Land resource digital intelligent management system and method based on Internet of Things

Abstract

The invention relates to the technical field of karst geological relics protection, in particular to an intelligent monitoring and evaluation system and method for a karst geological relics protection area, and the system integrates multi-source remote sensing data fusion, a GIS geographic information system, a space-time big data model, a deep learning algorithm and a block chain technology. An intelligent monitoring and evaluation system for the karst geological relics protection area is constructed; the system comprises a basic data acquisition module, a spatial information analysis module, a tourist flow evaluation module, a geological disaster early warning module, an ecological protection evaluation module, a real-time data monitoring module, a block chain storage module and a mobile science popularization module, and integration of data acquisition, analysis, storage and science popularization is realized. Through a three-dimensional digital model and real-time sensor data, geological relics spatial distribution, tourist dynamics and ecological changes are accurately analyzed, and tourist bearing capacity management and landslide risk early warning capability are improved. The block chain technology ensures data traceability, and the mobile science popularization module enhances tourist interaction experience.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical fields of Internet of Things, artificial intelligence and land resource management, and particularly relates to a land resource digital intelligent management system and method based on Internet of Things. BACKGROUND

[0002] With the rapid development of social economy, the scarcity and importance of land resources are increasingly highlighted. Traditional land resource management methods mainly rely on manual field survey and paper-based archive management, which has problems such as long data acquisition cycle, low precision, poor real-time performance, and serious data island, and is difficult to meet the needs of modern land resource fine management.

[0003] In recent years, the rapid development of new generation information technologies such as Internet of Things, artificial intelligence, big data and cloud computing provides a technical basis for the digitalization and intelligent transformation of land resource management. The existing technology usually uses a single sensor for data collection, and the data transmission and processing capacity is limited, which is difficult to realize the fusion analysis of multi-source heterogeneous data and cannot meet the demand of land resource dynamic monitoring in complex environment.

[0004] At present, although some areas have begun to try to use information technology for land resource management, there are generally problems such as single sensing means, difficult data sharing, limited analysis capability and insufficient decision support. Especially in complex terrain conditions, traditional monitoring means are difficult to fully perceive the state of land resources, resulting in lack of scientific basis for management decision-making.

[0005] Therefore, how to build a system that can fully perceive the state of land resources, efficiently transmit massive data, intelligently analyze complex data and scientifically support management decision-making has become a key technical problem to be solved in the field of land resource management. SUMMARY

[0006] The present application relates to the technical fields of Internet of Things, artificial intelligence and land resource management, and particularly relates to a land resource digital intelligent management system and method based on Internet of Things.

[0007] The application provides a land resource digital intelligent management system based on the Internet of Things, comprising: a multi-dimensional intelligent sensing module for collecting land resource utilization space information and time dynamic information; a heterogeneous network transmission module in communication connection with the multi-dimensional intelligent sensing module, for receiving multi-source land resource data sent by the multi-dimensional intelligent sensing module and transmitting the multi-source land resource data through a network; an intelligent data processing module in communication connection with the heterogeneous network transmission module, for fusing and analyzing the multi-source land resource data to obtain a processing result; and a decision support module in communication connection with the intelligent data processing module, for visualizing output based on the processing result and providing land resource management suggestions; wherein the decision support module sends management suggestion feedback information to the multi-dimensional intelligent sensing module and the heterogeneous network transmission module for parameter adjustment.

[0008] As a preference, the multi-dimensional intelligent sensing module comprises: an image acquisition unit for collecting high-resolution images of land resources and performing image fusion and feature extraction on the high-resolution images; an unmanned aerial vehicle monitoring unit for realizing intelligent inspection according to spatial and temporal changes of land resources through an unmanned aerial vehicle formation system; a sensor monitoring unit for obtaining soil and meteorological monitoring data through intelligent soil monitoring and a micro weather station array; and a positioning unit for obtaining positioning data of the location of land resources, land resource space planning data and right confirmation data.

[0009] As a preference, the heterogeneous network transmission module comprises: a 5G network transmission unit for realizing high-rate data transmission through a 5G network; a LoRa network transmission unit for realizing long-distance low-power consumption data transmission through a LoRa network; an NB-IoT network transmission unit for realizing wide-area Internet of Things connection through an NB-IoT network; and a network slicing unit for optimizing data transmission quality based on data types and priorities.

