Deployment Method and System for Geological Hazard Monitoring Network Based on Coverage Entropy Theory

Abstract

This invention relates to the field of geological disaster monitoring technology, specifically a method and system for deploying a geological disaster monitoring network based on coverage entropy theory. The method divides the monitoring area into several sub-regions and calculates the coverage entropy value of each sub-region. Sensor nodes are modeled as agents, and collaborative game optimization is performed based on the coverage entropy value and a constructed objective function. The positions and node states of the agents are updated to generate an optimized solution. The objective function consists of the coverage contribution value of each agent with a set coverage weight and an overlap penalty term. The coverage weight is dynamically adjusted by the geological disaster early warning level, which is obtained through the fusion calculation of real-time data from multiple sources in the monitoring area and the agents. This solution enables dynamic optimization and deployment of the geological disaster monitoring network, covering high-risk areas and improving robustness to cope with sudden geological disasters.

Description

Technical Field

[0001] This invention relates to the field of geological disaster monitoring technology, specifically to a method and system for deploying a geological disaster monitoring network based on the theory of coverage entropy. Background Technology

[0002] Geological disaster monitoring aims to predict and prevent natural disasters through scientific means in order to reduce loss of life and property. Common geological disaster monitoring methods include building a geological disaster network, monitoring relevant information on geological disasters, and taking preventive and early warning measures in advance.

[0003] Currently, the deployment of geological disaster monitoring networks mainly relies on static algorithms, such as uniform distribution, genetic algorithms, or empirical point placement, to arrange sensor nodes in the network. While this can achieve multi-objective optimization, it still has the following problems:

[0004] Static algorithms for layout have limitations, including low convergence efficiency, reliance on fixed weights, and difficulty in adapting to dynamic changes in geological disaster scenarios.

[0005] The monitoring and optimization process involves randomness, which may result in insufficient coverage of high-risk areas;

[0006] 3. Geological disasters often occur suddenly. Static algorithms have low tolerance for node failures and poor robustness, making them difficult to cope with sudden geological disasters.

[0007] Therefore, there is an urgent need for a geological disaster monitoring network deployment method and system based on the coverage entropy theory, which can dynamically optimize the deployment of the geological disaster monitoring network, cover high-risk areas, and improve robustness to cope with sudden geological disasters. Summary of the Invention

[0008] One of the objectives of this invention is to provide a geological disaster monitoring network deployment method based on the coverage entropy theory, which can dynamically optimize the deployment of the geological disaster monitoring network, cover high-risk areas, improve robustness, and cope with sudden geological disasters.

[0009] The first basic solution provided by this invention is a method for deploying a geological disaster monitoring network based on the theory of cover entropy, comprising:

[0010] The monitoring area is divided into several sub-regions, and the coverage entropy value of each sub-region is calculated.

[0011] Sensor nodes are modeled as agents. Based on the coverage entropy value and the constructed objective function, collaborative game optimization is performed to update the position and node state of the agents and generate an optimization scheme. The objective function consists of the coverage contribution value of the agents with coverage weights and an overlap penalty term. The coverage weights are dynamically adjusted by the geological disaster early warning level, which is obtained by fusing real-time data from multiple sources in the monitoring area.

[0012] Furthermore, the calculation of the coverage entropy value of each sub-region includes: calculating the coverage entropy value of each sub-region using a coverage entropy model;

[0013] The coverage entropy model is as follows:

[0014] ;

[0015] in Geological risk weights; For the first The probability that a sub-region is not monitored; coverage entropy value The higher the number, the higher the grade. The greater the risk of blind spots in the monitoring of individual sub-regions;

[0016] in :

[0017] ;

[0018] in The effective radius of the sensor, sub-region To sensor node The distance.

[0019] Furthermore, the cooperative game optimization includes: covering game;

[0020] Coverage game theory involves each agent moving independently and maximizing its own coverage contribution based on the coverage entropy value of each sub-region. The total contribution of all agents does not decrease, and the covering game ends when the displacements of all agents converge during their movements; where For intelligent agents The grid coverage entropy is equal to that of the agent. Coverage entropy value of the sub-region; This represents the signal strength.

