Method and system for AI practical training of smart city based on big data

Abstract

The invention relates to the technical field of smart city practical training, in particular to an AI smart city practical training method and system based on big data, and the method comprises an operation program which comprises the following steps: S1, collecting city information data based on multiple channels; step S2, data screening; s3, judging data validity in combination with the Internet of Things; step S4, modeling is carried out; and S5, carrying out city analysis on the model. According to the method, the operation program with nine steps is set, city information data are obtained in a diversified mode for cities, the city model with the simulation effect is established according to the data, and the man-machine interaction machine is set through the AI technology on the basis, so that real man-machine interaction experience can be conveniently carried out; the program content can be continuously enriched by adding big data; the AI smart city practical training method can be continuously enriched and innovated, diversified experience modes are increased, and the practical training experience feeling is improved.

Description

Technical Field

[0001] The present invention relates to the field of smart city training technology, and in particular to a method and system for AI smart city training based on big data. Background Art

[0002] Smart city often intersects with regional development concepts such as digital city, sensory city, wireless city, intelligent city, ecological city, low-carbon city, and even mixes with industry information concepts such as e-government, intelligent transportation, and smart grid. The interpretation of the concept of smart city often has different focuses. Some people believe that the key lies in technology application, some believe that the key lies in network construction, some believe that the key lies in human participation, and some believe that the key lies in smart effects. Some leading cities in urban information construction emphasize people-oriented and sustainable innovation. In short, wisdom is not just intelligence. Smart city is not just another term for smart city, or the intelligent application of information technology, but also includes human wisdom participation, people-oriented, sustainable development and other connotations. The "smart city" in the 21st century can make full use of information and communication technology to sense, analyze, and integrate various key information of the core system of urban operation, so as to make intelligent responses to various needs including people's livelihood, environmental protection, public safety, urban services, industrial and commercial activities, and create a better urban life for mankind.

[0003] At present, the rapid development of AI technology has provided a technical foundation for smart cities. In order to enrich the development model of smart cities, AI is often combined with practical training under existing conditions. For example, a smart city evaluation AI analysis system and a smart city evaluation method disclosed in announcement number CN109636132B provide a smart city evaluation AI analysis system that can be operated according to settings, which can analyze the mutual influencing factors between various industries in the operation of smart cities, access various core key data of the city operation system, and visualize these data at the same time, so as to provide management decision support for (including but not limited to) emergency command, urban management, public safety, environmental protection, intelligent transportation, infrastructure and other fields, so as to avoid snap decisions to the greatest extent, improve the healthy and orderly development of smart cities, and enhance the pertinence of construction suggestions for different smart cities. At the same time, the self-learning ability of the smart city evaluation AI analysis system can promote the construction of smart cities according to practice and development needs, and constantly correct the influencing factors in real time to apply to urban construction and operation in different development periods. The same as the video real-time processing AI engine system for smart city management disclosed in announcement number CN116382813B, which can provide rapid assembly capabilities based on workflow models, is suitable for AI engine customization in complex video stream scenarios, and realizes efficient, low-cost, and diversified empowerment of video stream processing, reduces the cost and cycle of video stream empowerment, and ensures the practicality of the AI ​​engine intelligent analysis platform. However, the above only uses AI to process data on the city, lacks high-dimensional analytical thinking, so it can only operate based on actual conditions, and also ignores important influencing factors of human civilization creation. In this process, the urban force model is not combined with human behavior, resulting in the single development of urban civilization following the trajectory set by AI, which will greatly reduce the opportunities for human behavior to interact with it, and it is impossible to truly predict the development direction of the city. It is also difficult to guarantee the experience and satisfaction during the training process. Summary of the invention

[0004] The purpose of the present invention is to solve the problem in the prior art that it is difficult to provide authenticity and diversity experience for human-computer interaction in establishing city models, and to propose a method and system for AI training smart cities based on big data.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for AI training smart city based on big data, including an operating procedure, wherein the operating procedure includes the following steps:

[0007] Step S1, collecting city information data based on multiple channels;

[0008] Step S2, data screening;

[0009] Step S3, judging the validity of data in combination with the Internet of Things;

[0010] Step S4, modeling;

[0011] Step S5, performing city analysis on the model;

[0012] Step S6, human-machine operation;

[0013] Step S7, AI performs city analysis on the actual operation behavior;

[0014] Step S8, record and save;

[0015] Step S9, expanding the database.

[0016] Preferably, the collecting of city information data based on multiple channels refers to obtaining comprehensive information data for the city, and the data screening refers to filtering and sorting the obtained comprehensive information data.

[0017] Preferably, the judging of data validity in combination with the Internet of Things refers to judging whether the sorted data has value for modeling by comparing it with the data in the Internet of Things, and the modeling refers to modeling using data that has been judged to have modeling value.