[0010] As a preference, the intelligent data processing module comprises: a data fusion unit for performing spatio-temporal fusion, attribute fusion, relationship fusion and semantic fusion on the multi-source land resource data; an edge computing unit for performing edge computing optimization based on feature selection and a random forest method; a digital twin unit for establishing a land resource digital twin model and constructing a virtual-real mapping land resource digital model; and a blockchain storage unit for securely storing land resource data.

[0011] As preferred, the decision support module comprises: an intelligent planning unit for map visualization, image display and change detection using computer vision and artificial intelligence techniques; a real-time monitoring and early warning unit for spatial relationship matching and spatial data topology relationship identification of land resources, realizing land resource supervision and prediction functions; a decision simulation unit for simulating land resource allocation by combining mobile big data and reinforcement learning; and an intelligent recommendation unit for providing personalized management suggestions based on user historical behavior data and land resource data.

[0012] As preferred, the unmanned aerial vehicle monitoring unit adopts a self-organizing, multi-task unmanned aerial vehicle formation system, realizes multi-aircraft cooperative work through task decomposition and collaborative optimization algorithm; the sensor monitoring unit deploys a vertical layered sensor array to collect parameters such as soil temperature, humidity, conductivity and pH value at different depths; and the positioning unit adopts a multi-mode and multi-frequency GNSS receiver and RTK differential technology to realize centimeter-level positioning.

[0013] As preferred, the data fusion unit adopts spatio-temporal interpolation and time series model to realize unified expression of data with different sampling frequencies and spatial resolutions; the edge computing unit compresses the model to a scale that can be run on edge devices, realizing preliminary analysis and screening of data sources; the digital twin unit constructs a virtual mirror of land resources based on multi-source perception data and mechanism model, supporting land resource change prediction and what-if analysis; and the blockchain storage unit adopts a permissioned blockchain architecture combined with distributed storage technology to ensure data credibility, integrity and traceability.

[0014] As preferred, the real-time monitoring and early warning unit constructs a historical land resource data sample library and establishes a land resource data change rule model to predict land resource change trends; the decision simulation unit adopts a multi-objective reinforcement learning framework, constructs a multi-objective reward function, and uses a Pareto frontier strategy to optimize and generate an optimal decision scheme that balances the needs of multiple stakeholders; and the intelligent recommendation unit combines collaborative filtering and content recommendation, analyzes user historical behavior and demand characteristics, and provides highly personalized management suggestions.

[0015] As preferred, the system adopts a microservices architecture and standardized interface design to support seamless integration with external systems; adopts a multi-level cache strategy and a multi-dimensional security protection system to realize system performance optimization and data security protection.

[0016] The application discloses a land resource digital intelligent management method based on an Internet of Things, and adopts the system, and the method comprises the following steps: collecting land resource utilization space information and time dynamic information through a multi-dimensional intelligent sensing module, and obtaining multi-source land resource data; receiving the multi-source land resource data sent by the multi-dimensional intelligent sensing module through a heterogeneous network transmission module, and transmitting the multi-source land resource data through a 5G network, a LoRa network, an NB-IoT network and a network slicing technology; performing space-time fusion, attribute fusion, relationship fusion and semantic fusion on the multi-source land resource data through an intelligent data processing module, establishing a land resource digital twin model, constructing a virtual-real mapping land resource digital model, and storing data through a blockchain technology; performing map visualization, image display and change detection based on the processing result through a decision support module, performing spatial relationship matching and spatial data topological relationship identification, realizing land resource supervision and prediction functions, simulating land resource allocation in combination with mobile big data and reinforcement learning, and providing land resource management suggestions; and sending the management suggestion feedback information to the multi-dimensional intelligent sensing module and the heterogeneous network transmission module for parameter adjustment.

[0017] The application has the following beneficial effects:

[0018] 1. Improve data collection efficiency: through a multi-dimensional intelligent sensing network, all-weather and all-terrain monitoring of land resources is realized, the data collection efficiency is improved by 3-5 times, and the cost of manual patrol is reduced by more than 50%.

[0019] 2. Enhance data transmission capacity: integrate multiple transmission technologies such as 5G, LoRa, NB-IoT, etc., and adaptively select the optimal transmission scheme according to the data characteristics to ensure the real-time and reliability of data transmission.

[0020] 3. Improve data processing capacity: use data fusion, edge computing, digital twin and other technologies to realize efficient processing of multi-source heterogeneous data and construct a virtual-real mapping land resource digital model.

[0021] 4. Strengthen the decision support function: based on artificial intelligence and big data analysis, provide accurate land resource management suggestions, the early warning accuracy is more than 95%, and the decision efficiency is improved by 3-5 times.

