Smart park early warning method and system based on multi-source heterogeneous data

Abstract

The invention relates to a smart park early warning method and system based on multi-source heterogeneous data, and the method comprises the following steps: S1, data collection, S2, data preprocessing, S3, risk early warning model construction and training, S4, risk identification and early warning, and S5, early warning response: generating corresponding early warning information and an emergency disposal scheme according to the risk type and the risk level, the early warning information is pushed to a park management terminal and a related person in charge, and park intelligent equipment is linked to execute emergency disposal operation; according to the method, data fusion is realized, the incidence relation among the multi-source data is fully mined, and comprehensive data support is provided for accurate early warning; multiple risk types such as equipment faults, environment safety and personnel abnormity can be accurately identified; the missing report rate and the false report rate are greatly reduced, and sufficient time is gained for risk disposal; the problems that a traditional early warning system is lagged in response and poor in collaboration are solved, and the park risk disposal efficiency and the management intelligence level are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of smart park management technology, specifically a smart park early warning method and system based on multi-source heterogeneous data. Background Technology

[0002] With the acceleration of urbanization and the rapid development of digital technology, smart parks, as comprehensive carriers integrating industry, living, and services, have become an important component of new urban construction. Smart parks deploy a large number of IoT devices, monitoring terminals, and business management systems, generating massive amounts of multi-source heterogeneous data. This includes equipment operation data, such as elevator operating parameters, power distribution equipment load, and air conditioning unit status; environmental sensing data, such as air quality, temperature, humidity, and concentration of toxic and harmful gases; personnel flow data, such as access control records, personnel trajectories in video surveillance, and visitor registration information; energy consumption data, such as electricity, water resources, and gas usage; and security alarm data, such as infrared alarms and smoke alarms.

[0003] However, the current systems within the park operate independently, hindering effective data sharing from multiple sources and reducing the comprehensiveness of early warning systems. Existing early warning methods often rely on single-dimensional data or simple threshold judgments, such as triggering fire warnings solely based on smoke sensor thresholds, leading to high rates of missed and false alarms and making early prediction difficult. Existing early warning systems mostly only provide alarm notifications, lacking intelligent classification of warning levels, risk tracing, and automatic matching of emergency response plans. This prevents park management personnel from quickly locating risk sources and accurately taking countermeasures, delaying risk response opportunities. Therefore, those skilled in the art provide a smart park early warning method and system based on multi-source heterogeneous data to address the problems mentioned in the background. Summary of the Invention

[0004] The purpose of this invention is to provide a smart park early warning method and system based on multi-source heterogeneous data, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A smart park early warning method based on multi-source heterogeneous data includes the following steps:

[0007] S1. Data Acquisition: Collect heterogeneous data within the smart park through multi-source data acquisition terminals. The heterogeneous data includes structured data, semi-structured data, and unstructured data.

[0008] S2. Data Preprocessing: Perform unified preprocessing on the collected multi-source heterogeneous data, including data cleaning, data transformation, data fusion and feature extraction;

[0009] S3. Risk Warning Model Construction and Training: Construct a multi-dimensional risk warning model based on deep learning. The warning model includes a feature input layer, a feature fusion layer, a risk identification layer, and a warning output layer. Input the preprocessed feature vectors into the warning model, train the model using historical risk data of the park and labeled samples, optimize the model parameters, and enable the model to identify and predict different types of risks.

[0010] S4. Risk Identification and Early Warning: Input the pre-processed feature vectors into the trained risk early warning model. The model identifies current and potential risks in the park and determines the risk type and risk level through multi-dimensional feature correlation analysis.

[0011] S5. Early Warning Response: Based on the risk type and risk level, generate corresponding early warning information and emergency response plans, push the early warning information to the park management terminal and relevant responsible persons, and coordinate with the park's smart devices to execute emergency response operations.

[0012] As a further aspect of the present invention: the multi-source data acquisition terminal includes an IoT sensing terminal, a video surveillance terminal, an access control terminal, an energy metering terminal, and a mobile acquisition terminal; the IoT sensing terminal is used to collect ambient temperature and humidity, air quality, concentration of toxic and harmful gases, and equipment operating current / voltage / temperature parameters; the video surveillance terminal is used to collect video and image data of public areas, equipment rooms, and entrances / exits of the park; the access control terminal is used to collect personnel entry and exit records and identity information; and the energy metering terminal is used to collect real-time consumption data of electricity, water resources, and gas.

