A management and control system for island space resource normalization monitoring

Abstract

The application belongs to the technical field of data processing and administrative supervision, and discloses a management and control system for island space resource normalization monitoring, which comprises: a module for establishing a verification credit score account for suspected change patches identified by remote sensing; a scene correlation rule engine which stores weight and activity parameters and can dynamically modulate credit scores by using universal auxiliary information; and a module for closed-loop feedback adjustment of the weight and activity of the rules according to the verification results, which contains a forced dormancy and opportunistic activation mechanism for the rules. The application builds an adaptive resource scheduling and rule self-evolution mechanism based on verification credit, so that the judgment logic of the supervision system changes the analysis focus from isolated physical changes to comprehensive evaluation of changes combined with space-time scene information, and adjusts the judgment rules through closed-loop feedback, so that the system can adapt to new risk patterns.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of island space resource normal monitoring management and control system, belong to data processing and administrative supervision technical field. BACKGROUND

[0002] It is essential to monitor the state of key facilities and environment for a long time, and it is a common technical strategy in the field to use sensing optical cable for distributed monitoring due to its characteristics of passive, anti-electromagnetic interference and long-distance continuous measurement. The common method is to obtain temperature strain or vibration information along the line by analyzing the specific physical quantity change of backscattered light in the optical cable, and then determine whether the macro state of the monitoring object has changed.

[0003] However, when this monitoring method is transferred from the laboratory or ideal working condition to the real complex and uncertain island environment, the inherent design limitation begins to appear. The existing monitoring system generally regards the optical cable itself as a uniform passive sensing medium. The core of its analysis model is to identify those abnormal sections with clear signal characteristics caused by physical events such as large deformation and severe temperature change. However, this method is greatly restricted in monitoring sensitivity and reliability for identifying early risk hidden dangers with slow change process and weak signal characteristics, or highly similar to environmental background noise, such as the gradual destruction of vegetation in the initial stage of illegal construction of small foundation protection zone or the slight displacement of key facilities in adverse weather background.

[0004] At a deeper level, the root of this limitation lies in an inherent contradiction within existing technologies: the conflict between reliance on high signal-to-noise ratio (SNR) anomaly signals and the need to identify early risks with low SNR. To avoid false alarms caused by massive environmental noise, such as tides, waves, and normal human activities, the system must set high alarm thresholds. However, this inevitably comes at the cost of sacrificing the ability to detect weak, slowly changing risk signals. While the industry has attempted to filter out noise using more complex algorithms, this often introduces a high computational burden and may filter out atypical, genuine risk signals due to over-optimization of the model. Specifically... Existing technologies suffer from the following shortcomings: 1. Monitoring systems lack effective and reliable identification mechanisms for early risk events characterized by subtle and gradual changes in signal features; 2. The system's judgment of monitoring results heavily relies on isolated changes in physical quantities, lacking the ability to fuse and analyze multi-dimensional contextual information at the time of the event, making it difficult to distinguish between high-risk man-made anomalies and low-risk natural or legitimate changes; 3. Under the current monitoring logic, limited and costly on-site verification resources are often heavily consumed in identifying various suspected anomalies with similar signal features but vastly different actual risk values, leading to delayed responses to truly time-sensitive risks. Therefore, the technical problem this invention aims to solve is how to establish a new monitoring and processing mechanism that, without significantly increasing system complexity, can effectively identify and distinguish the most valuable early subtle risk signals from massive amounts of low signal-to-noise ratio background data, and prioritize limited verification resources towards these key targets. Summary of the Invention

[0005] This invention provides a management and control system for routine monitoring of island spatial resources. Its main purpose is to solve the problem that existing technologies are unable to effectively identify and distinguish the most valuable early weak risk signals when faced with massive amounts of low signal-to-noise ratio background data, which leads to low efficiency in the allocation of high-value verification resources.

[0006] To achieve the above objectives, the present invention provides a management and control system for routine monitoring of island spatial resources, the management and control system comprising:

[0007] A credit account creation module is used to create a verification credit score account and set an initial credit score for each suspected change patch identified by remote sensing imagery;

[0008] A context association rule engine is used to store multiple context association rules, and each context association rule is associated with a configurable weight parameter and an activity parameter updated by the system.