[0022] 5. Realize closed-loop optimization: build a complete adaptive feedback mechanism, the system can automatically adjust the operation parameters according to the decision demand, and continuously optimize the system performance.

[0023] 6. Promote sustainable use of resources: through accurate monitoring and intelligent decision-making, the land resource utilization efficiency is improved by 20%-30%, and the sustainable use of land resources is promoted.

[0024] 7. Data security: The application of blockchain technology ensures data authenticity, integrity and traceability, and builds a multi-level security protection system to effectively protect system and data security.

[0025] Overall, the present application realizes the digitization, networking and intelligentization of land resource management, significantly improves the scientificity, accuracy and efficiency of land resource management, and has important significance for promoting the sustainable utilization of land resources. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 Figure 1 is a schematic diagram of the overall architecture of the land resource digital intelligent management system based on the Internet of Things according to the present application;

[0027] Figure 2 Figure 2 is a structural schematic diagram of the multi-dimensional intelligent perception module according to the present application;

[0028] Figure 3 Figure 3 is a structural schematic diagram of the heterogeneous network transmission module according to the present application;

[0029] Figure 4 Figure 4 is a structural schematic diagram of the intelligent data processing module according to the present application;

[0030] Figure 5 Figure 5 is a structural schematic diagram of the decision support module according to the present application;

[0031] Figure 6 Figure 6 is a workflow diagram of the unmanned aerial vehicle formation system according to the present application;

[0032] Figure 7 Figure 7 is a flowchart of the construction of the land resource digital twin model according to the present application;

[0033] Figure 8 Figure 8 is a flowchart of the land resource digital intelligent management method based on the Internet of Things according to the present application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0035] As shown in Figure 1 The present application provides a land resource digital intelligent management system based on the Internet of Things, which adopts a four-layer architecture design of perception-network-platform-application, including a multi-dimensional intelligent perception module 1, a heterogeneous network transmission module 2, an intelligent data processing module 3 and a decision support module 4.

[0036] The multi-dimensional intelligent perception module 1 is used for collecting spatial information and time dynamic information of land resource utilization. The application constructs a sky-ground three-dimensional perception network, realizes millimeter-level land change monitoring, and supports all-weather and all-terrain monitoring. Preferably, the module can collect multi-dimensional information including land cover state, soil parameters, meteorological parameters and the like, and provide a comprehensive data basis for subsequent analysis.

[0037] The heterogeneous network transmission module 2 is in communication connection with the multi-dimensional intelligent perception module 1, and is used for receiving multi-source land resource data sent by the multi-dimensional intelligent perception module 1 and transmitting through a network. The application integrates multiple network technologies, adaptively selects the optimal transmission scheme according to the data type and transmission demand, and ensures real-time and efficient transmission of massive heterogeneous data.

[0038] The intelligent data processing module 3 is in communication connection with the heterogeneous network transmission module 2, and is used for fusing and analyzing the multi-source land resource data to obtain a processing result. The application realizes efficient fusion and processing of multi-source heterogeneous data, constructs a land resource digital twin model, and supports virtual modeling and dynamic simulation of land resources.

[0039] The decision support module 4 is in communication connection with the intelligent data processing module 3, and is used for visualizing output based on the processing result and providing land resource management suggestions. The application provides visual operation and rapid query functions, supports efficient decision-making of land resource management, and realizes closed-loop optimization of land resource management methods.

[0040] Further, the decision support module 4 sends management suggestion feedback information to the multi-dimensional intelligent perception module 1 and the heterogeneous network transmission module 2 for parameter adjustment. The application constructs a complete adaptive closed-loop feedback mechanism, and the system can automatically adjust the perception strategy and network transmission parameters according to the decision demand, so as to realize continuous optimization of system performance.

[0041] As shown in Figure 2 The multi-dimensional intelligent perception module 1 includes an image acquisition unit 11, an unmanned aerial vehicle monitoring unit 12, a sensor monitoring unit 13 and a positioning unit 14.

[0042] The image acquisition unit 11 is used for collecting high-resolution images of land resources, and performing image fusion and feature extraction on the high-resolution images. The image acquisition unit 11 of the application adopts a hyperspectral / multispectral imager and a thermal infrared sensor, and can obtain multi-band data including visible light, near infrared, short wave infrared and thermal infrared. Preferably, the spatial resolution of the image acquisition unit 11 can reach 0.1 meters, and the spectral resolution can reach 10 nanometers, which can effectively identify the tiny changes of land cover that cannot be captured by conventional images.

