation data interfaces (Mianyang urban traffic monitoring, garbage classification point monitoring, and Fujiang River water quality monitoring interfaces); the dynamic fusion engine's mapping rule base is preset with two types of basic rules: "associating social security payment data with population and household registration data by ID card number" and "associating Fujiang River water quality data with government service data of riverside streets by administrative division code"; the governance strategy base is preset with three types of basic strategies adapted to Mianyang's "science and technology city + river basin city" characteristics: "daily traffic management strategy", "garbage classification supervision strategy" and "basic strategy for watershed water quality protection"; the feedback optimization module's optimization cycle is set to 1 day to adapt to the data increment scale of small and medium-sized cities (approximately 8,000 pieces of government data per day).

[0024] Operation process: Data access: Social security payment data from the Mianyang City "One-Stop Government Service" platform (3,200 records per day, including payment data for employees of technology companies), morning rush hour congestion data from urban traffic monitoring (7:30-9:00, concentrated on main roads such as Linyuan Road and Fucheng Road, 2,800 records per day), and Fujiang River water quality monitoring data (pH value, dissolved oxygen, etc., collected every 2 hours, 12 records per day) are accessed into the system; The identity verification unit verifies the identity of each interface visitor (such as the Municipal Government Service Center, the Municipal Transportation Bureau, and the Municipal Ecological Environment Bureau) through OAuth2.0. After successful verification, the data preprocessing unit standardizes the data into JSON format, removes abnormal data from traffic monitoring (such as blank records caused by equipment failure) based on the 3σ principle, and uses the average water quality of the same basin to fill in missing values. The processed data is then transmitted to the dynamic fusion engine.

[0025] Data Fusion: The dynamic fusion engine's tag generation unit generates a spatiotemporal tag for social security payment data, "2024-10-10 09:15_510703" (Fucheng District, Mianyang City), and a business tag for "Government Services_Social Security Payment_Technology Enterprises." It also generates a spatiotemporal tag for traffic congestion data, "2024-10-10 08:20_510703," and a business tag for "Urban Operation_Traffic Congestion_Main Roads." Finally, it generates a tag for water quality data, "2024-10..." The spatiotemporal label "-1008:00_510703_Fujiang River Basin" and the business label "Urban Operation_Water Quality Monitoring_Basin Protection" are used; the real-time matching unit calls the "Regional Association Rules" in the mapping rule library to merge the three types of data into "Association Data of Peak Social Security Payments for Employees of Science and Technology Enterprises in Fucheng District, Traffic Congestion on Main Roads, and Fujiang River Water Quality", and synchronously store it in the full data pool of the dual databases (using Hadoop distributed storage to adapt to the data storage needs of Mianyang Science and Technology City).

[0026] Strategy Generation: The scenario identification unit of the scenario-based governance center identifies the scenario as "Daily Operation Monitoring Scenario (including traffic and government service sub-scenarios)" based on the characteristics of the integrated data ("The peak period for social security payments by science and technology enterprises in Fucheng District is concentrated between 8:30 and 10:00, which overlaps with the morning peak congestion period on Linyuan Main Road, and the water quality indicators of the Fu River are normal"). The strategy generation unit calls the "Daily Traffic Management Strategy" in the governance strategy library, adjusts the parameters in combination with real-time data, and generates a dynamic governance instruction: "Add two temporary parking spaces around the government service hall of Fucheng Science and Technology City (open from 7:30 to 10:30), optimize bus routes 15 and 28 to detour around the congested section of Linyuan Main Road, and add two special windows for social security payments for science and technology enterprises in the government service hall". The instruction verification unit of the instruction issuance unit calls the "Mianyang Urban Traffic Management Compliance Rules" and "Government Service Window Setting Specifications" to verify the instruction. After confirming that there is no compliance risk, the instruction is transmitted to the terminal execution units of the Mianyang Municipal Traffic Management Bureau and the Municipal Government Service Center.

[0027] Feedback Optimization: The following day, the feedback optimization module's effect data collection unit collected execution effect data showing that "temporary parking space utilization rate was 92%, congestion time was reduced by 18 minutes after the No. 15 bus route was rerouted, and the processing time at the social security window for technology enterprises was reduced from an average of 20 minutes to 12 minutes." The data encryption unit encrypted the data using the SM4 algorithm and then transmitted it to the optimization analysis unit. The optimization analysis unit compared the collected data with the expected effect data in the governance strategy library ("temporary parking space utilization rate ≥ 85%, congestion time reduced by ≥ 15 minutes, processing time reduced to ≤ 15 minutes"), and calculated optimization parameters ("extend the opening time of temporary parking spaces to 11:00, and assign one business counselor to the social security window for technology enterprises") using the gradient descent algorithm. The parameter output unit transmitted the optimization parameters to the dynamic fusion engine (updating "technology enterprise data association rules") and the scenario-based governance hub (updating "window resource scheduling parameters"), while also updating the governance strategy library of the dual databases.