[0009] a credit score dynamic modulation module configured to obtain a general-purpose auxiliary information associated with the suspected change plot in space-time, and invoke a scenario correlation rule engine to dynamically modulate the credit scores of a plurality of verification credit score accounts by using the general-purpose auxiliary information;

[0010] a verification resource scheduling module configured to sort a plurality of suspected change plots according to the dynamically modulated verification credit scores, and schedule verification resources to verify the suspected change plots in the front of the sorting according to the sorting result;

[0011] a closed-loop feedback adjustment module configured to obtain a verification result of the verification resources, and first adjust a weight parameter of a scenario correlation rule that triggered the verification and update an activity parameter of the scenario correlation rule according to the verification result, and then temporarily reduce the weight parameter of the scenario correlation rule when the weight parameter and the activity parameter of the scenario correlation rule both reach a first threshold value stored in the system, and schedule the verification resources to verify a suspected change plot identified by another scenario correlation rule whose activity parameter is lower than a third threshold value stored in the system with a second probability stored in the system.

[0012] Preferably, the credit score dynamic modulation module, when dynamically modulating by using the general-purpose auxiliary information, is configured to reduce the credit score of a corresponding verification credit score account when the general-purpose auxiliary information indicates that the change of the suspected change plot matches the characteristics of an archived natural phenomenon model or a publicly announced legal activity plan, and increase the credit score of the corresponding verification credit score account when the general-purpose auxiliary information indicates that the change of the suspected change plot matches the characteristics of an archived potential illegal behavior pattern.

[0013] Preferably, the control system further comprises a key carrier database configured to store geographic location information of facilities or regions that have an impact on regional systematic security, and the control system is configured to set a disaster response mode that is activated when the number of monitored suspected change plots exceeds a number threshold value stored in the system within a certain time period; in the state that the disaster response mode is activated, the dynamic modulation operation of the credit score dynamic modulation module is further configured to: assign a verification credit score of a suspected change plot that coincides with a geographic location stored in the key carrier database with a system preset highest priority value.

[0014] Preferably, when the management and control system schedules the UAV to check the suspected change plot, the management and control system further comprises: an opportunity investigation corridor generation module, configured to set an opportunity investigation corridor with a determined width around the flight route when the flight route is planned based on the suspected change plot with a higher ranking; an in-flight scanning instruction module, configured to instruct the UAV to scan the opportunity investigation corridor when the UAV flies along the flight route, so as to obtain the real scene features of other suspected change plots in the corridor; and a dynamic task adjustment module, configured to update the check credit score of other suspected change plots in the corridor in real time according to the obtained real scene features, and generate an adjustment instruction to suspend the current check task or add a new check task after the current check task is completed based on the updated check credit score.

[0015] Preferably, the management and control system further comprises: a decision stalemate identification module, configured to identify the suspected change plot as a stalemate target if the suspected change plot is associated with a communicable active entity when the modulated check credit score is in a stalemate interval defined by a fourth threshold value and a fifth threshold value stored in the system; and an information probe module, configured to automatically send a challenge instruction containing a unique identification code and having a preset data format to the active entity through a communication channel for the stalemate target, and re-modulate the check credit score of the stalemate target.

[0016] Preferably, the credit score dynamic modulation module dynamically modulates the credit score of the check credit score account of any suspected change plot according to the following formula: wherein, is the modulated check credit score, is the initial credit score, is the total number of scenario association rules hit by the general auxiliary information triggering the current modulation is the weight parameter of the scenario association rule hit by the first rule, is a risk factor bound with the risk qualitative of the scenario association rule of the first rule, and the risk factor is derived from a lookup table stored in the system and mapping the rule identification and the risk factor value. Preferably, the general auxiliary information comprises at least one of the following: tide data information, weather data information, fishing moratorium calendar information, vessel automatic identification system data information, and publicized engineering project information.

[0017] Preferably, the check resource is a UAV, and the resource is selected from a law enforcement ship and a manual interpretation expert.

[0018] Preferably, the check resource is a UAV, and the resource is selected from a law enforcement ship and a manual interpretation expert.