[0021] Furthermore, the objective function is:

[0022] ;

[0023] in Total contribution value;

[0024] For intelligent agents The grid cover entropy, ,in For intelligent agents Sub-region;

[0025] Signal strength;

[0026] For intelligent agents The distance to the sub-region with the highest coverage entropy value along the movement route;

[0027] The effective radius of the intelligent agent;

[0028] This is the overlap penalty coefficient;

[0029] This is an overlap indicator function;

[0030] ,in The overlap threshold, For overlap, ,in For intelligent agents Effective monitoring area on a two-dimensional plane.

[0031] Furthermore, during the movement, the agent adjusts its movement based on the position update strategy;

[0032] The location update strategy is as follows:

[0033]

[0034] in For intelligent agents The direction of movement and step size;

[0035] The learning rate;

[0036] These are the entropy gradient coefficients;

[0037] This represents the maximum coverage entropy value.

[0038] Furthermore, during the movement, the agent adjusts the signal strength according to the signal strength adjustment strategy;

[0039] The signal strength adjustment strategy is as follows:

[0040]

[0041] in For intelligent agents The next moment The signal strength;

[0042] For intelligent agents At present The signal strength;

[0043] To adjust the step size;

[0044] The sign function determines the direction of adjustment:

[0045]

[0046] The entropy threshold is set based on the geological risk type of the monitored area.

[0047] Furthermore, the cooperative game optimization also includes: cooperative game;

[0048] Cooperative game theory includes: agents forming alliances, evaluating each agent's relative contribution to the alliance, and controlling the node states of agents based on the magnitude of their relative contributions.

[0049] Furthermore, the node state of the control agent includes:

[0050] Intelligent agents establish a communication topology and broadcast node state data to form an alliance;

[0051] Calculate the coverage contribution value for each agent. ;

[0052] Calculate the total contribution value based on the objective function. ;

[0053] Compare the coverage contribution value and the total contribution value of the agents to obtain the relative contribution. ;

[0054] Obtain the number of agents in the alliance, and calculate the redundancy threshold based on the number of agents. ;in This is a redundancy threshold. The number of agents in the alliance;

[0055] Determine if the relative contribution of the agent is less than the threshold. If yes, keep the agent; otherwise, put the agent into a dormant state.

[0056] Update the node state data of the agents, determine whether all agents in the alliance have performed relative contribution assessments, and if so, end the optimization and generate an optimization scheme.

[0057] Furthermore, the coverage weight is dynamically adjusted based on the geological disaster early warning level. The geological disaster early warning level is obtained through the fusion calculation of multi-source real-time data from the intelligent agent and the monitoring area, including:

[0058] Obtain the displacement rate of the intelligent agent and the rainfall in the monitored area;

[0059] The geological disaster early warning level is calculated based on the agent's displacement rate and the rainfall in the monitored area. :

[0060]

[0061] in and These are the displacement rate of the agent and the normalized value of rainfall in the monitored area, respectively.

[0062] According to the geological disaster early warning level Adjust coverage weight ;

[0063] Update the covering weights in the objective function;

[0064] It also includes: monitoring whether the change in the geological disaster early warning level exceeds the preset threshold. If so, the collaborative game optimization is carried out again; if not, the change in the geological disaster early warning level continues to be monitored.

[0065] The second objective of this invention is to provide a geological disaster monitoring network deployment system based on the coverage entropy theory, which can dynamically optimize the deployment of the geological disaster monitoring network, cover high-risk areas, and improve robustness to cope with sudden geological disasters.

[0066] The present invention provides a second basic scheme: a geological disaster monitoring network deployment system based on the coverage entropy theory, which adopts the above-mentioned geological disaster monitoring network deployment method based on the coverage entropy theory.

[0067] Beneficial effects: First, this scheme divides the monitoring area into several sub-regions and calculates the coverage entropy value of each sub-region. The magnitude of the coverage entropy value can characterize the risk of the sub-region being a monitoring blind spot, so that it can be given higher priority in subsequent monitoring layout. This method is the first to apply the information entropy theory to the field of geological disaster monitoring network deployment, and establishes a brand-new risk assessment system for monitoring blind spots, so as to more accurately identify risks. By quantifying the risk of monitoring blind spots through mathematical modeling, it overcomes the subjectivity of traditional experience-based point placement and significantly improves the accuracy of high-risk area identification.