[0028] Implementation results: During peak hours for social security contributions for employees of technology companies in Fucheng District, Mianyang City, traffic congestion rate decreased from 42% to 19%; satisfaction with social security services for technology companies increased by 38%; and the efficiency of correlation analysis between water quality monitoring data of the Fu River and government service data along the river increased by 50%, which is in line with Mianyang's urban positioning of "Science and Technology City + Watershed Governance".

[0029] Example 2 This invention discloses a system for dynamic governance of a city's intelligent central hub driven by multi-source government data, such as... Figure 1 As shown, the specific implementation method of this embodiment is as follows: Emergency management scenarios in large cities (taking Chengdu, Sichuan Province as an example) System configuration: The multi-source data access layer adds emergency event data interfaces (Chengdu Emergency Management Bureau early warning interface, fire and rescue brigade alarm interface, and Chengdu Meteorological Bureau rainstorm early warning interface) and third-party collaborative data interfaces (Sichuan Provincial Geological Disaster Monitoring Center landslide early warning interface and Chengdu Metro operation data interface); the dynamic fusion engine's mapping rule library adds two types of rules adapted to Chengdu's "mega-city + mountainous and hilly terrain" characteristics: "associating meteorological early warning data and geological disaster data by timestamp + administrative division code" and "associating fire alarm data and metro operation data by regional radius (within 1 kilometer)"; the governance strategy library adds three types of special strategies: "rainstorm emergency response strategy", "geological disaster prevention strategy", and "metro operation linkage emergency strategy"; the optimization cycle of the feedback optimization module is set to 8 minutes (adapting to the high frequency and fast response requirements of mega-city emergency events), and the data encryption unit is synchronously connected to the Chengdu Municipal Government Cloud security system.

[0030] Operation process: Data access: Chengdu Meteorological Bureau issued an "Orange Rainstorm Warning" at 09:30 on October 15, 2024 (the main urban area is expected to receive ≥50mm of rainfall in the next 3 hours); Sichuan Provincial Geological Disaster Monitoring Center simultaneously issued a "Blue Landslide Warning for the Longquanshan Area"; the Fire and Rescue Brigade reported at 09:45 that "a flood alarm has been raised in the underground garage of a residential area on Dongda Street in Jinjiang District (the water depth has reached 40cm)"; Chengdu Metro operation data showed that "the pedestrian flow around Dongmen Bridge Station on Line 2 has increased by 30% compared to weekdays (the morning rush hour has not yet subsided)"; after OAuth2.0 authentication, the data preprocessing unit removed duplicate push data from the meteorological warnings and supplemented the specific location information of the garage in the fire alarm data (through GIS map matching). The processed data was then transmitted to the dynamic fusion engine.

[0031] Data Fusion: The dynamic fusion engine's tag generation unit generates a unified spatiotemporal tag "2024-10-15 09:45_510104" (Jinjiang District, Chengdu) for four types of data. It adds an additional business tag "Longquanshan Area_Landslide Warning" to the geological disaster data and "Line 2_Dongmen Bridge Station_Peak Passenger Flow" to the metro data. The real-time matching unit calls "timestamp + administrative division association rules" and "regional radius association rules" to merge the data into "composite emergency data of Jinjiang District rainstorm waterlogging + Longquanshan landslide warning + metro peak passenger flow", which is then pushed to the scenario-based governance center and the full data pool.