[0019] ​Preferably, the response behavior analysis module adjusts the specific operation rule of the response behavior characteristics of the active entity to the query instruction as follows: if the response of the active entity is not received within a certain time period and the position of the active entity does not change, the verification credit score is increased; if the response of the active entity is received within a certain time period and the position of the active entity moves away from the suspected change spot, the verification credit score is decreased.

[0020] Preferably, the closed-loop feedback adjustment module adjusts the specific operation rule of the weight parameter of the scenario correlation rule triggered in the current verification as follows: when the data characteristics of the verification result match the data characteristics stored in the high-risk event characteristic library in the system, the weight parameter of the corresponding scenario correlation rule triggered in the current modulation is increased; when the data characteristics of the verification result match the data characteristics stored in the no-risk event characteristic library in the system, the weight parameter of the corresponding scenario correlation rule triggered in the current modulation is decreased.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] 1. The method establishes a verification credit score account for the suspected change spot identified by the remote sensing image, dynamically modulates the score by using the tide shipping and other inclusive auxiliary information, and adjusts the scenario correlation rule according to the verification result on which the modulation is based; the combination of these technical features makes the analysis and judgment of the monitoring target no longer rely on the physical form of the spot itself, but considers the spot change in a dynamic scenario containing time environment and surrounding activities, and the judgment logic of the system gradually approaches the recognition of the potential intention behind the change through the continuous operation of verification and feedback, rather than staying in the simple classification of physical phenomena.

[0023] 2. When the unmanned aerial vehicle is dispatched to verify the target spot with a high verification credit score, the method does not only perform a point-to-point flight task, but also presets an opportunity survey corridor based on the flight route and scans other low-score spots in the corridor during the flight to obtain real scene characteristics; this process changes the execution stage of the unmanned aerial vehicle verification task from a simple displacement process to a secondary information collection and value discovery process, and the obtained real scene characteristics can be used to update the verification credit score of the spots in the corridor in real time, thereby dynamically adjusting the current verification task, so that the investment of single verification resource has the opportunity to discover and confirm potential risk points that are not in the priority verification sequence due to weak initial signals.

[0024] 3、When the system faces a suspected change plot with intermediate credit score, insufficient decision information, and associated with a communicable active entity such as a ship, instead of passive waiting or directly upgrading the verification method, the method first sends an automated challenge instruction to the entity through the standard communication channel; the system is not concerned about the content of the reply to the challenge, but the response behavior characteristics of the entity after receiving the instruction, and this behavior characteristics is a new information input, which is used to adjust the credit score of the plot, which provides a way to generate decision basis actively, avoiding the premature consumption of monitoring resources or ignoring potential risks.

[0025] 4、The feedback step of the method for adjusting the weight of the scenario correlation rule also includes setting an activity parameter for the rule, temporarily reducing the weight of the rule that reaches the merit threshold for weight and activity, and raising the verification priority of the plot identified by the low-activity rule with a preset low probability; this establishes a new metabolism mechanism at the rule level in the system, which actively creates opportunities for emerging or rare pattern rules that are not mainstream to obtain verification and feedback, avoiding the solidification of the system's judgment ability due to excessive reliance on verified successful experience, and helping to maintain the identification sensitivity to unknown or new risk patterns. BRIEF DESCRIPTION OF DRAWINGS

[0026] Fig. 1 The adaptive processing flowchart of the present application based on credit score and closed-loop feedback;

[0027] Fig. 2 The dynamic evaluation and hierarchical disposal decision flowchart of a single plot of the present application;

[0028] Fig. 3 The monitoring system deployment and data flow architecture diagram of multi-source information fusion of the present application. DETAILED DESCRIPTION

[0029] In order to make the technical solutions and advantages of the present application clearer, the present application will be further described in detail below. Obviously, the described embodiments are part of the embodiments of the present application, rather than 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.