[0018] Preferably, the city analysis based on the model refers to analyzing the city model after modeling, and the human-computer operation refers to human-computer interaction experience based on the city model.

[0019] Preferably, the AI ​​performing city analysis on practical operation behavior refers to analyzing the human-computer interaction experience through AI, the record preservation refers to preserving the human-computer interaction experience with reference value, and the database expansion refers to integrating the human-computer interaction experience with reference value into the city model.

[0020] A system that applies the above-mentioned method of AI training smart city based on big data includes a front-end management machine and a training cloud platform. The training cloud platform includes a data acquisition terminal, a data analyzer, a central processing unit and a management and control platform. The data acquisition terminal, the data analyzer, and the central processing unit are all signal-connected to the management and control platform, and the front-end management machine is electrically connected to the training cloud platform.

[0021] Preferably, the data collection terminal includes a primary information source and a secondary information source;

[0022] The data analyzer includes a data analysis and screening module, a data fusion module and a data modeling module;

[0023] The management and control platform includes an operation module, a human-computer interaction machine, a fault processing machine, a record analysis module and a capacity expansion module.

[0024] Preferably, the first-hand information sources include field investigations, questionnaire surveys and urban monitoring systems, and the second-hand information sources include internal historical data sources, statistical materials and official announcements;

[0025] The data modeling module includes an environment modeling section, a still life modeling section and an animal modeling section;

[0026] The fault handling machine includes an intelligent inspection machine, a fault repair device and a manual team;

[0027] The expansion module includes a city expansion module, an interactive expansion module, a blending module and a learning module.

[0028] Preferably, the number of the human-computer interaction machines is ten per group, and the number is at least ten groups.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] 1. The present invention sets an operation procedure with nine steps, so that firstly, city information data is obtained in a diversified manner for the city, and a city model with a simulation effect is established based on the data. On this basis, a human-computer interaction machine is set up through AI technology to facilitate a real human-computer interaction experience, and the program content can be continuously enriched by adding big data.

[0031] 2. The present invention enriches and diversifies data acquisition methods and uses AI to implement screening and extraction of information data, so as to simulate the real situation of the city to a higher degree, establish a model after directional fusion, and also can test various urban conditions in this way.

[0032] 3. The present invention provides a large number of human-computer interaction machines to facilitate a large number of human-computer interaction experiences. While obtaining economic value, it can also provide more data references for urban information data, which helps to choose a variety of urban development methods.

[0033] 4. The present invention provides updated data for the city model and sets up a learning module with learning and imitation capabilities, so that the method of AI training smart city can be continuously enriched and innovated, increase the diversity of experience methods, and enhance the training experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A flowchart of a method for AI training smart cities based on big data proposed by the present invention;

[0035] Figure 2 A schematic diagram of the structure of a system for applying a method for AI training smart cities based on big data proposed by the present invention;

[0036] Figure 3A schematic diagram of the structure of primary and secondary information sources of a system for applying a method for AI training smart cities based on big data proposed by the present invention;

[0037] Figure 4 A schematic diagram of the data modeling module structure of a system for applying a method for AI training smart cities based on big data proposed by the present invention;

[0038] Figure 5 A schematic diagram of the structure of a fault processing machine of a system for applying a method for AI training smart cities based on big data proposed by the present invention;

[0039] Figure 6 This is a schematic diagram of the expansion module structure of a system for applying a method for AI training smart cities based on big data proposed in the present invention.

[0040] In the figure:

[0041] 1. Front-end management machine;

[0042] 2. Training PTZ;

[0043] 201. Data collection terminal;

[0044] 20101, first-hand information source; 2010101, field investigation; 2010102, questionnaire survey; 2010103, urban monitoring system;

[0045] 20102, Second-hand information sources; 2010201, Internal historical data; 2010202, Statistical materials; 2010203, Official announcements;

[0046] 202. Data analyzer;

[0047] 20201. Data analysis and screening module; 20202. Data fusion module;

[0048] 20203, data modeling module; 2020301, environment modeling module; 2020302, still life modeling module; 2020303, animal modeling module;

[0049] 203, central processing unit;

[0050] 204, management and control platform; 20401, operation module; 20402, human-computer interaction machine;

[0051] 20403, fault handling machine; 2040301, intelligent inspection machine; 2040302, fault repair instrument; 2040303, manual team;

[0052] 20404, record analysis module;

[0053] 20405, expansion module; 2040501, city expansion module; 2040502, interactive expansion module; 2040503, integration module; 2040504, learning module. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0055] Reference Figure 1-Figure 6 , a method for AI training smart city based on big data, including an operation procedure, the operation procedure including the following steps:

[0056] Step S1, collecting urban information data based on multiple channels. It should be noted that the first-hand information on urban development is obtained through field investigation 2010101, questionnaire survey 2010102 and urban monitoring system 2010103, and the second-hand information on urban development is obtained through internal historical data source 2010201, statistical materials 2010202 and official announcements 2010203.