[0032] Strategy Generation: The scenario recognition unit identifies the scenario as a "composite emergency response scenario (primarily rainstorm, supplemented by geological disaster and public transportation linkage)" based on the characteristics of the fused data ("orange rainstorm warning + backflow of water in underground parking garage + peak passenger flow in surrounding subway stations + suburban landslide warning"). The strategy generation unit calls the "rainstorm emergency response strategy" and adjusts the parameters based on real-time data ("water depth of 40cm exceeds the warning threshold by 30cm, drainage should be prioritized; peak passenger flow in subway stations should be avoided to prevent people from staying"), generating "1. Dispatch two large water pumps from Chengdu Drainage Company to Jinjiang District East..." 1. Arrive at Dajie Community within 30 minutes; 2. Notify Chengdu Metro Line 2 Dongmen Bridge Station to open 2 additional exits and arrange 3 staff members to guide the flow of people; 3. The Longquanyi District Emergency Management Bureau organized 3 patrol teams to investigate potential hazards in the Longquanshan area; 4. Push the dynamic instruction "Rainstorm Avoidance Tips for the Main Urban Area" through the "Chengdu Release" government new media platform; After the instruction verification unit calls the "Chengdu Emergency Response Procedure Specifications" and "Metro Operation Safety Rules" to verify the compliance of the instruction, it is issued to the terminal execution units of the Municipal Drainage Company, Chengdu Metro, Longquanyi District Emergency Management Bureau, and Municipal Cyberspace Administration.

[0033] Feedback Optimization: 8 minutes later, the effect data collection unit collected execution data showing that "the water pumping truck is en route (expected to arrive in 25 minutes), the Dongmen Bridge Station has opened additional exits, the Longquanshan patrol team has set off, and the government new media notification has been pushed out (with over 500,000 views)." The optimization analysis unit compared the expected results ("the water pumping truck departs within 30 minutes, and the additional exits are opened within 10 minutes") and calculated optimization parameters ("In the next similar scenario, dispatch the water pumping truck to the main urban standby point 5 minutes in advance, and simultaneously push the instruction to open additional subway exits to station security personnel"). The parameter output unit fed back the optimization parameters to the dynamic fusion engine (updating the "emergency resource dispatch mapping rules") and the scenario-based governance center (updating the "subway linkage emergency parameters") in real time to ensure more efficient subsequent emergency response.

[0034] Implementation Results: In Chengdu's Jinjiang District, the time for water accumulation in underground parking garages to decrease from 40cm to the safe threshold (10cm) was shortened to 55 minutes (originally 120 minutes); the peak passenger flow duration at Dongmen Bridge Station was shortened from 45 minutes to 22 minutes; no new landslide hazard points were added in the Longquanshan area; and the public's emergency awareness rate in the main urban area reached 92% during the rainstorm, fully meeting the complex needs of emergency management in Chengdu, a megacity.

[0035] Example 3 This invention discloses a system for dynamic governance of a city's intelligent central hub driven by multi-source government data, such as... Figure 1 As shown, the specific implementation method of this embodiment is as follows: Urban governance scenario of administrative division adjustment (taking Tongchuan District of Dazhou City, Sichuan Province as an example, in 2024, some townships of Tongchuan District were transferred to the jurisdiction of Dachuan District). System Configuration: The administrative division coding of the multi-source data access layer adopts the redundant coding system described in claim 6. The original "Fuxing Town, Tongchuan District, Dazhou City" is assigned the standard code 511702105 and the redundant code DZTXC001. The "Fuxing Town" (belonging to Dachuan District) after the relocation is assigned the standard code 511703108, and the redundant code is still associated with DZTXC001. The mapping rule library of the dynamic fusion engine adds two types of rules to adapt to the administrative division adjustment: "historical government data is associated with the redundant code, and new data is associated with the standard code" and "cross-regional data query is matched by the dual index of redundant code + standard code". The governance strategy library adds "government service strategy during the transition period of administrative division adjustment" and "cross-regional data sharing and collaboration strategy", which mainly covers high-frequency cross-regional businesses such as social security, household registration, and real estate registration.

[0036] Operation process: Data access: The system integrates the following data: "Historical social security payment data of residents of Fuxing Town, Tongchuan District before the transfer" (86,000 records, involving 32,000 residents), "New social security payment data of Fuxing Town, Dachuan District after the transfer" (320 records per day), and household registration migration data between Tongchuan District and Dachuan District (85 records per day, including cross-district migration records of residents of Fuxing Town); the identity verification unit performs dual verification (OAuth2.0 + IP whitelist) for the interface access of Dazhou Municipal Social Security Bureau and Municipal Public Security Bureau to ensure cross-district data security; the data preprocessing unit unifies the encoding format of the two types of administrative region data (synchronizes the original household registration encoding format of Tongchuan District to the standard of Dachuan District) and completes the township affiliation information in the historical data (marked "belonged to Tongchuan District before the transfer in 2024").