[0068] Secondly, in this scheme, each sensor node in the monitoring network is modeled as an intelligent agent, which then performs cooperative game optimization based on the coverage entropy value and the constructed objective function, updates the position and node state of the intelligent agent, and generates an optimization scheme.

[0069] The objective function guides the optimization direction in the collaborative game optimization process. It consists of the coverage contribution value and overlap penalty term of each agent with a coverage weight. By adjusting the coverage weight, the proportion of the coverage contribution value and overlap penalty term in the objective function can be changed. The coverage weight is dynamically adjusted by the geological disaster early warning level. The geological disaster early warning level is obtained by fusing real-time data from multiple sources in the agent and the monitoring area, thereby establishing a dynamic mapping relationship between geological disaster risk and optimization objective, and realizing adaptive adjustment of coverage weight. For monitoring areas with different types of geological disasters, the optimization objective is automatically adjusted according to real-time monitoring data. Compared with the traditional fixed weight algorithm, which suffers a sharp decline in performance when the environment changes abruptly, this scheme can be steplessly adjusted between different objective requirements to meet the needs of different scenarios. Moreover, multi-source data fusion decision-making, which integrates multi-dimensional data to generate optimization strategies, has higher accuracy than single-index decision-making. In addition, there is cooperation among the agents in the collaborative game optimization. If any agent fails, the optimization process will compensate for the monitoring gap caused by its failure. The sensor node failure tolerance is high, which improves robustness and can cope with sudden geological disasters.

[0070] Finally, this solution combines information entropy theory and agent game theory, which is more optimized than single agent game theory, improves convergence efficiency during the game process, and adapts to the dynamic changes in geological disaster scenarios.

[0071] In summary, this solution enables dynamic optimization and deployment of the geological disaster monitoring network, covering high-risk areas and enhancing robustness to cope with sudden geological disasters. Attached Figure Description

[0072] Figure 1 This is a flowchart illustrating an embodiment of the geological disaster monitoring network deployment method based on the coverage entropy theory of the present invention.

[0073] Figure 2 This is a schematic diagram of the coverage game process in an embodiment of the geological disaster monitoring network deployment method based on coverage entropy theory of the present invention;

[0074] Figure 3 This is a schematic diagram of the collaborative game process in an embodiment of the geological disaster monitoring network deployment method based on the coverage entropy theory of the present invention. Detailed Implementation

[0075] The following detailed description illustrates the specific implementation method:

[0076] Example 1

[0077] This embodiment is basically as shown in the appendix. Figure 1 As shown: A method for deploying a geological hazard monitoring network based on coverage entropy theory, including:

[0078] The monitoring area is divided into sub-areas, and the coverage entropy value of each sub-area is calculated using the coverage entropy model. The coverage entropy value is used to represent the monitoring priority, ensuring that high-risk areas are covered first.

[0079] Specifically, the monitoring area is divided into several sub-areas, each sub-area is a grid, and each grid is an independent unit;

[0080] The coverage entropy model is used to calculate the coverage entropy value of each sub-region; the coverage entropy model is as follows:

[0081] ;

[0082] in Geological risk weights are dynamically adjusted based on slope and lithology. For the first The probability that a sub-region is not monitored; coverage entropy value The higher the number, the higher the grade. The greater the risk of blind spots in the monitoring of individual sub-regions;

[0083] in :

[0084] ;

[0085] in The effective radius of the sensor, sub-region To sensor node distance, This refers to the number of sensors; in this embodiment... .

[0086] Sensor nodes are modeled as intelligent agents, and collaborative game optimization is performed to update the position and node state of the agents and generate optimization schemes. The agents have the ability to make movement decisions, adjust node states, and adjust signal strength. Movement decisions adjust their own positions based on coverage entropy; node state (active / dormant) adjustment selectively dormants redundant nodes when resources are limited; signal strength adjustment optimizes the monitoring range.