[0030] The invention discloses a kind of island space resource normality monitoring management and control system, it includes a credit account establishment module, a scene correlation rule engine, a credit score dynamic modulation module, a check resource scheduling module and a closed-loop feedback adjustment module;When system runs, first by credit account establishment module for remote sensing image identified multiple suspected change map spot establishment check credit score account, subsequently, credit score dynamic modulation module calls scene correlation rule engine, and adjust each account score in combination with the acquired universal auxiliary information, check resource scheduling module is sorted according to the score after adjustment to map spot and dispatches resource to be verified, finally, closed-loop feedback adjustment module adjusts the parameter of corresponding rule in scene correlation rule engine according to verification result, forms a complete processing procedure for optimizing administrative supervision resource allocation;In a specific embodiment, when marine management department is monitored in jurisdictional area, the technical problem of uninhabited island is that remote sensing technology can obtain a large number of suspected change map spots, and field verification resource is limited, it is difficult to effectively distinguish all suspected changes;To deal with this problem, the credit account establishment module of the present application is configured as: receiving the geographic information data of multiple suspected change map spots output by upstream remote sensing image processing system, and establishing a data structured check credit score account for each map spot, the account includes map spot unique identifier, geographic coordinates, static attribute and a numerical field, i.e. The module is assigned to account initial credit score according to the static attribute of map spot Through a predetermined rule base, for example, a map spot located in the sea base point island, it Can be set to 20, and a map spot of the same area located in ordinary uninhabited island, it It is set to 10, this step will indistinguishable map spot data be processed into structured account data containing initial value evaluation.

[0031] Given that static attributes of land parcels alone are insufficient to assess their risk, the system employs the following procedures for dynamic evaluation: A credit score dynamic modulation module is activated to acquire general auxiliary information spatiotemporally related to suspected changed land parcels. This information includes tidal data, meteorological data, fishing moratorium calendar information, Automatic Identification System (AIS) data for waterways, and publicly disclosed project information. Simultaneously, the system's internal scenario association rule engine pre-stores multiple scenario association rules. Each rule defines the relationship between land parcel changes and specific general auxiliary information; for example, rule one: IF (land parcel type = mudflat change) AND (time). =During spring tide) THEN (risk assessment = natural correlation), Rule 2: IF (patch type = tidal flat change) AND (time = low tide) AND (nearby AIS vessel type = sand dredger) THEN (risk assessment = strong correlation with violation); After the credit score dynamic modulation module obtains information, it calls the engine to traverse each account. If the spatiotemporal characteristics of the patches in the account match the conditions of a certain rule, a score modulation is triggered, transforming the analysis of isolated physical changes into a comprehensive assessment in a multi-dimensional scenario; To ensure the determinism of the modulation process, each scenario association rule is associated with a weight parameter that can be updated by the system. and a risk factor This risk factor It originates from a lookup table within a system that maps rule identifiers to risk factor values; for example, rules that are qualitatively classified as naturally correlated. The rule is set to -1, and the risk is characterized as a strong correlation with the violation. Set to +1; the specific operating rules for the credit score dynamic modulation module to modulate the score of any account are defined by the following formula: ,in, This is the dynamically modulated credit score. As the initial credit score, This represents the total number of rules that were matched in this modulation. For the first The weight parameters of the rule, This is the risk factor tied to the rule, and the quantification procedure ensures the objectivity and repeatability of score modulation.

[0032] After calculating the credit scores for all suspected changed map features, the verification resource scheduling module was configured to allocate credit scores to all verification accounts based on their latest scores. The scores are sorted in descending order to generate a priority verification list. Accordingly, based on this sorting result, the module prioritizes available verification resources, such as drones, law enforcement vessels, or human experts, for verification of the top-ranked map features. This scheduling mechanism directs verification resources to map features with higher verification value after evaluation, thereby improving the targeting of resource allocation. To enable the rule system to adapt to changes in the environment and behavioral patterns, the closed-loop feedback adjustment module executes the following procedures: After a field verification is completed, the verification result is input into this module. The module matches the result data characteristics with the high-risk or no-risk event feature database within the system to determine the validity of the verification. If a high-credit-score map feature is confirmed as a high-risk event, the weight parameter of the corresponding scenario association rule triggering this modulation is increased. If the event is confirmed to be risk-free, the weight parameter of the corresponding rule will be reduced. While adjusting the weights, the system updates the rule's activity parameter to record the frequency of its invocation. Furthermore, when both the weight and activity parameters of a rule reach a first threshold stored in the system, the weight parameter of that rule is temporarily reduced for a subsequent period, providing other rules with an opportunity to be verified. In addition, the system schedules verification resources with a second probability stored in the system to verify the patches identified by rules whose activity parameters are below a third threshold stored in the system. This provides the possibility for rules targeting rare or novel patterns to be tested and activated, thus forming a closed loop that can dynamically adjust the effectiveness of rules to adapt to changes in the monitoring scenario.