[0037] Data Fusion: The dynamic fusion engine's tag generation unit generates a spatiotemporal tag "2024-09-30 (before relocation)_DZTXC001" and a business tag "Government Services_Social Security Payment_Historical Data" for historical data, and generates a spatiotemporal tag "2024-10-05 (after relocation)_511703108" and a business tag "Government Services_Social Security Payment_New Data" for new data; the real-time matching unit calls the "redundant coding association rule" to automatically associate the historical social security data of the same resident with the new data (e.g., resident Zhang's historical data is matched to the new data through DZTXC001), and at the same time generates a "Fuxing Town Resident Cross-District Social Security Data Association Report", which is synchronized to the full data pool and the Dazhou Municipal Social Security Bureau's terminal execution unit.

[0038] Strategy Generation: The scenario identification unit identifies the scenario as "government service optimization scenario (transition of administrative division adjustment)" based on the characteristics of the fused data ("centralized demand for cross-district data association in Fuxing Town, and the need for synchronous updates of resident household registration and social security data"). The strategy generation unit calls the "government service strategy during the transition period" and, in conjunction with the "cross-district handling" policy requirements of Dazhou City, generates the following instructions: "1. Establish a 'special window for administrative division adjustment' at the Fuxing Town Government Service Center in Dachuan District to handle historical data inquiries and cross-district business; 2. Open the 'Tongchuan-Dachuan' social security data sharing channel, so residents do not need to repeatedly submit historical payment certificates; 3. Send SMS reminders to residents of Fuxing Town to inform them of the operation process for linking household registration and social security." The instruction verification unit verifies the instructions against the "Dazhou City Administrative Division Adjustment Government Service Guarantee Plan" and then issues them to the Dachuan District Government Service Center, the Municipal Social Security Bureau, and the Municipal Mobile Company.

[0039] Feedback Optimization: The feedback optimization module collects performance data every 24 hours. On the first day, it collected information such as "126 transactions were processed at the dedicated window, the data sharing channel utilization rate was 88%, and the resident SMS reading rate was 91%". The optimization analysis unit compared the expected results ("daily processing volume of the dedicated window ≥ 100 transactions, utilization rate ≥ 80%) and calculated optimization parameters ("add one staff member familiar with Tongchuan District's business to the dedicated window, and upgrade SMS notifications to a dual reminder of 'government APP pop-up + SMS'"). The parameter output unit updated the optimization parameters to the governance strategy database and simultaneously synchronized them to the Tongchuan District Government Service Center (to assist in training staff).

[0040] Implementation Results: During the administrative division adjustment of Fuxing Town, Dazhou City, the average processing time for residents' social security services across districts was reduced from 50 minutes to 15 minutes. The accuracy rate of historical data queries reached 100%, with no data gaps or duplicate submissions. The complaint rate for cross-district government services dropped from 18% at the beginning of the adjustment to 0.3%, meeting the government service requirements of Sichuan Province for "smooth transition of administrative division adjustments".

[0041] Example 4 This invention discloses a system for dynamic governance of a city's intelligent central hub driven by multi-source government data, such as... Figure 1 As shown, the specific implementation method of this embodiment is as follows: Cross-regional collaborative governance scenario (taking the Chengdu-Deyang-Meishan-Ziyang metropolitan area in Sichuan Province as an example, focusing on cross-regional supervision of enterprises in the four cities of Chengdu, Deyang, Meishan and Ziyang) System Configuration: The multi-source data access layer adds a "Chengdu-Deyang-Meishan-Ziyang" government data sharing interface (data interaction interface between Chengdu Municipal Market Supervision Bureau, Deyang Municipal Taxation Bureau, Meishan Municipal Human Resources and Social Security Bureau, and Ziyang Municipal Ecological Environment Bureau), which complies with the "Chengdu-Deyang-Meishan-Ziyang Government Data Sharing Management Measures"; the dynamic fusion engine's mapping rule library adds two types of collaborative rules: "Cross-city enterprise data associated with the 'Chengdu-Deyang-Meishan-Ziyang' unified social credit code" and "Enterprise supervision data matched by business type (taxation / social security / environmental protection)"; the governance strategy library adds "Chengdu-Deyang-Meishan-Ziyang cross-regional joint supervision strategy" and "Cross-city government service collaborative strategy", and sets up dedicated functional modules such as "joint law enforcement instruction generation" and "cross-city data feedback"; the optimization cycle of the feedback optimization module is set to 12 hours to adapt to the timeliness requirements of cross-city collaborative governance.