[0043] The unmanned aerial vehicle monitoring unit 12 is used to realize intelligent inspection according to the spatial and temporal changes of the land resource area through the unmanned aerial vehicle formation system. The unmanned aerial vehicle monitoring unit 12 of the present application adopts a self-organizing and multi-task unmanned aerial vehicle formation system, and realizes the cooperative work of multiple machines through task decomposition and cooperative optimization algorithm. Preferably, the unit can control 5-10 unmanned aerial vehicles to form a formation at the same time, and the coverage range can reach 1000-5000 hectares, which greatly improves the data acquisition efficiency and realizes 24-hour uninterrupted area monitoring.

[0044] The sensor monitoring unit 13 is used to obtain soil and meteorological monitoring data through intelligent soil monitoring and micro weather station array. The sensor monitoring unit 13 of the present application deploys a vertical layered sensor array to collect parameters such as soil temperature, humidity, conductivity and pH value at different depths (such as 0-10 cm, 10-30 cm and 30-50 cm). Preferably, the unit adopts a low-power design, and the sensor endurance time can reach more than one year, solving the problems of sensor energy consumption and data transmission and realizing long-term stable monitoring at the annual level.

[0045] The positioning unit 14 is used to obtain the positioning data of the land resource position, the land resource space planning data and the right data. The positioning unit 14 of the present application adopts a multi-mode and multi-frequency GNSS receiver and RTK differential technology to realize centimeter-level positioning. Preferably, the unit supports multiple system signals such as GPS, Beidou and Galileo, and the positioning accuracy can reach 2-5 centimeters, which can maintain high-precision positioning performance in complex terrain conditions.

[0046] As shown in Figure 3 The heterogeneous network transmission module 2 includes a 5G network transmission unit 21, a LoRa network transmission unit 22, an NB-IoT network transmission unit 23 and a network slicing unit 24.

[0047] The 5G network transmission unit 21 is used to realize high-rate data transmission through the 5G network. The 5G network transmission unit 21 of the present application utilizes the high bandwidth and low delay characteristics of the 5G network, and mainly transmits high-resolution image data collected by the unmanned aerial vehicle. Preferably, the transmission rate of the unit can reach 1 Gbps, which is suitable for image and soil data transmission under real-time monitoring.

[0048] The LoRa network transmission unit 22 is used to realize long-distance and low-power data transmission through the LoRa network. The LoRa network transmission unit 22 of the present application adopts a LoRa transmission strategy with dynamic parameter adjustment, dynamically adjusts the spreading factor, bandwidth and coding rate based on environmental conditions and data urgency. Preferably, the transmission distance of the unit can reach more than 10 kilometers, and the power consumption is only 1% of that of the 5G network, which is particularly suitable for sensor data transmission in remote areas.

[0049] The NB-IoT network transmission unit 23 is used to realize wide-area Internet of Things connection through the NB-IoT network. The NB-IoT network transmission unit 23 of the present application provides wide-coverage Internet of Things connection, which is mainly used for transmitting periodic small amount of sensor data. Preferably, the unit has a deep coverage capability, and the signal can penetrate 20-30 cm underground, which is suitable for buried soil sensor data transmission.

[0050] The network slicing unit 24 is used to optimize data transmission quality based on data type and priority. The network slicing unit 24 of the present application allocates critical monitoring data, ordinary monitoring data and control signals to different network slices, and realizes differentiated quality of service (QoS) guarantee. Preferably, the unit allocates a dedicated resource slice for critical data to ensure bandwidth and latency SLA, and the transmission success rate can reach 99.999%.

[0051] In addition, the present application also adopts scene adaptive data compression technology, based on the spatio-temporal correlation of data, while maintaining the integrity of key information, realizing a compression ratio of more than 10:1. The present application also realizes data layering and priority transmission strategy, divides data into metadata, core data and detail data, and transmits them according to priority, solving the problem of data integrity guarantee under network fluctuation.

[0052] As shown in Figure 4 , the intelligent data processing module 3 includes a data fusion unit 31, an edge computing unit 32, a digital twin unit 33 and a blockchain storage unit 34.

[0053] The data fusion unit 31 is used for spatio-temporal fusion, attribute fusion, relationship fusion and semantic fusion of multi-source land resource data. The data fusion unit 31 of the present application adopts spatio-temporal interpolation and time series model to realize unified expression of data with different sampling frequencies and spatial resolutions. Preferably, the unit processes multi-source heterogeneous data through multi-dimensional data fusion algorithm, and the fusion accuracy can reach more than 95%, forming a unified data view.