[0057] Step S2, data screening, uses the data screening module 20201 to identify and judge the primary information and the secondary information, selects the information data with practical value, and uses the data fusion module 20202 to fuse the data information. In the data fusion stage, in order to enhance the consistency of the data, the fusion of the data after the distributed resource configuration is realized. The data after the distributed resource configuration includes video analysis results and traffic flow data; introduces a fusion enhancement algorithm, which aims to adjust and unify the data format and scale of different data sources through an adaptive adjustment mechanism.

[0058] The specific implementation can be expressed as: Vadjusted = θ (Vsource) Voriginal + β (Vsource, Vtarget)

[0059] in:

[0060] Vadjusted is the preliminary fused dataset with improved data consistency;

[0061] Voriginal is the original data value, which comes from video surveillance, traffic sensors, and environmental monitoring equipment, reflecting the real-time status of the city;

[0062] Vsource is the type of data source, indicating what kind of monitoring device the data comes from;

[0063] Vtarget is the target data format, which indicates the unified format and standard expected to be achieved after data fusion;

[0064] α(·) is a scale adjustment function, which adjusts the scale of the data according to the different data sources to achieve consistency between data from different sources.

[0065] For example, traffic flow data and environmental monitoring data have different scales and need to be adjusted for unified processing.

[0066] α(Vsource)=asource·Vsource+bsource

[0067] Among them, asource and bsource are adjustment coefficients based on the data source type Vsource;

[0068] Vsource is the data source type, such as video surveillance and traffic monitoring;

[0069] asource is the scale adjustment coefficient corresponding to the data source type, which is used to adjust the scale of the data to make it compatible with other data sources;

[0070] bsource is the offset corresponding to the data source type, which is used to further fine-tune the data for consistency;

[0071] β(Vsource, Vtarget) is an offset adjustment function that adjusts the format deviation of the data according to the data source type and the target data format. For example, the timestamp format in the video surveillance analysis result is unified with the timestamp format of the traffic monitoring system.

[0072] β(Vsource,Vtarget)=csource,target·Vsource+dsource,target

[0073] Among them, csource,target and dsource,target are adjustment coefficients based on the data source type and the target data format, and are used to adjust the format deviation of the data according to the data source type and the target data format;

[0074] In order to avoid inefficient data fusion when processing large-scale data.

[0075] Step S3, judging the validity of data in combination with the Internet of Things, and combining it with the real-time situation of urban development to ensure the validity and authenticity of the information data.

[0076] Step S4, modeling, by setting the data modeling module 20203, determine the functional attributes of the required city model according to the behavior event target parameters, such as selecting the name of the model, the path where the model is stored, and the threshold of the initialization model. Input data structure, the data is a BGR image frame; output data structure, the output data is a list of target information in the image.

[0077] Step S5, conduct city analysis on the model, and use AI to conduct comprehensive data verification on the generated city model to determine the differences between it and the real-time development situation.

[0078] Step S6, human-computer operation, realizes online communication between humans and AI technology through human-computer interaction machine 20402. Both ERP and computer kiosks use today's popular graphical user interface (GUI), voice user interface (VUI) is used for speech recognition and synthesis systems, and emerging multimodal and graphical user interfaces (GUI) allow humans to interact with specific role agents in a way that other interface paradigms cannot achieve. Users can interact directly with hardware for human input and output such as displays through a graphical user interface. Users interact with computers through this software interface using given input and output (I / O) hardware. The software and hardware are matched, so the processing of user input is fast enough and the delay of computer output does not disrupt the workflow.

[0079] In step S7, AI performs city analysis on actual operation behaviors.

[0080] Step S8, record preservation. After the human-computer interaction is realized, different training experiences will be formed. Big data comparison and judgment are performed through the record analysis module 20404 to retain human-computer interaction experience information data with reference value.

[0081] Step S9, expanding the database, uploading the human-computer interaction experience information data with reference value to the central processor 203 through the expansion module 20405 for permanent backup, and filling it into the city model at the same time.

[0082] The system using the above-mentioned AI training smart city method based on big data includes a front-end management machine 1 and a training cloud platform 2. By setting the front-end management machine 1, manual back-end management of all equipment and instruments in the training cloud platform 2 can be performed.

[0083] The training gimbal 2 includes a data acquisition terminal 201, a data analyzer 202, a central processing unit 203 and a management and control platform 204. The data acquisition terminal 201, the data analyzer 202, and the central processing unit 203 are all signal-connected to the management and control platform 204, and the front-end management machine 1 is electrically connected to the training gimbal 2, so as to realize the uploading and downloading of instructions and the forward and backward transmission of information.