[0042] Operation process: Data access: The Chengdu Municipal Market Supervision Bureau reported that "a certain auto parts company (Unified Social Credit Code 91510114XXXXXX) has irregularities in its production ledgers"; the Deyang Municipal Taxation Bureau reported that "the company's branch in Deyang has a record of delayed tax declaration"; the Meishan Municipal Human Resources and Social Security Bureau reported that "the company's factory in Meishan has two complaints about social security arrears payments"; and the Ziyang Municipal Ecological Environment Bureau provided "wastewater discharge compliance data of the company's Ziyang production base". The data from the four cities were accessed through the "Chengdu-Deyang-Meishan-Ziyang" government data sharing platform. The identity verification unit adopted a joint authentication mechanism among the four cities (each municipal government department needs to log in through its own OAuth2.0 account). The data preprocessing unit standardized the format of the company name and branch name (to avoid format differences between "Chengdu XX Company" and "Chengdu XX Limited Liability Company") and removed duplicate complaint records.

[0043] Data Fusion: The tag generation unit of the dynamic fusion engine generates a unified spatiotemporal tag for the data from the four cities, namely “2024-10-08_Chengdu-Deyang-Meishan-Ziyang_91510114XXXXXX”, and adds business tags such as “Enterprise Supervision_Taxation” and “Enterprise Supervision_Social Security” to each type of data; the real-time matching unit calls the “Unified Social Credit Code Association Rules” to merge the data from the four cities into “Cross-city Supervision Panoramic Data of the Auto Parts Enterprise”, and simultaneously generates a “List of Enterprise Issues” (including three types of issues: non-standard ledgers, tax delays, and social security complaints), which is pushed to the end execution unit of the scenario-based governance center and the joint supervision team of the four cities.

[0044] Strategy Generation: The scenario identification unit identifies the scenario as "Cross-regional enterprise supervision scenario (Chengdu-Deyang-Meishan-Ziyang collaboration)" based on the characteristics of the fused data ("Enterprises have multiple regulatory issues across four cities, requiring joint handling"). The strategy generation unit calls the "Cross-regional joint supervision strategy" and, in conjunction with the "Enterprise Joint Supervision" requirements in the "Chengdu-Deyang-Meishan-Ziyang Integrated Development Three-Year Action Plan (2024-2026)," generates the following instructions: "1. The Chengdu Municipal Market Supervision Bureau will lead a special joint enforcement team, in conjunction with relevant departments of Deyang, Meishan, and Ziyang, to inspect the enterprise's branches within 3 days; 2. A real-time data sharing channel for supervision among the four cities will be opened, with enforcement progress synchronized daily; 3. The enterprise is required to submit a rectification plan within 5 days, and the four cities will jointly accept the rectification upon completion." The instruction verification unit verifies the instruction through the "Chengdu-Deyang-Meishan-Ziyang" integrated government service compliance platform and then issues it to the market supervision, tax, human resources and social security, and environmental protection departments of the four cities.

[0045] Feedback Optimization: 12 hours later, the effect data collection unit collected progress data showing that "the four cities have formed a joint law enforcement team, the shared channel has been opened, and enterprises have received rectification notices"; the optimization analysis unit calculated optimization parameters ("dividing law enforcement inspections into three special groups: 'ledger verification,' 'tax verification,' and 'social security verification,' to improve inspection efficiency; adding a 'rectification progress upload' module to the shared channel"); the parameter output unit fed back the optimization parameters to the scenario-based governance center (updating joint law enforcement strategies) and the dynamic fusion engine (updating cross-city data association rules), and simultaneously filed them with the "Chengdu-Deyang-Meishan-Ziyang" city integration office.

[0046] Implementation results: The response time for cross-city joint law enforcement of auto parts enterprises in the Chengdu-Deyang-Meishan-Ziyang metropolitan area has been shortened from an average of 96 hours to 36 hours, the timely submission rate of enterprise rectification plans has reached 100%, and the efficiency of data sharing among the four cities has increased by 75%, effectively supporting the government affairs in Sichuan Province's "Chengdu-Deyang-Meishan-Ziyang integration" strategy.

[0047] Fundamentals of Experimental Design Unified testing conditions Test dataset: 30 days of multi-source government data from a new first-tier city (with a resident population of 8 million+), including three core data categories: ① Government service data (social security payments, real estate registration, etc., totaling 1.2 million records, averaging 40,000 records per day); ② City operation data (traffic monitoring, water quality monitoring, garbage classification data, etc., totaling 3.6 million records, averaging 120,000 records per day); ③ Emergency event data (rainstorm warnings, fire alarms, traffic accident data, etc., totaling 12,000 records, averaging 400 records per day).

[0048] Test hardware environment: All test systems were deployed on servers with the same configuration (CPU: Intel Xeon Gold 6330, memory: 128GB DDR4, storage: 2TB SSD, network bandwidth: 1000Mbps) to eliminate the influence of hardware differences.