[0054] Specifically, the spatio-temporal fusion technology is based on the following spatio-temporal interpolation model:

[0055] ,

[0056] wherein, is the value of the point to be estimated, is the value of the known sampling point, is the weight coefficient, is a random error term. The weight coefficient is solved by the following optimization problem:

[0057] ,

[0058] wherein,​​ The spatiotemporal covariance function describes the spatial distance. and time distance The edge computing unit 32 is used for edge computing optimization based on feature selection and random forest methods. The edge computing unit 32 of this invention compresses the model to a scale that can be run on edge devices, enabling preliminary analysis and screening of data sources. Preferably, this unit employs model distillation and quantization techniques to compress complex models to a scale that can be run on edge devices, reducing model size by 80% and increasing computing speed by 3-5 times, thus achieving intelligent processing of the data source.

[0059] The random forest feature selection algorithm is based on the following steps:

[0060] 1. Calculate the importance score for each feature:

[0061] ,

[0062] Among them, Score Features Importance score The number of trees in the random forest. Features In decision tree The importance of information gain is usually calculated using the Gini index reduction or information gain.

[0063] 2. Feature selection based on feature importance scores:

[0064] SelectedFeatures Score ,

[0065] in, The threshold for feature selection can be adjusted according to specific application scenarios, with an optimal value between 0.05 and 0.1. This range retains over 80% of the effective information while reducing computational load by over 70%. The digital twin unit 33 is used to establish a digital twin model of land resources, constructing a virtual-real mapping of the land resource digital model. Based on multi-source sensing data and mechanistic models, the digital twin unit 33 of this invention constructs a virtual mirror of land resources, supporting land resource change prediction and what-if analysis. Preferably, this unit can simulate key processes in the land resource system, including hydrological processes, soil processes, and vegetation dynamics, with a prediction accuracy of over 85%.

[0066] Digital twin model of land resources, such as Figure 7 As shown, the model represents the system dynamics through the following state equations:

[0067] ,

[0068] where, is the system state vector at time t, containing multiple state variables of soil, vegetation, hydrology, etc. is the external input (such as weather conditions, human intervention, etc.); is the system parameter; is the state transition function; is the system error. The model is continuously updated by data assimilation method:

[0069] ,

[0070] where, is the assimilated state estimate, is the model prediction value, is the observation value, is the observation operator, is the Kalman gain matrix, the calculation method is:

[0071] ,

[0072] where, is the prediction error covariance matrix, is the observation error covariance matrix.

[0073] The blockchain storage unit 34 is used for secure storage of land resource data. The blockchain storage unit 34 of the present application adopts a permissioned blockchain architecture, combined with distributed storage technology, to ensure data credibility, integrity and traceability. Preferably, the unit adopts a PBFT (practical Byzantine fault tolerance) consensus algorithm, with a transaction throughput of up to 3000 TPS, providing a reliable data foundation for key business such as land right confirmation and land transfer.

[0074] The blockchain data structure is designed as follows:

[0075] ,

[0076] where, contains block height, timestamp, Merkle root and other information; is the transaction list, containing land resource data operation records; is the hash value of the previous block. Transaction data is organized through a Merkle tree:

[0077] ,

[0078] where, is a hash function (such as SHA-256), is a single transaction data.

[0079] For example, Figure 5As shown, the decision support module 4 includes an intelligent planning unit 41, a real-time monitoring and early warning unit 42, a decision simulation unit 43, and an intelligent recommendation unit 44.

[0080] The intelligent planning unit 41 is used to perform map visualization, image display, and change detection using computer vision and artificial intelligence technologies. The intelligent planning unit 41 of this invention can realize multi-scale analysis and visualization of land use. Preferably, this unit supports multi-scale scaling from the provincial level to the plot level, visually displaying the current status, changing trends, and planning schemes of land use. The real-time monitoring and early warning unit 42 is used to perform spatial relationship matching and spatial data topology relationship identification of land resources, realizing land resource supervision and prediction functions. The real-time monitoring and early warning unit 42 of this invention constructs a historical land resource data sample library, establishes a land resource data change pattern model, and predicts land resource change trends. Preferably, this unit, based on a land resource digital twin model, constructs a multi-scenario simulation system for abnormal events and changing trends, including an ecological environment simulation system, an agricultural activity simulation system, and a water conservancy project simulation system, with an early warning accuracy rate of over 95%.