[0087] Cooperative game optimization includes: overlay game and cooperative game;

[0088] The overlay game is the competitive phase of the agents, such as... Figure 2As shown, this includes: each agent moves independently and maximizes its own coverage contribution value through movement based on the coverage entropy value of each sub-region. The total contribution value of all agents does not decrease, and the covering game ends when the displacements of all agents converge.

[0089] in For intelligent agents (i.e., sensor nodes) The grid cover entropy, ,in For intelligent agents Sub-region; Signal strength;

[0090] The total contribution value is calculated using the constructed objective function;

[0091] The objective function consists of the coverage contribution values ​​of agents with separate coverage weights and an overlap penalty term, with the sum of the two coverage weights being one:

[0092] ;

[0093] in Total contribution value;

[0094] For grid cover entropy, ,in For intelligent agents Sub-region;

[0095] Signal strength (0.3~1.0), dynamically adjusted according to power: Normalized value;

[0096] For intelligent agents Distance to the high entropy region Unit: m;

[0097] The effective radius of the intelligent agent is determined by hardware performance; in this embodiment, it is 200m. (Unit: m)

[0098] The overlap penalty coefficient controls the penalty intensity of redundant monitoring. It is calibrated through gradient testing and has a value range of (0,1]. In this embodiment, it is set to 0.8.

[0099] This is an overlap indicator function with a value range of {0,1}. In this embodiment, it is activated when the overlap is greater than 30%.

[0100] ,in The overlap threshold is set to 0.3 in this embodiment. For overlap, ,in For intelligent agents The effective monitoring area on a two-dimensional plane is usually modeled as a circular area.

[0101] During movement, the agent adjusts its movement and signal strength according to a position update strategy and a signal strength adjustment strategy.

[0102] The location update strategy is as follows:

[0103]

[0104] in For intelligent agents The direction of movement and step size;

[0105] The learning rate controls the convergence speed; in this embodiment, it is set to 2.

[0106] This is the entropy gradient coefficient, which is set to 1 in this embodiment;

[0107] This represents the maximum coverage entropy value.

[0108] The signal strength adjustment strategy is as follows:

[0109]

[0110] in For intelligent agents The next moment The signal strength, normalized value [0.3, 1.0];

[0111] For intelligent agents At present The signal strength is initially set to 0.7 in this embodiment, and the normalized value is [0.3, 1.0].

[0112] To adjust the step size, it is set to 0.05 in this embodiment;

[0113] The sign function determines the direction of adjustment:

[0114]

[0115] Entropy threshold;

[0116] The settings are determined based on the geological risk type of the monitoring area; for example: debris flow 0.5, landslide 0.4, collapse 0.3.

[0117] The covering game ends when the displacements of all agents converge, including:

[0118] If the intelligent agent Is the displacement difference between adjacent time points less than the convergence threshold? If the displacements of all agents converge, the game ends; where... For intelligent agents At any moment displacement, For intelligent agents At any moment displacement, In this embodiment, the convergence threshold is set to 1m.

[0119] Cooperative game theory is the alliance phase of agents, which includes: agents forming an alliance, evaluating the relative contribution of each agent in the alliance, and controlling the node state of the agents based on the size of the relative contribution, that is, controlling the working state of the agents.

[0120] The specific process is as follows: Figure 3 As shown:

[0121] Intelligent agents establish a communication topology and broadcast node state data to form an alliance;

[0122] Calculate the coverage contribution value for each agent. ;

[0123] Calculate the total contribution value based on the constructed objective function. ;

[0124] Compare the coverage contribution value and the total contribution value of the agents to obtain the relative contribution. ;

[0125] Obtain the number of agents in the alliance, and calculate the redundancy threshold based on the number of agents. ;in This is a redundancy threshold, and the setting should be adjusted according to the actual situation. In this embodiment, it is set to 1.2. This represents the number of agents in the alliance, i.e., the number of sensors.

[0126] The system determines whether the agent's relative contribution is less than a threshold. If yes, the agent is retained; otherwise, the agent is put into a dormant state. Specifically, if the agent's relative contribution is less than the threshold, the agent is marked. That is, mark it as reserved, otherwise mark the agent. That is, marked as dormant; for the marked The intelligent agent sends a sleep command to switch it to a low-power mode; a tag... The intelligent agent continues to work;

[0127] Update the node state data of the agents, and determine whether all agents in the alliance have performed relative contribution assessment. If so, end the optimization and generate an optimization scheme. If not, continue to perform relative contribution assessment for the agents that have not performed relative contribution assessment.