[0084] The data collection terminal 201 includes primary information sources 20101 and secondary information sources 20102. The primary information sources 20101 include field investigations 2010101, questionnaire surveys 2010102 and urban monitoring systems 2010103. The secondary information sources 20102 include internal historical data sources 2010201, statistical materials 2010202 and official announcements 2010203.

[0085] The data analyzer 202 includes a data analysis screening module 20201, a data fusion module 20202 and a data modeling module 20203. The data modeling module 20203 includes an environmental modeling module 2020301, a still life modeling module 2020302 and an animal modeling module 2020303. By separately modeling moving objects and moving objects and then organically combining the two, the diversification of urban models can be achieved.

[0086] The management and control platform 204 includes an operation module 20401, a human-computer interaction machine 20402, a fault processing machine 20403, a record analysis module 20404 and a capacity expansion module 20405. The fault processing machine 20403 includes an intelligent inspection machine 2040301, a fault repair instrument 2040302 and a manual team 2040303;

[0087] The expansion module 20405 includes a city expansion template block 2040501, an interactive expansion block 2040502, an integration block 2040503 and a learning block 2040504.

[0088] The number of human-computer interaction machines 20402 is ten in a group, and the number is at least ten groups, which helps to realize multi-person synchronous training and provide more information data for big data.

[0089] It should be noted that the specific models and specifications of the front-end management machine 1, data analyzer 202, central processing unit 203, fault handling machine 20403, human-computer interaction machine 20402, intelligent inspection machine 2040301 and fault repair device 2040302 need to be selected and determined based on the actual specifications of the device. The specific selection and calculation method adopts the existing technology in this field, so it will not be repeated.

[0090] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for AI training smart city based on big data, including an operating procedure, characterized in that: The operating procedure comprises the following steps: Step S1, collecting city information data based on multiple channels; Step S2, data screening; Step S3, judging the validity of data in combination with the Internet of Things; Step S4, modeling; Step S5, performing city analysis on the model; Step S6, human-machine operation; Step S7, AI performs city analysis on the actual operation behavior; Step S8, record and save; Step S9, expanding the database.

2. According to the method of AI training smart city based on big data according to claim 1, it is characterized in that: The collecting of city information data based on multiple channels refers to obtaining comprehensive information data for the city, and the data screening refers to filtering and sorting the obtained comprehensive information data.

3. The method of AI training smart city based on big data according to claim 2 is characterized in that: The combination of the Internet of Things to determine the validity of data refers to determining whether the data sorted out by comparing it with the data in the Internet of Things has the value for modeling, and the modeling refers to modeling using the data that has been determined to have modeling value.

4. The method of AI training smart city based on big data according to claim 3 is characterized in that: The city analysis based on the model refers to analyzing the city model after modeling, and the human-computer operation refers to human-computer interaction experience based on the city model.

5. The method of AI training smart city based on big data according to claim 4 is characterized in that: The AI ​​conducting city analysis on practical operations refers to analyzing the human-computer interaction experience through AI, the record preservation refers to preserving the human-computer interaction experience with reference value, and the database expansion refers to integrating the human-computer interaction experience with reference value into the city model.

6. A system involving a method for AI training smart city based on big data as described in claim 5, comprising a front-end management machine (1) and a training cloud platform (2), characterized in that: The training platform (2) comprises a data acquisition terminal (201), a data analyzer (202), a central processing unit (203) and a control platform (204); the data acquisition terminal (201), the data analyzer (202) and the central processing unit (203) are all connected to the control platform (204) by signals, and the front-end management machine (1) is electrically connected to the training platform (2).

7. The system of the method for AI training smart city based on big data according to claim 6 is characterized in that: The data collection terminal (201) includes a primary information source (20101) and a secondary information source (20102); The data analyzer (202) comprises a data analysis and screening module (20201), a data fusion module (20202) and a data modeling module (20203); The management and control platform (204) includes an operation module (20401), a human-computer interaction machine (20402), a fault processing machine (20403), a record analysis module (20404) and a capacity expansion module (20405).

8. The system of the method for AI training smart city based on big data according to claim 7 is characterized in that: The first-hand information sources (20101) include field investigations (2010101), questionnaire surveys (2010102) and urban monitoring systems (2010103), and the second-hand information sources (20102) include internal historical data sources (2010201), statistical materials (2010202) and official announcements (2010203); The data modeling module (20203) includes an environment modeling section (2020301), a still life modeling section (2020302) and an animal modeling section (2020303); The fault handling machine (20403) includes an intelligent inspection machine (2040301), a fault repair device (2040302) and a manual team (2040303); The expansion module (20405) includes a city expansion template block (2040501), an interactive expansion block (2040502), an integration block (2040503) and a learning block (2040504).

9. The system of the method for AI training smart city based on big data according to claim 8 is characterized in that: The number of the human-computer interaction machines (20402) is ten per group, and the number is at least ten groups.

Images