[0013] As a further aspect of the present invention: the structured data includes equipment operating parameters, access control records, energy consumption data, and visitor registration information; the semi-structured data includes equipment logs and system operation logs; and the unstructured data includes video surveillance data, image data, audio data, and environmental perception images.

[0014] As a further aspect of the present invention: the data cleaning is used to remove abnormal data, fill in missing data, and remove data redundancy; the data conversion is used to convert heterogeneous data of different formats into a unified data format; the data fusion includes feature-based fusion and decision-based fusion, feature-based fusion is used to extract key features from each data source and splice them together, and decision-based fusion is used to perform weighted fusion of the analysis results from each data source; the feature extraction is used to extract feature vectors related to park risks from the preprocessed data, including equipment failure features, environmental anomaly features, personnel anomaly features, and energy anomaly features.

[0015] As a further aspect of the present invention: the risk types include equipment failure risk, environmental safety risk, personnel safety risk, and energy leakage risk; the risk levels are divided into three levels: general warning, relatively severe warning, and serious warning.

[0016] As a further aspect of the present invention: In step S5, the emergency response plan includes risk tracing information, contact information of the responsible person, and guidelines for the response process; the linkage of park intelligent equipment to perform emergency response operations includes: when a fire risk is identified, linking fire alarm equipment to issue an alarm, start the sprinkler system, and close the gas valve; when an equipment failure risk is identified, linking the equipment management system to issue a shutdown prompt and push a maintenance work order; when an abnormal flow of personnel is identified, linking the video surveillance terminal to focus on the target area and push early warning information to the security personnel terminal.

[0017] A smart park early warning system based on multi-source heterogeneous data includes: a multi-source data acquisition module, a data preprocessing module, an early warning model training module, a risk identification and early warning module, an early warning response module, and a data storage module;

[0018] The multi-source data acquisition module includes an IoT sensing unit, a video surveillance unit, an access control unit, an energy metering unit, and a log acquisition unit; the multi-source data acquisition module is used to collect structured data, semi-structured data, and unstructured data within the smart park through a multi-source data acquisition terminal;

[0019] The data preprocessing module includes a data cleaning unit, a format conversion unit, a feature extraction unit, and a data fusion unit. The data preprocessing module is used to perform data cleaning, data conversion, data fusion, and feature extraction on the collected multi-source heterogeneous data, and output feature vectors related to park risks.

[0020] The early warning model training module is used to construct a multi-dimensional risk early warning model based on deep learning, and uses historical risk data of the park and labeled samples to train and optimize the model;

[0021] The risk identification and early warning module is used to input real-time feature vectors into the trained early warning model, identify risk types and risk levels, and generate early warning information.

[0022] The early warning response module is used to generate an emergency response plan based on the early warning information, push the early warning information to relevant terminals, and coordinate with the park's smart devices to perform emergency response operations.

[0023] The data storage module is used to store the collected raw data, preprocessed data, model training data, early warning records, and emergency response logs.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] 1. This application solves the problem of independent and inconsistent data formats in traditional industrial parks by using a unified data acquisition terminal and preprocessing process; it realizes data fusion, fully explores the correlation between multi-source data, and provides comprehensive data support for accurate early warning.

[0026] 2. This application adopts a multi-task early warning model based on deep learning, which can accurately identify various risk types such as equipment failure, environmental safety, and personnel abnormalities; it can make early predictions of potential risks. Compared with traditional single threshold early warning methods, it significantly reduces the false alarm and missed alarm rates, and buys sufficient time for risk disposal.

[0027] 3. This application can automatically match emergency response plans according to risk levels, link park smart devices to perform emergency operations, and push early warning information to relevant responsible persons through multiple channels. This solves the problems of delayed response and poor coordination of traditional early warning systems, and greatly improves the efficiency of risk handling and the level of intelligent management in the park. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of a smart park early warning method and system based on multi-source heterogeneous data. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Please see Figure 1 In this embodiment of the invention, a smart park early warning method based on multi-source heterogeneous data includes the following steps:

[0031] S1. Data Acquisition: Collect heterogeneous data within the smart park through multi-source data acquisition terminals. The heterogeneous data includes structured data, semi-structured data, and unstructured data.