[0033] Example 1: In the daily monitoring work of an island management department, the system received two suspected change patches of different natures on the same day. Its processing flow, by introducing multi-dimensional contextual information, effectively distinguished the change patches of different natures. Within a specific regulatory cycle, the credit account establishment module identified and generated two initial credit scores. Both are suspected change patches with a score of 10, namely patch A and patch B. Patch A shows a change in the edge morphology of a large mudflat, while patch B is a small change in surface color within a secluded harbor. In a monitoring method that relies solely on the physical characteristics of remote sensing images, managers may prioritize patch A because it is larger, or they may choose it randomly because they cannot distinguish the risks of the two, potentially leading to a misallocation of verification resources. In one embodiment of the present invention, the credit score dynamic modulation module is triggered and calls the scenario association rule engine, while simultaneously accessing tidal data and automatic identification system data for waterways. For patch A, the scenario association rule engine matches a rule whose condition is that the change time of patch A coincides with the spring tide cycle shown in the tidal data. The risk factor bound to this rule is... The score was -1, and the system accordingly downgraded the credit score of patch A. Meanwhile, for patch B, the engine triggered another rule, the condition being that the change in patch B occurred during nighttime low tide, and that the Automatic Identification System (AIS) data showed a sand dredging vessel lingering at that location for an extended period. This rule was linked to a risk factor. The credit score is increased by +1. Based on this, the credit score dynamic modulation module increases the credit score of patch B. This processing method enables the system to effectively suppress misjudgments caused by normal natural phenomena without reducing its sensitivity to weak signals.

[0034] Furthermore, the resource scheduling module verifies the dynamically modulated... The scores were ranked, and at this moment, the verification credit score of patch B was higher than that of patch A. The system then instructed a drone to fly to the location of patch B first for verification. This action changed the decision-making basis for verification from simply relying on the physical characteristics of the patches to verifying the verification value based on multi-dimensional scenario assessment. This changed the goal of administrative management from indiscriminately seeing all changes to prioritizing the verification of events with high-risk characteristics. Ultimately, the verification resources were directed to the location of patch B, confirming an illegal sand mining operation, while patch A did not consume any verification resources. At the same time, the weight parameters of the relevant scenario association rules that triggered this successful judgment were also determined. The corresponding addition was made to the closed-loop feedback adjustment module; this complete process, by establishing credit accounts for isolated physical changes, dynamically modulating them using inclusive auxiliary information, and then using the verification results to provide closed-loop feedback on the modulation rules, forms a resource allocation architecture that adaptively matches limited administrative supervision resources with constantly changing risk dynamics.

[0035] Example 2: To verify the effectiveness of the closed-loop feedback adjustment mechanism in the technical solution of this invention, the following simulation experiment was conducted. The experimental platform was constructed based on real remote sensing images of a certain sea area for 24 consecutive months, full-time hydrological and meteorological data, waterway vessel records, and a manually verified ground authenticity database. This database marked 128 high-risk events and more than 4,500 risk-free natural change events that occurred during this period. To evaluate the system's learning ability, all rule weight parameters of the scenario association rule engine were used in the experiment. The initial values ​​were all set to 0.1 to simulate a system with no prior experience; the experiment included a control group and an experimental group, with the control group's system based solely on the initial credit score. For target ranking, the experimental group adopted the complete technical solution of this invention. Both systems simultaneously processed simulation data from 24 consecutive months and output their respective top 10 highest-priority verification targets. The output results were then compared with a ground-based realism database to count the number of correctly identified high-risk events. In the experimental group, after each round of comparison, its closed-loop feedback adjustment module updated the rule weight parameters based on the comparison results. .