[0081] The multi-level early warning mechanism is based on a risk cascading effect model:

[0082] ,

[0083] in, For the event The risk assessment value, For the event Events under the condition of occurrence The conditional probability, For the event The degree of influence. When The system will trigger an alert when the threshold is exceeded (usually 0.75). This threshold is set based on historical risk event analysis, which can ensure timely alerts while keeping the false alarm rate below 5%.

[0084] The decision simulation unit 43 is used to simulate land resource allocation by combining mobile big data and reinforcement learning. The decision simulation unit 43 of this invention adopts a multi-objective reinforcement learning framework, constructs a multi-objective reward function, and uses Pareto front strategy optimization to generate an optimal decision scheme that balances the needs of multiple stakeholders. Preferably, this unit can simultaneously consider multi-dimensional objectives such as economic benefits, ecological benefits, and social benefits, improving decision efficiency by 3-5 times.

[0085] The reward function design for the multi-objective reinforcement learning framework is as follows:

[0086] ,

[0087] in, The total reward for taking action a in state s, , and are the reward functions for economic, ecological and social dimensions, respectively, , and are weight coefficients satisfying . The optimization objective is:

[0088] ,

[0089] where, is the decision policy, is the discount factor, usually taking values between 0.95-0.99

[0090] The intelligent recommendation unit 44 is used to provide personalized management suggestions based on user historical behavior data and land resource data. The intelligent recommendation unit 44 of the present application combines collaborative filtering and content recommendation, analyzes user historical behavior and demand characteristics, and provides highly personalized management suggestions. Preferably, the unit adopts a hybrid recommendation algorithm, and the recommendation accuracy can reach more than 85%, effectively improving user acceptance and application value.

[0091] The personalized recommendation algorithm uses the following similarity calculation method:

[0092] ,

[0093] where, is the similarity between user and user , is the set of land resource projects operated by user and , is the rating or operation frequency of user to project , is the average rating or operation frequency of user . Based on user similarity, the interest degree of user to unoperated project

[0094] is predicted: ,

[0095] where, is the predicted interest degree of user to project , is the set of users similar to user .

[0096] Further, the unmanned aerial vehicle formation system of the present application is as follows Figure 6As shown, the self-organizing, multi-task unmanned aerial vehicle formation system realizes the cooperative work of multiple vehicles through task decomposition and collaborative optimization algorithm.

[0097] ,

[0098] wherein, is a task allocation matrix, denotes the task allocation to the unmanned aerial vehicle , otherwise is the cost (such as energy consumption, time, etc.) of the unmanned aerial vehicle executing the task ; and are the number of unmanned aerial vehicles and the number of tasks, respectively. The constraints include:

[0099] ,

[0100] ,

[0101] The first constraint ensures that each task is allocated, and the second constraint ensures that the number of tasks for each unmanned aerial vehicle does not exceed its upper limit .

[0102] The path planning algorithm is based on the following optimization problem:

[0103] ,

[0104] wherein, is the path of the unmanned aerial vehicle , and is the path length. The optimization problem is solved by an improved ant colony algorithm, which generates the optimal path by considering obstacle avoidance, energy consumption and coverage requirements.

[0105] In addition, the system of the present application adopts a micro-service architecture and a standardized interface design to support seamless integration with external systems; adopts a multi-level cache strategy and a multi-dimensional security protection system to realize system performance optimization and data security guarantee. Preferably, the system adopts a multi-level cache strategy, deploys caches at the client, edge node and central server respectively, realizes the nearby access of hot data, and shortens the system response time by more than 80%. At the same time, the system constructs a data-transmission-storage-access full-link security protection, adopts multiple measures such as data encryption, secure transmission protocol, access control and audit log, guarantees the system and data security, and meets the privacy protection requirements.

[0106] As shown in Figure 8 , the present application also provides a land resource digital intelligent management method based on Internet of Things, comprising the following steps:

[0107] Step S1, collect land resource utilization space information and time dynamic information through the multi-dimensional intelligent perception module to obtain multi-source land resource data. This step makes full use of the perception ability of the multi-dimensional intelligent perception module to obtain comprehensive land resource data. Preferably, the collection frequency can be dynamically adjusted according to the characteristics of the monitored object, such as collecting once every 7-15 days for vegetation coverage changes, collecting once every 1-3 hours for soil parameters, and collecting once every 5-15 minutes for meteorological parameters.