[0049] Test software environment: The operating system is CentOS 8.5, the database is MySQL 8.0 (comparison ratio 1 / 2) and Hadoop 3.3.4 (this invention, distributed storage of full data), and the programming language is Java 11 to ensure the consistency of the software environment.

[0050] Evaluation Metrics: Four core metrics are set: ① Data fusion efficiency (the amount of data fused per unit time, unit: data items / second); ② Scenario adaptation accuracy (the degree of matching between the generated governance instructions and the actual scenario requirements, calculated through manual verification and business rule validation, unit: %); ③ Governance response time (the time from data access to instruction issuance to the end execution unit, unit: seconds); ④ Effect feedback optimization rate (the improvement of secondary governance metrics after optimization based on execution effect data, unit: %).

[0051] Experimental Example 1 Rule self-learning performance test of dynamic fusion engine 1. Test parameters Testing period: 14 days (the first 7 days are the rule initialization phase, and the last 7 days are the self-learning iteration phase). Dynamic fusion engine configuration: Spatiotemporal label generation frequency (synchronized with data access, generated in real time), initial rules of the mapping rule base (15 basic association rules, such as "social security data - population data associated by ID card number" and "traffic congestion data - regional government service data associated by administrative division code"), rule self-learning cycle (the rules are iteratively updated every day at 24:00 based on the fusion data and feedback results of the day).

[0052] 2. Testing Process Days 1-7: Data fusion is performed using the initial mapping rules. The data fusion efficiency and fusion error rate (amount of data with association errors / total amount of data fused) are recorded daily. Days 8-14: Enable the rule self-learning function. The mapping rules will be automatically adjusted daily based on the previous day's fusion error data (such as "failure to associate cross-regional enterprise data"). (Add "joint association rule of enterprise unified social credit code - administrative division code"), and record the daily data fusion efficiency and fusion error rate.

[0053] 3. Test Results Conclusion: After enabling rule self-learning, the data fusion efficiency improved by 44.5% ((1850-1280) / 1280≈44.5%), and the fusion error rate decreased by 75% ((3.2-0.8) / 3.2=75%), proving that the self-learning capability of the dynamic fusion engine of this invention can significantly optimize the data fusion effect.

[0054] Experiment Example 2 Emergency Scenario Response Testing of the Scenario-Based Governance Center 1. Test parameters Test scenario: Simulates a combined emergency scenario of "red rainstorm warning + road flooding + traffic congestion" (based on real rainstorm event data from June 15, 2024, collected in the test dataset). Configuration of the scenario-based governance hub: scenario recognition threshold (water depth ≥ 30cm triggers "rainstorm emergency response scenario", congestion duration ≥ 10 minutes triggers "traffic diversion sub-scenario"), strategy generation parameters (for every 5cm increase in water depth, add 1 water pumping truck dispatch instruction, for every 5 minutes increase in congestion, expand the closure area of ​​1 road segment).

[0055] 2. Testing Process Input the following simulated data into the system: ① Meteorological data (a red rainstorm warning was issued at 14:00, expected to last for 2 hours); ② Traffic data (at 14:05, a main road had 32cm of water accumulation, with a congestion duration of 12 minutes, affecting 2 bus routes); ③ Emergency data (at 14:08, two alarms were triggered regarding backflow of water from surrounding residential areas along this road section). Record the entire process time from data access to the generation and issuance of instructions, and verify the matching degree between the instruction content and the scenario requirements (such as the number of water pumping trucks to be dispatched, bus route detour plans, and community evacuation instructions).

[0056] 3. Test Results Governance response time: 4 minutes 28 seconds (data access 1 minute 12 seconds → fusion 1 minute 05 seconds → scene recognition 0 minutes 41 seconds → policy generation 0 minutes 30 seconds → instruction issuance 0 minutes 20 seconds); Scene adaptation accuracy: 100% (The generated instructions include "dispatch 2 water pumping trucks to the target road section, temporarily detour 2 bus routes, and notify residents of 1 nearby community not to go out for the time being", which is a perfect match with the optimal instructions determined by manual judgment). Subsequent execution results: 30 minutes after the instruction was issued, the water depth dropped to 18cm, the congestion time was shortened to 3 minutes, and there were no new backflow alarms, proving that the scenario-based governance center can quickly generate instructions that are accurately adapted to emergency scenarios.