[0128] The coverage weight in the objective function is dynamically adjusted based on the geological disaster early warning level. The geological disaster early warning level is obtained by fusing real-time data from multiple sources in the monitoring area and the agent. The higher the geological disaster early warning level, the greater the coverage weight.

[0129] Specifically, the displacement rate of the agent and the rainfall in the monitored area are obtained;

[0130] The geological disaster early warning level is calculated based on the agent's displacement rate and the rainfall in the monitored area. :

[0131]

[0132] in and These are the normalized values ​​of the agent's displacement rate and the rainfall in the monitored area, respectively, ranging from 0 to 1;

[0133] According to the geological disaster early warning level Adjust coverage weight ;

[0134] Update the covering weights in the objective function;

[0135] monitor If the change exceeds a preset threshold, then the collaborative game optimization is performed again; otherwise, monitoring continues. The change in value; in this embodiment, the preset threshold is 0.1.

[0136] The sensor nodes are optimized according to the optimization scheme.

[0137] The specific implementation process is as follows, taking the deployment of a monitoring network in a debris flow-prone area as an example:

[0138] The monitoring area is divided into sub-regions, and the coverage entropy value of each sub-region is calculated using the coverage entropy model, including:

[0139] During the geographic data acquisition phase, equipment was used to collect data from the monitoring area, including: LiDAR scanning: using a RIEGLVZ-4000 laser scanner with a point spacing of 0.5m; and drone aerial photography: using a DJI M300 RTK to acquire orthophotos with a resolution of 5cm.

[0140] In the geographic data preprocessing stage, data processing is performed. The data processing flow involves processing the collected data from the monitoring area. In this embodiment, the raw electromagnetic cloud data is processed, including point cloud filtering, ground point extraction, DEM generation, and grid division. Grid division employs grid-based modeling and uses ArcGIS Pro for spatial analysis: a 50×50 Fishnet grid (20m×20m) is created, and an attribute field, elevation standard deviation, is added to each grid (sub-region). ,slope lithology coefficient ;

[0141] During the coverage entropy modeling and blind spot risk assessment phase, the probability of a sub-region not being monitored is:

[0142]

[0143] in =200m;

[0144] Based on the coverage entropy model, calculate the coverage entropy value for each sub-region:

[0145]

[0146] in The geological risk weight ranges from 1.0 to 1.5.

[0147] The sensor nodes are modeled as intelligent agents, and cooperative game optimization is performed to update the positions and node states of the agents, including:

[0148] During the agent initialization phase, hardware parameters are set, including a quad-core processor to support a main control module for real-time decision-making, a sensor with a displacement accuracy of ±1mm, a communication module using the LoRaWAN protocol and a 500m radius, and a 12V / 10Ah lithium battery pack for power supply; the software logic is set to sense data, coverage contribution calculation, position update decision, and communication coordination.

[0149] In the coverage game phase, each agent moves independently and maximizes its own coverage contribution based on the coverage entropy value of each sub-region. The total contribution of all agents does not decrease, and the covering game ends when the displacements of all agents converge during their movements; where Initially set to 0.7, adjusted by step size. =0.05 adjustment; η=2 in the position update strategy to ensure migration to high-entropy regions, such as the direction of maximum entropy gradient.

[0150] In the cooperative game phase, agents form alliances, with each group consisting of 5-10 agents (sensor nodes). Redundancy is eliminated, and redundant nodes are put into hibernation when the overlap is greater than 30%. If the relative contribution of an agent is less than a certain threshold, that agent is put into hibernation.

[0151] Sensor nodes are deployed in an optimized manner according to the optimization scheme. The mesh structure supports multi-hop communication. When a sensor node fails, the adjacent sensor nodes automatically enhance the signal strength to compensate.