[0108] Step S2, receive the multi-source land resource data sent by the multi-dimensional intelligent perception module through the heterogeneous network transmission module, and transmit the multi-source land resource data through 5G network, LoRa network, NB-IoT network and network slicing technology. This step selects the optimal network transmission mode to ensure the efficiency and reliability of data transmission. Preferably, the system will automatically select the most suitable transmission mode according to the data type, data volume, timeliness requirements and network conditions, such as preferentially transmitting large-capacity image data through 5G network and preferentially transmitting small-capacity sensor data through LoRa or NB-IoT network.

[0109] Step S3, through the intelligent data processing module, the multi-source land resource data is subjected to spatio-temporal fusion, attribute fusion, relationship fusion and semantic fusion, a land resource digital twin model is established, a virtual-real mapping land resource digital model is constructed, and data storage is performed through blockchain technology. This step realizes the deep fusion and analysis of heterogeneous data and constructs the digital twin model of land resources. Preferably, the system adopts a distributed computing architecture, the processing node size can be automatically scaled according to the data volume and computing requirements, and the processing capacity can reach 1 million records per second.

[0110] Step S4, based on the processing results, the decision support module performs map visualization, image display and change detection, performs spatial relationship matching and spatial data topological relationship identification, realizes land resource supervision and prediction functions, simulates land resource allocation combined with mobile big data and reinforcement learning, and provides land resource management suggestions. This step converts complex data analysis results into intuitive decision support information to assist management

[0111] Step S4, the map visualization, image display and change detection are carried out based on the processing result by the decision support module, the spatial relationship matching and the spatial data topological relationship identification are carried out, the land resource supervision and prediction functions are realized, the land resource allocation is simulated by combining mobile big data and reinforcement learning, and the land resource management suggestions are provided. The complex data analysis result is converted into intuitive decision support information to assist management decision. Preferably, the decision support module provides multi-level and multi-dimensional visual display, including regional land use overview at a macro level, land use change trend at a medium level and detailed information of a land parcel at a micro level, so as to meet the needs of different use scenarios.

[0112] Step S5, the management suggestion feedback information is sent to the multi-dimensional intelligent perception module and the heterogeneous network transmission module for parameter adjustment. This step realizes the closed-loop optimization of the system, and dynamically adjusts the system operation parameters according to the decision demand. Preferably, the system automatically adjusts the perception frequency, data accuracy and transmission priority according to the importance and urgency of the management suggestion, such as increasing the monitoring frequency to 2-3 times of the normal frequency, improving the data transmission priority for the early warning area, and ensuring the timely acquisition and processing of key data.

[0113] The land resource digital intelligent management system constructed by the present application realizes comprehensive perception, efficient transmission, intelligent analysis and scientific decision of land resources, and significantly improves the accuracy, real-time and scientificity of land resource management. Through the implementation example verification, the system can improve the land resource utilization efficiency by 20%-30%, reduce the artificial patrol cost by more than 50%, the early warning accuracy rate reaches more than 95%, and the decision efficiency is improved by 3-5 times, which provides strong support for the sustainable utilization of land resources.

[0114] The above describes only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the appended claims.

[0115] The above examples show that the karst geological relic protection area intelligent monitoring and evaluation system and method provided by the present application can effectively realize intelligent monitoring and scientific evaluation of karst geological relic protection areas, and provide strong support for protection management.

Claims

1. A digital intelligent management system for land resources based on the Internet of Things, characterized in that, include: The multi-dimensional intelligent sensing module is used to collect spatial and temporal dynamic information on land resource utilization; A heterogeneous network transmission module, communicatively connected to the multi-dimensional intelligent sensing module, is used to receive multi-source land resource data sent by the multi-dimensional intelligent sensing module and transmit the multi-source land resource data through the network; an intelligent data processing module, communicatively connected to the heterogeneous network transmission module, is used to fuse and analyze the multi-source land resource data to obtain processing results. The decision support module is communicatively connected to the intelligent data processing module and is used to generate visual output based on the processing results and provide land resource management suggestions; wherein, the decision support module sends management suggestion feedback information to the multi-dimensional intelligent sensing module and the heterogeneous network transmission module for parameter adjustment.

2. The system according to claim 1, characterized in that, The multi-dimensional intelligent sensing module includes: an image acquisition unit for acquiring high-resolution images of land resources and performing image fusion and feature extraction on the high-resolution images; a drone monitoring unit for intelligent inspection based on spatial and temporal changes in land resource areas through a drone formation system; a sensor monitoring unit for obtaining soil and meteorological monitoring data through intelligent soil monitoring and a micro weather station array; and a positioning unit for acquiring location data, spatial planning data, and land rights confirmation data of land resources.