[0032] The structured data includes equipment operating parameters, access control records, energy consumption data, and visitor registration information;

[0033] The semi-structured data includes device logs and system operation logs;

[0034] The unstructured data includes video surveillance data, image data, audio data, and environmentally conscious images;

[0035] S2. Data Preprocessing: Perform unified preprocessing on the collected multi-source heterogeneous data, including data cleaning, data transformation, data fusion and feature extraction;

[0036] The data cleaning is used to remove abnormal data, fill in missing data, and remove data redundancy;

[0037] The data conversion is used to convert heterogeneous data of different formats into a unified data format;

[0038] The data fusion includes feature-based fusion and decision-based fusion. Feature-based fusion is used to extract key features from each data source and concatenate them, while decision-based fusion is used to perform weighted fusion of the analysis results from each data source.

[0039] The feature extraction is used to extract feature vectors related to park risks from the preprocessed data, including equipment failure features, environmental anomaly features, personnel anomaly features, and energy anomaly features.

[0040] S3. Risk warning model construction and training: Construct a multi-dimensional risk warning model based on deep learning. The warning model includes a feature input layer, a feature fusion layer, a risk identification layer, and a warning output layer.

[0041] The preprocessed feature vectors are input into the early warning model. The model is trained using historical risk data of the park and labeled samples to optimize the model parameters, so that the model has the ability to identify and predict different types of risks.

[0042] S4. Risk Identification and Early Warning: Input the pre-processed feature vectors into the trained risk early warning model. The model identifies current and potential risks in the park and determines the risk type and risk level through multi-dimensional feature correlation analysis.

[0043] The risk types include equipment failure risk, environmental safety risk, personnel safety risk, and energy leakage risk.

[0044] The risk levels are divided into three levels: general warning, relatively severe warning, and serious warning.

[0045] S5. Early Warning Response: Based on the risk type and risk level, generate corresponding early warning information and emergency response plans, push the early warning information to the park management terminal and relevant responsible persons, and coordinate with the park's smart devices to execute emergency response operations.

[0046] The multi-source data acquisition terminal includes an IoT sensing terminal, a video surveillance terminal, an access control terminal, an energy metering terminal, and a mobile acquisition terminal. The IoT sensing terminal is used to collect ambient temperature and humidity, air quality, concentration of toxic and harmful gases, and equipment operating current / voltage / temperature parameters. The video surveillance terminal is used to collect video and image data from public areas, equipment rooms, and entrances / exits of the park. The access control terminal is used to collect personnel entry and exit records and identity information. The energy metering terminal is used to collect real-time consumption data of electricity, water resources, and gas.

[0047] In step S5, the emergency response plan includes risk tracing information, contact information of responsible persons, and guidelines for handling procedures. The coordinated emergency response operations involving the park's intelligent equipment include: when a fire risk is identified, triggering the fire alarm equipment to issue an alarm, activating the sprinkler system, and closing the gas valve; when an equipment malfunction risk is identified, triggering the equipment management system to issue a shutdown notice and push a maintenance work order; and when an abnormal flow of personnel is identified, triggering the video surveillance terminal to focus on the target area and push warning information to security personnel terminals.

[0048] A smart park early warning system based on multi-source heterogeneous data includes: a multi-source data acquisition module, a data preprocessing module, an early warning model training module, a risk identification and early warning module, an early warning response module, and a data storage module;

[0049] The multi-source data acquisition module includes an IoT sensing unit, a video surveillance unit, an access control unit, an energy metering unit, and a log acquisition unit; the multi-source data acquisition module is used to collect structured data, semi-structured data, and unstructured data within the smart park through a multi-source data acquisition terminal;

[0050] The data preprocessing module includes a data cleaning unit, a format conversion unit, a feature extraction unit, and a data fusion unit. The data preprocessing module is used to perform data cleaning, data conversion, data fusion, and feature extraction on the collected multi-source heterogeneous data, and output feature vectors related to park risks.

[0051] The early warning model training module is used to construct a multi-dimensional risk early warning model based on deep learning, and uses historical risk data of the park and labeled samples to train and optimize the model;

[0052] The risk identification and early warning module is used to input real-time feature vectors into the trained early warning model, identify risk types and risk levels, and generate early warning information.

[0053] The early warning response module is used to generate an emergency response plan based on the early warning information, push the early warning information to relevant terminals, and coordinate with the park's smart devices to perform emergency response operations.

[0054] The data storage module is used to store the collected raw data, preprocessed data, model training data, early warning records, and emergency response logs.