[0036] The test results show that in the long-term operation, the test group using the closed-loop feedback adjustment mechanism is continuously superior to the control group relying only on static physical feature analysis in the number of high-risk event identification; in the initial stage of the test, i.e., the first month, the correct identification numbers of the test group and the control group are both 1, and at this time, the system performances of the two groups have no difference; with the test proceeding, by the 12th month, among the total of 6 high-risk events in the month, the correct identification number of the test group has increased to 5, while the identification number of the control group is 0; this trend is continued in the subsequent test, and by the 24th month, among the total of 4 high-risk events in the month, the test group identifies all 4, and the control group only identifies 1; the appearance of this performance difference is directly related to the change of the average weight of the core rules in the test group , which increases from the initial 0.10 to 0.62 by the 12th month, and reaches 0.93 by the 24th month, while the control group lacks this adjustment mechanism, and its judgment basis does not change; data confirms that compared with the static analysis method, the technical solution of the present application can improve the identification and priority ranking of high-risk events under complex background through closed-loop feedback learning, so as to more effectively guide the limited verification resources to the real risk points.

[0037] Embodiment 3: This embodiment combines Figs. 1 to 3 to explain a kind of island space resource normalization monitoring management and control system, as Fig. 1 , the process starts from the suspected change plot determined by the remote sensing image identification result processing, first, credit account establishment module is established for each plot and is given initial score value with exclusive credit score account, subsequently, credit score dynamic modulation module comprehensively external input's universal auxiliary information, such as tide, weather and ship AIS etc., and call the rule, weight and activity parameter stored in scenario correlation rule engine, dynamically assess the credit score of each account, based on the score after evaluation, verification resource scheduling module executes sequencing operation, to schedule high value target to carry out priority verification, and generate priority verification list and scheduling instruction output to verification execution terminal, after obtaining verification result, the result is fed back to closed-loop feedback adjustment module, this module updates and optimizes the rule weight and activity in scenario correlation rule engine according to the result, to form a complete self-adaptive closed-loop processing architecture.

[0038] as Fig. 2As shown, after receiving the map patch data, the system assigns an initial credit score to it through the initial establishment phase, and then enters the core state of dynamic evaluation. In this state, if the credit score of the map patch increases, its state transitions to high priority pending verification, and after verification, it is archived as a high-risk event or a risk-free event based on the result. If the credit score of the map patch decreases, its state transitions to low priority, and it can be directly archived as risk-free if it is not scheduled or after the cycle ends. In particular, when the credit score enters the preset deadlock range, the system will trigger the decision deadlock handling mechanism. After sending a challenge command, the system will feed back new information to the dynamic evaluation phase based on the response behavior to break the deadlock and drive the further evolution of the state.

[0039] like Fig. 3 As shown, its core is a monitoring and command center, which houses application servers and database servers. Together, these servers support the island's spatial resource management system and manage the core database. The center receives two main types of external input information: remote sensing image data from satellite remote sensing systems and general auxiliary information from auxiliary information sources. Based on its internal processing logic, it issues instructions to external execution units. This is mainly done by issuing scheduling instructions containing flight path planning to UAVs and receiving their verification data feedback, or by issuing scheduling instructions to law enforcement vessels to perform on-site verification. At the same time, the system also interacts with expert workstations to support manual data interpretation and feedback, thus forming a complete operational system of information input, intelligent processing, instruction output, and result feedback.

[0040] Example 4: Before deploying the control system of this invention to a new island monitoring area, one engineering problem it faces is calibrating a series of core operating parameters of the scenario association rule engine within the system to adapt its initial state to the unique environment and event patterns of the area. To this end, this example discloses a parameter calibration procedure based on historical data backtracking analysis. This procedure utilizes a historical dataset containing ground reality annotations to provide a reproducible process for setting the core parameters of the system. This procedure first applies to the closed-loop feedback adjustment module for adjusting the rule weight parameters. Adjustment step size Setting this parameter requires a trade-off between learning rate and system stability. Too small a step size leads to slow weight updates, while too large a step size may cause weight values ​​to oscillate during feedback adjustments. The calibration method involves using the first six months of historical data to set a set of adjustment step sizes, covering the range of 0.01 to 0.2 with 0.01 intervals. Independent short-cycle simulations are then performed, and the curves showing the system's correct recognition rate for high-risk events over time are recorded for each step size setting. Finally, the adjustment step size corresponding to the curve that allows the recognition rate to reach a stable state most quickly without significant oscillations is selected as the operating parameter for deployment in that region. .