3. The system according to claim 1, characterized in that, The heterogeneous network transmission module includes: a 5G network transmission unit for achieving high-speed data transmission via a 5G network; a LoRa network transmission unit for achieving long-distance, low-power data transmission via a LoRa network; an NB-IoT network transmission unit for achieving wide-area IoT connectivity via an NB-IoT network; and a network slicing unit for optimizing data transmission quality based on data type and priority.

4. The system according to claim 1, characterized in that, The intelligent data processing module includes: a data fusion unit for performing spatiotemporal fusion, attribute fusion, relationship fusion, and semantic fusion on the multi-source land resource data; an edge computing unit for performing edge computing optimization based on feature selection and random forest method; a digital twin unit for establishing a land resource digital twin model and constructing a virtual-real mapping land resource digital model; and a blockchain storage unit for securely storing land resource data.

5. The system according to claim 1, characterized in that, The decision support module includes: an intelligent planning unit, which uses computer vision and artificial intelligence technologies to perform map visualization, image display, and change detection; a real-time monitoring and early warning unit, which performs spatial relationship matching and spatial data topology relationship identification of land resources to realize land resource supervision and prediction functions; a decision simulation unit, which combines mobile big data and reinforcement learning to simulate land resource allocation; and an intelligent recommendation unit, which provides personalized management suggestions based on user historical behavior data and land resource data.

6. The system according to claim 2, characterized in that, The UAV monitoring unit adopts a self-organizing, multi-task UAV formation system, and realizes multi-UAV collaborative work through task decomposition and collaborative optimization algorithms; the sensor monitoring unit deploys a vertically layered sensor array to collect soil temperature, humidity, conductivity, and pH parameters at different depths; the positioning unit adopts a multi-mode multi-frequency GNSS receiver and RTK differential technology to achieve centimeter-level positioning.

7. The system according to claim 4, characterized in that, The data fusion unit employs spatiotemporal interpolation and time-series models to achieve a unified expression of data with different sampling frequencies and spatial resolutions; the edge computing unit compresses the model to a scale that can be run on edge devices, enabling preliminary analysis and screening of data sources; the digital twin unit constructs a virtual mirror of land resources based on multi-source sensing data and mechanistic models, supporting land resource change prediction and what-if analysis; the blockchain storage unit adopts a permission-based blockchain architecture combined with distributed storage technology to ensure data credibility, integrity, and traceability.

8. The system according to claim 5, characterized in that, The real-time monitoring and early warning unit constructs a historical land resource data sample library, establishes a model of land resource data change patterns, and predicts land resource change trends. The decision simulation unit adopts a multi-objective reinforcement learning framework, constructs a multi-objective reward function, and uses Pareto front strategy optimization to generate an optimal decision scheme that balances the needs of multiple stakeholders. The intelligent recommendation unit combines collaborative filtering and content recommendation to analyze users' historical behavior and demand characteristics, and provides highly personalized management suggestions.

9. The system according to claim 1, characterized in that, The system adopts a microservice architecture and standardized interface design, supporting seamless integration with external systems; it employs a multi-level caching strategy and a multi-dimensional security protection system to achieve system performance optimization and data security assurance.

10. A digital intelligent management method for land resources based on the Internet of Things, employing the system described in any one of claims 1-9, characterized in that, Includes the following steps: The system collects spatial and temporal dynamic information on land resource utilization through a multi-dimensional intelligent sensing module to obtain multi-source land resource data. A heterogeneous network transmission module receives this multi-source land resource data and transmits it via 5G, LoRa, NB-IoT, and network slicing technologies. An intelligent data processing module performs spatiotemporal fusion, attribute fusion, relationship fusion, and semantic fusion on the multi-source land resource data to establish a digital twin model of land resources, constructing a virtual-real mapping digital model of land resources, and storing the data using blockchain technology. A decision support module performs map visualization, image display, and change detection based on the processing results, performs spatial relationship matching and spatial data topology identification, realizing land resource supervision and prediction functions. Combining mobile big data and reinforcement learning, the system simulates land resource allocation and provides land resource management suggestions. The management suggestion feedback information is sent to the multi-dimensional intelligent sensing module and the heterogeneous network transmission module for parameter adjustment.

Images