[0057] Experimental Example 3 Closed-loop optimization effect test of feedback optimization module 1. Test parameters Test scenario: Optimization of routine government services (focusing on the governance goal of "processing time for social security payment business", with the initial goal of "reducing the average processing time from 25 minutes to 15 minutes"); Feedback optimization module configuration: frequency of effect data collection (collect business processing time data of the government service hall once per hour), optimization algorithm (gradient descent algorithm, learning rate 0.01, iteration step size 1 hour), and optimization parameter output method (real-time update of the "window resource scheduling strategy" parameter of the scenario-based governance center).

[0058] 2. Testing Process Day 1: The system initially generates a basic strategy of "opening 8 social security payment windows, with each window serving an average of 50 people per day" and records the average processing time of the day (22 minutes). Days 2-7: The feedback optimization module collects processing time data daily. If the processing time exceeds the target (15 minutes), the optimization parameters are calculated using the gradient descent algorithm (e.g., "for every additional window, the processing time is expected to be shortened by 1.2 minutes"), and the strategy is updated (9 windows are opened on day 2, 10 windows are opened on day 4, and 2 online processing windows are added on day 6). Record the average daily processing time and the optimization rate of the strategy optimization results.

[0059] 3. Test Results Conclusion: After 6 days of closed-loop optimization, the average processing time for social security payments was reduced from 22 minutes to 14.5 minutes, achieving and exceeding the initial target. The feedback optimization rate reached a maximum of 14.0%, proving that the feedback optimization module can achieve continuous iterative improvement in governance effectiveness.

[0060] Comparative Example 1 Data fusion system based on static rules (corresponding to CN202410876543.2) 1. System Configuration The data fusion was performed using the "preset static association rules" disclosed in the patent. The number of rules was fixed at 15 (consistent with the initial rules in Experiment Example 1 of this invention, without self-learning function). There is no contextualized governance center; it only outputs static reports after data fusion. The no-feedback optimization module does not update the fusion rules and output content throughout the entire process.

[0061] 2. Test Results (same test period and same dataset as Experiment 1) Comparative Example 2 Basic statistical urban intelligent hub (corresponding to CN202310567890.1) 1. System Configuration It only has data storage (MySQL single-point storage) and basic statistical analysis functions (such as "average daily social security processing volume" and "top 10 congested road sections"). The governance output is a static statistical report, which requires manual generation of governance instructions based on the report. There is no dynamic fusion engine (data is stored only according to its source and is not associated with fusion) and no feedback optimization module.

[0062] 2. Test Results (Same emergency scenario as Experiment Example 2) Comparative Example 3 This invention removes the feedback optimization module from the system (comparison of missing core modules). 1. System Configuration It retains the multi-source data access layer, dynamic fusion engine, scenario-based governance hub, and dual databases; Remove the feedback optimization module, meaning that execution effect data will not be collected and fusion rules and governance strategies will not be updated (the strategies will remain fixed at the initial version).

[0063] 2. Test Results (Same government service scenario as Experiment 3, test period 7 days) Experimental conclusions Core technology advantages verification: Through the collaborative design of "dynamic fusion engine (including rule self-learning) + scenario-based governance center + feedback optimization module", this invention significantly outperforms existing technologies in data fusion efficiency (101.1% improvement compared to Comparative Example 1), scenario adaptation accuracy (100% vs. 0% of existing technologies), governance response time (71.7% reduction in emergency scenarios compared to Comparative Example 2), and continuous optimization capability (21.6% reduction in processing time for government service scenarios compared to Comparative Example 3). Module necessity verification: Comparative example 3 proves that after removing the feedback optimization module, the system cannot achieve continuous iteration of governance effect, further highlighting the necessity of the feedback optimization module as the "closed-loop core" and proving that the feedback optimization module combines the other modules into an organic whole; Authenticity and reproducibility: All experiments are based on a unified hardware / software environment, with a clear dataset and quantitative indicators. The data is reproducible, and direct comparison with existing technology patents (CN202410876543.2, CN202310567890.1) clearly demonstrates the advantages of this invention.