[0041] Further, the procedure calibrates three key thresholds of the control rule metabolism mechanism. After a complete simulation run covering all historical data is completed, the system statistically sorts the activity parameters of all activated rules in the scenario-related rule engine, takes the 95th percentile of the distribution as the first threshold for triggering forced dormancy, and takes the 5th percentile of the distribution as the third threshold for defining dormant rules. The second probability of the scheduling verification resource for opportunistic verification of the identified plot by the dormant rules is set according to the proportion of rare high-risk events in the historical data set. The probability value is set to be linearly related to the proportion value, thereby associating the exploration behavior intensity of the system with the diversity of the regional risk pattern. The execution of the calibration procedure provides a set of configuration benchmarks based on historical data analysis for the initial deployment of the system, replacing the experience-dependent parameter setting method.

[0042] In order to couple the judgment logic of the management and control system to the administrative priorities and environmental characteristics of a specific regulatory area before it is put into practical application, a standardized data preset procedure needs to be executed. The procedure first acts on the construction of the key carrier database. Through an interactive interface, the operator imports the planning data released by the official of the region, which contains the geographic location and protection level of various types of infrastructure and sensitive areas, into the system. The system analyzes and extracts the facilities or areas marked as having an impact on the regional systematic safety, including geographic entities such as territorial base points, submarine cable landing points, and key protected coral reef core areas. The geographic coordinates and official classification attributes of these entities are structured and stored in the key carrier database. This process provides an object list for the priority response logic in the subsequent disaster response mode.

[0043] Further, the procedure uses a historical remote sensing image data set that has been completed with ground truth labeling to fill the high-risk and no-risk event feature libraries. For each confirmed event in the historical data set, the system automatically extracts its multi-dimensional data features in the remote sensing image, including texture descriptor spectral index and morphological parameters. After associating these feature vectors with the authenticity label of the event representing illegal structures or tidal erosion, they are respectively stored in the high-risk event feature library or the no-risk event feature library. At the same time, the system also statistically analyzes the number of newly added suspected change plots per unit time in the historical data set, calculates the mean and standard deviation under normal weather conditions and during extreme weather periods such as typhoons, and sets a number threshold for activating the disaster response mode based on the statistical results. The threshold is set to be three standard deviations above the mean number under normal conditions.

[0044] Example 6: To ensure that the control system of the application can make accurate quantitative judgments on the response behavior of the real scene features and communicable active entities obtained by the unmanned aerial vehicle in a specific regulatory area, the system needs to perform an engineering procedure for building a localized baseline reference model before formal operation; the procedure aims to generate a bimodal baseline reference model containing visual features and behavioral responses, which will serve as the core basis for subsequent system adjustments and decision-making deadlock handling; when constructing the visual feature sub-model of the baseline model, the operator uses a historical aerial image dataset containing ground truth annotations to extract the standardized real scene feature vectors of the confirmed high-risk targets, such as small illegal structures or specific types of vegetation damage, from the dataset. The vector is composed of a set of quantifiable parameters such as image edge density distribution and specific spectral band reflectance ratio. These feature vectors and their corresponding risk type labels are stored together in the system's real scene feature template library, serving as the basis for visual judgment. When the unmanned aerial vehicle performs on-the-go scanning within the opportunity survey corridor, the real scene feature vector of any image patch obtained by the unmanned aerial vehicle will be compared with the template vector in the library for cosine similarity. The comparison result provides a quantitative input for the subsequent dynamic task adjustment module to update the credit score.

[0045] When constructing the behavioral response sub-model of the baseline model, the procedure defines a standardized behavior state transition judgment logic. After the information probe module sends a challenge instruction to a ship entity associated with a deadlock target, the response behavior analysis module determines the behavior state of the entity within 15 minutes after the instruction is sent based on its speed and position data. If the entity's speed increases from below 1 knot to above 3 knots and the angle between the heading angle and the location of the suspected change patch is greater than 90 degrees, its behavior state is determined to be evasive, triggering a decrease in the corresponding credit score. If the entity's position does not change and its speed remains below 1 knot, its behavior state is determined to be confrontational, triggering an increase in the corresponding credit score. This model converts the qualitative assessment of response behavior into a deterministic state machine judgment based on spatiotemporal parameters.