[0152] First, this scheme divides the monitoring area into several sub-regions and calculates the coverage entropy value of each sub-region. The magnitude of the coverage entropy value can characterize the risk of the sub-region being a monitoring blind spot, so that it can be given higher priority in subsequent monitoring deployment. This method is the first to apply information entropy theory to the field of geological disaster monitoring network deployment, establishing a brand-new risk assessment system for monitoring blind spots, so as to more accurately identify risks. By quantifying the risk of monitoring blind spots through mathematical modeling, it overcomes the subjectivity of traditional experience-based point placement, and significantly improves the accuracy of high-risk area identification. Furthermore, the geological risk weight in the coverage entropy model of this scheme is dynamically adjusted according to slope and lithology, realizing dynamic priority adjustment. The regional monitoring priority is updated in real time according to changes in the geological environment, while the traditional static weight allocation method cannot achieve this dynamic adaptation. This scheme is more versatile and applicable to various geological disaster scenarios such as landslides and debris flows, and has a wider applicability than dedicated algorithms for specific scenarios.

[0153] Secondly, in this scheme, each sensor node in the monitoring network is modeled as an intelligent agent, which then performs cooperative game optimization based on the coverage entropy value and the constructed objective function, updates the position and node state of the intelligent agent, and generates an optimization scheme.

[0154] The objective function guides the optimization direction in the collaborative game optimization process. It consists of the coverage contribution value and overlap penalty term of each agent with a coverage weight. By adjusting the coverage weight, the proportion of the coverage contribution value and overlap penalty term in the objective function can be changed. The coverage weight is dynamically adjusted by the geological disaster early warning level. The geological disaster early warning level is obtained by fusing real-time data from multiple sources in the agent and the monitoring area, thereby establishing a dynamic mapping relationship between geological disaster risk and optimization objective, and realizing adaptive adjustment of coverage weight. For monitoring areas with different types of geological disasters, the optimization objective is automatically adjusted according to real-time monitoring data. Compared with the traditional fixed weight algorithm, which suffers a sharp decline in performance when the environment changes abruptly, this scheme can be steplessly adjusted between different objective requirements to meet the needs of different scenarios. Moreover, multi-source data fusion decision-making, which integrates multi-dimensional data to generate optimization strategies, has higher accuracy than single-index decision-making. In addition, there is cooperation among the agents in the collaborative game optimization. If any agent fails, the optimization process will compensate for the monitoring gap caused by its failure. The sensor node failure tolerance is high, which improves robustness and can cope with sudden geological disasters.

[0155] Furthermore, the collaborative game optimization adopts a two-stage architecture of autonomous optimization and group coordination, breaking through the limitations of traditional single-stage optimization algorithms. It balances speed and global optimality. The autonomous optimization stage enables rapid response and solves the problem of long computation time in traditional centralized algorithms. The group coordination stage ensures reasonable allocation of global resources, avoids local optima, and achieves higher resource utilization. It reduces redundant nodes through intelligent sleep and wake-up mechanisms, which can reduce equipment costs by more than 30% compared to traditional full-coverage deployment methods. It supports automatic compensation for node failures and has significantly better robustness than traditional networks with fixed topology.

[0156] Finally, this solution combines information entropy theory and intelligent agent game theory, which is more optimized than single intelligent agent game theory, improves convergence efficiency in the game process, and adapts to the dynamic changes in geological disaster scenarios. It also improves deployment efficiency, shortening the entire process time from data input to generating an optimized solution by more than 50%, reducing overall costs. While ensuring monitoring quality, it reduces equipment investment and maintenance costs by 30%-40%, enhances emergency response capabilities, and increases the response speed to sudden geological disasters by more than 3 times.

[0157] In summary, this solution enables dynamic optimization and deployment of the geological disaster monitoring network, covering high-risk areas and enhancing robustness to cope with sudden geological disasters.

[0158] This embodiment also provides a geological disaster monitoring network deployment system based on the coverage entropy theory, which adopts the above-mentioned geological disaster monitoring network deployment method based on the coverage entropy theory.