[0064] The above description is only a preferred embodiment, experimental example, and comparative example of the present invention, and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A system for dynamic governance of an urban intelligent hub driven by multi-source government data, characterized in that, It includes a multi-source data access layer, a dynamic fusion engine, a scenario-based governance hub, a feedback optimization module, and dual databases; The output of the multi-source data access layer is connected to the input of the dynamic fusion engine; The output of the dynamic fusion engine is connected to the input of the scenario-based governance center and the full data pool in the dual databases. The dynamic fusion engine is used to construct a dynamic mapping relationship of multi-source government data based on spatiotemporal tags and business tags to realize real-time classification and fusion of data. The output of the scenario-based governance center is connected to the input of the feedback optimization module and the execution unit at the end of the urban governance process, respectively. The scenario-based governance center is used to call the basic strategies in the governance strategy library based on the fused data characteristics and generate dynamic execution instructions that are adapted to the current governance scenario. The output of the feedback optimization module is connected to the input of the dynamic fusion engine, the input of the scenario-based governance center, and the governance strategy library in the dual databases, respectively. The feedback optimization module is used to collect the execution effect data of the governance end execution unit, generate optimization parameters and feed them back to the dynamic fusion engine and the scenario-based governance center to achieve closed-loop optimization of the governance chain; The governance strategy library of the dual databases is bidirectionally connected to the scenario-based governance center.

2. The system according to claim 1, characterized in that, The multi-source data access layer includes government service data interfaces, city operation data interfaces, emergency event data interfaces, and third-party collaboration data interfaces; The multi-source data access layer also includes a data preprocessing unit, which is used to standardize the format of the accessed data, remove outliers, and fill in missing values. The processed data is then transmitted to the dynamic fusion engine.

3. The system according to claim 2, characterized in that, Each data interface of the multi-source data access layer is equipped with an authentication unit. The authentication unit uses the OAuth2.0 protocol to authenticate the interface visitor and allows data access after successful authentication.

4. The system according to claim 1, characterized in that, The dynamic fusion engine includes a tag generation unit, a mapping rule base, and a real-time matching unit; The tag generation unit is used to generate unique spatiotemporal tags and business tags for the accessed multi-source government data. The spatiotemporal tags include the timestamp of the data generation and the administrative division code, and the business tags include the government affairs field and business type to which the data belongs. The mapping rule base stores dynamic mapping rules trained based on historical governance cases and government data characteristics; The real-time matching unit is used to call rules in the mapping rule base according to the spatiotemporal labels and business labels of the data to complete the real-time matching and fusion of data and governance scenarios.

5. The system according to claim 4, characterized in that, The mapping rule base includes a rule self-learning unit, which is used to iteratively train the mapping rules every quarter based on the historical fusion data in the full data pool and the optimization records of the feedback optimization module, and update the rule parameters in the mapping rule base.

6. The system according to claim 1, characterized in that, The scenario-based governance hub includes a scenario identification unit, a strategy generation unit, and an instruction issuance unit; The scene recognition unit is used to identify the type of the current urban governance scene based on the fusion data characteristics output by the dynamic fusion engine. The scene types include daily operation monitoring scene, emergency event handling scene, and government service optimization scene. The strategy generation unit is used to call the basic strategies in the governance strategy library, adjust the strategy parameters in combination with the real-time characteristics of the fused data, and generate dynamic governance strategies. The instruction issuing unit is used to convert dynamic governance strategies into standardized execution instructions and transmit them to the urban governance terminal execution unit.

7. The system according to claim 6, characterized in that, The instruction issuing unit includes an instruction verification unit, which is used to verify the execution instruction by calling the compliance rules in the governance strategy library before the instruction is issued. After the verification is passed, the instruction is transmitted to the city governance terminal execution unit.

8. The system according to claim 1, characterized in that, The feedback optimization module includes an effect data acquisition unit, an optimization analysis unit, and a parameter output unit; The effect data collection unit is used to collect execution progress data, target achievement rate data, and public feedback data from the end-level execution units of urban governance. The optimization analysis unit is used to compare the collected effect data with the expected effect data in the governance strategy library, and to calculate the mapping rule optimization parameters of the dynamic fusion engine and the strategy parameter optimization values ​​of the scenario-based governance center through the gradient descent algorithm. The parameter output unit is used to transmit the optimization parameters to the dynamic fusion engine and the scenario-based governance center, respectively.

9. The system according to claim 8, characterized in that, The effect data acquisition unit includes a data encryption unit, which uses the SM4 symmetric encryption algorithm to encrypt the acquired effect data. The encrypted data is then transmitted to the optimization analysis unit.

10. The system according to claim 1, characterized in that, The full data pool of the dual databases is used to store the fused data output by the dynamic fusion engine. The full data pool adopts a distributed storage architecture and supports fast data retrieval by spatiotemporal tags and business tags. The governance strategy library is used to store basic strategies for different governance scenarios. The governance strategy library supports strategy updates based on the optimization parameters of the feedback optimization module, and the update frequency is consistent with the optimization cycle of the feedback optimization module.

Images