[0055] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A smart park early warning method based on multi-source heterogeneous data, characterized in that, Includes the following steps: S1. Data Acquisition: Collect heterogeneous data within the smart park through multi-source data acquisition terminals. The heterogeneous data includes structured data, semi-structured data, and unstructured data. S2. Data Preprocessing: Perform unified preprocessing on the collected multi-source heterogeneous data, including data cleaning, data transformation, data fusion and feature extraction; S3. Risk warning model construction and training: Construct a multi-dimensional risk warning model based on deep learning. The warning model includes a feature input layer, a feature fusion layer, a risk identification layer, and a warning output layer. The preprocessed feature vectors are input into the early warning model. The model is trained using historical risk data of the park and labeled samples to optimize the model parameters, so that the model has the ability to identify and predict different types of risks. S4. Risk Identification and Early Warning: Input the pre-processed feature vectors into the trained risk early warning model. The model identifies current and potential risks in the park and determines the risk type and risk level through multi-dimensional feature correlation analysis. S5. Early Warning Response: Based on the risk type and risk level, generate corresponding early warning information and emergency response plans, push the early warning information to the park management terminal and relevant responsible persons, and coordinate with the park's smart devices to execute emergency response operations.

2. The smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The multi-source data acquisition terminal includes an IoT sensing terminal, a video surveillance terminal, an access control terminal, an energy metering terminal, and a mobile acquisition terminal. The IoT sensing terminal is used to collect ambient temperature and humidity, air quality, concentration of toxic and harmful gases, and equipment operating current / voltage / temperature parameters. The video surveillance terminal is used to collect video and image data of public areas, equipment rooms, and entrances / exits in the park. The access control terminal is used to collect personnel entry and exit records and identity information. The energy metering terminal is used to collect real-time consumption data of electricity, water resources, and gas.

3. The smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The structured data includes equipment operating parameters, access control records, energy consumption data, and visitor registration information; the semi-structured data includes equipment logs and system operation logs; and the unstructured data includes video surveillance data, image data, audio data, and environmental perception images.

4. The smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The data cleaning is used to remove abnormal data, fill in missing data, and remove data redundancy; the data transformation is used to convert heterogeneous data of different formats into a unified data format; the data fusion includes feature-based fusion and decision-based fusion, feature-based fusion is used to extract key features from each data source and splice them, and decision-based fusion is used to perform weighted fusion of the analysis results from each data source; the feature extraction is used to extract feature vectors related to park risks from the preprocessed data, including equipment failure features, environmental anomaly features, personnel anomaly features, and energy anomaly features.

5. The smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The risk types include equipment failure risk, environmental safety risk, personnel safety risk, and energy leakage risk; the risk levels are divided into three levels: general warning, relatively severe warning, and serious warning.

6. The smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, In step S5, the emergency response plan includes risk tracing information, contact information of the responsible person, and guidelines for the response process; The emergency response operations performed by the park's intelligent equipment include: when a fire risk is detected, triggering the fire alarm equipment to sound an alarm, activating the sprinkler system, and shutting off the gas valve; When a risk of equipment failure is detected, the system will issue a shutdown notice and push a maintenance work order; when a risk of abnormal personnel movement is detected, the system will focus on the target area and push early warning information to the security personnel's terminal.

7. The system of the smart park early warning method based on multi-source heterogeneous data according to any one of claims 1-6, characterized in that, include: The system includes a multi-source data acquisition module, a data preprocessing module, an early warning model training module, a risk identification and early warning module, an early warning response module, and a data storage module. The multi-source data acquisition module includes an IoT sensing unit, a video surveillance unit, an access control unit, an energy metering unit, and a log acquisition unit; the multi-source data acquisition module is used to collect structured data, semi-structured data, and unstructured data within the smart park through a multi-source data acquisition terminal; The data preprocessing module includes a data cleaning unit, a format conversion unit, a feature extraction unit, and a data fusion unit. The data preprocessing module is used to perform data cleaning, data conversion, data fusion, and feature extraction on the collected multi-source heterogeneous data, and output feature vectors related to park risks. The early warning model training module is used to construct a multi-dimensional risk early warning model based on deep learning, and uses historical risk data of the park and labeled samples to train and optimize the model.

8. The system of a smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The risk identification and early warning module is used to input real-time feature vectors into the trained early warning model, identify risk types and risk levels, and generate early warning information.

9. The system of a smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The early warning response module is used to generate an emergency response plan based on the early warning information, push the early warning information to relevant terminals, and coordinate with the park's smart devices to perform emergency response operations.

10. The system of a smart park early warning method based on multi-source heterogeneous data according to claim 1, characterized in that, The data storage module is used to store the collected raw data, preprocessed data, model training data, early warning records, and emergency response logs.

Images