[0159] The above descriptions are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A geological disaster monitoring network deployment method based on coverage entropy theory, characterized in that, The method comprises the following steps: dividing the monitoring area into a plurality of sub-areas, calculating the coverage entropy value of each sub-area, comprising: calculating the coverage entropy value of each sub-area through a coverage entropy model; wherein the coverage entropy model is: ； wherein is a geological risk weight; is the probability that the sub-area is not monitored; wherein : ； wherein is the effective radius of the sensor, is the sub-area to the sensor node distance, is the number of sensors; modeling the sensor nodes as agents, and according to the coverage entropy value and a constructed objective function, performing collaborative game optimization, updating the position and node state of the agents, and generating an optimization scheme; wherein the objective function is composed of the coverage contribution value of the agents with set coverage weights and an overlap penalty term; the coverage weight is dynamically adjusted by the geological disaster warning level, and the geological disaster warning level is obtained by fusion calculation of multi-source real-time data of the agents and the monitoring area; The cooperative game optimization comprises: each intelligent agent independently moves, and maximizes a coverage contribution value of each intelligent agent according to a coverage entropy value of each sub-region , and a total contribution value of all intelligent agents is not reduced, until a movement displacement of all intelligent agents converges, and the coverage game ends; wherein is a grid coverage entropy of the intelligent agent , and is equal to a coverage entropy value of a sub-region where the intelligent agent is located; is a signal strength. 2.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 1, characterized in that, the objective function is: ； wherein is the total contribution value; grid coverage entropy for the agent , wherein sub-region where the agent is located; S is signal strength; for the agent a distance of a sub-region with the highest coverage entropy value on the movement route; effective radius for the agent; is an overlap penalty coefficient; is an overlap indicator function; wherein is a threshold for overlap, is an overlap, wherein is an agent effective monitoring area on a two-dimensional plane. 3.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 1, characterized in that, during the movement process, the agents adjust the movement according to the position update strategy; wherein the position update strategy is: wherein is the moving direction and step length of the agent ​ learning rate; is the entropy gradient coefficient; is the maximum coverage entropy value. 4.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 1, characterized in that, during the movement process, the agents adjust the signal strength according to the signal strength adjustment strategy; wherein the signal strength adjustment strategy is: wherein is an agent , the signal strength at the next time instance; for the agent , the signal strength at the current time ; to adjust the step size; is a sign function that determines the direction of adjustment: The entropy threshold is set according to the geologic risk type of the monitored area. 5.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 1, characterized in that, the collaborative game optimization further comprises: forming an alliance among the agents, evaluating the relative contribution degree of each agent in the alliance, and controlling the node state of the agent according to the size of the relative contribution degree. 6.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 5, characterized in that, the control of the node state of the agent comprises: building a communication topology among the agents, and broadcasting the node state data of the agents to form an alliance; calculating a coverage contribution value for each agent ; According to the objective function, the total contribution value is calculated ; Compare the coverage contribution value and the total contribution value of the agent to obtain a relative contribution degree ; Obtain the number of agents in the alliance, and according to the number of agents, calculate a judgment threshold value for redundancy judgment ; wherein is a redundancy threshold value, is the number of agents in the alliance; judging whether the relative contribution degree of the agent is less than a judgment threshold, if yes, the agent is retained, if not, the agent is controlled to sleep; updating the node state data of the agent, and judging whether all the agents in the alliance have judged the relative contribution degree, if yes, the optimization is ended, and the optimization scheme is generated. 7.The method for geological disaster monitoring network deployment based on the theory of coverage entropy according to claim 2, characterized in that, the coverage weight is dynamically adjusted by the geological disaster warning level, comprising: obtaining the displacement rate of the agent and the rainfall of the monitoring area; According to the displacement rate of the intelligent agent and the rainfall of the monitoring area, the geological disaster warning level is calculated : wherein and are the displacement rate of the agent and the rainfall normalized value of the monitoring area, respectively; According to the geological disaster warning level , adjust the coverage weight ; updating the coverage weight in the objective function; further comprising: monitoring whether the change amount of the geological disaster warning level exceeds a preset threshold, if yes, re-performing the collaborative game optimization, if not, continuing to monitor the change amount of the geological disaster warning level.

8. A geological disaster monitoring network deployment system based on coverage entropy theory, characterized in that, The geological disaster monitoring network deployment method based on the coverage entropy theory according to any one of claims 1-7 is adopted.

Images