[0046] Example 7: To further improve the adaptability and judgment accuracy of the system in a specific regulatory area, the following fine-tuning procedure for core module parameters can also be performed during the initialization phase of system deployment; before the control system is put into operation, the working parameters of the decision-making deadlock identification module are calibrated, and the fourth threshold and the fifth threshold used to define the deadlock interval are determined by the following procedure: First, extract a ground truth database containing historical verification results, calculate the dynamic modulation credit score of all suspected change patches in the database, and statistically analyze the overall numerical distribution; then, associate the score with the final verification result of the patch, i.e., high-risk event or no-risk event; the fifth threshold A score point is determined, and all patches with scores higher than the score point have a proportion of high-risk events no less than a high-confidence parameter preset in the system, for example, 0.9 A fourth threshold value is determined as another score point, and all patches with scores lower than the score point have a proportion of no-risk events no less than a low-risk confirmation parameter preset in the system, for example, 0.9 The calibration procedure makes the determination of the stalemate interval directly related to the historical risk data characteristics of a specific regulatory area.

[0047] For the response behavior analysis module, the specific operation rules of the re-modulation of the verification credit score of the stalemate target are further optimized by a localized response baseline model, and the construction procedure of the model is as follows: the system first analyzes the response behavior of all active entities, such as ships, after receiving the standardized interrogation instruction, and establishes statistical distribution models of response time and position and velocity vector changes of the entities according to the entity type; when a new stalemate target appears and completes the interrogation, the response behavior parameters of the associated active entity will be input into the statistical distribution model of the same type of entity for comparison, and the system calculates a deviation index according to the comparison, which quantifies the degree of deviation of the current response behavior from the baseline of the same type of behavior; finally, the re-modulation operation of the verification credit score is set to be proportional to the numerical value of the deviation index, and this method uses a statistical baseline to replace a fixed rule to handle the behavior differences of different active entities in different scenarios.

[0048] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. An island space resource normalization monitoring management and control system, characterized in that, The management and control system comprises: a credit account establishment module, configured to establish a verification credit score account for each suspected change plot identified by the remote sensing image and set an initial credit score for the account; a scenario correlation rule engine, configured to store a plurality of scenario correlation rules, and each scenario correlation rule is associated with a configurable weight parameter and an activity parameter updated by the system; a credit score dynamic modulation module, configured to obtain universal auxiliary information associated with the suspected change plot in space and time, and call the scenario correlation rule engine to dynamically modulate the scores of the plurality of verification credit score accounts by using the universal auxiliary information; a verification resource scheduling module, configured to sort the plurality of suspected change plots according to the dynamically modulated verification credit scores, and schedule verification resources to verify the suspected change plots in the front of the sorting according to the sorting result; a closed-loop feedback adjustment module, configured to obtain the verification result of the verification resources, and according to the verification result, first adjust the weight parameter of the scenario correlation rule triggering the verification and update the activity parameter of the scenario correlation rule, and then when the weight parameter and the activity parameter of a scenario correlation rule both reach a first threshold value stored in the system, temporarily reduce the weight parameter of the scenario correlation rule in a subsequent period, and with a second probability stored in the system, schedule verification resources to verify the suspected change plot identified by another scenario correlation rule whose activity parameter is lower than a third threshold value stored in the system; Wherein, the specific operation rules of the credit score dynamic modulation module for dynamically modulating the score of the verification credit account of any suspected change plot are defined by the following formula: Wherein, is the verification credit score after dynamic modulation, is the initial credit score, is the total number of scenario correlation rules hit by the inclusive information triggering this modulation, is the weight parameter of the scenario correlation rule hit by the first rule, is a risk factor bound with the risk qualitative of the first scenario correlation rule, and the risk factor is derived from a lookup table stored in the system and mapping the rule identification and the risk factor value. and the universal auxiliary information includes at least one of tide data information, weather data information, fishing moratorium calendar information, automatic identification system data information of the channel ship, and announced engineering project information. 2.The island space resource normalizing monitoring and management system according to claim 1, characterized in that, When the credit score dynamic modulation module dynamically modulates by using the universal auxiliary information, when the universal auxiliary information indicates that the change of the suspected change plot matches the characteristics of the archived natural phenomenon model or the characteristics of the announced legal activity plan, the credit score dynamic modulation module is configured to reduce the score of the corresponding verification credit score account; and when the universal auxiliary information indicates that the change of the suspected change plot matches the characteristics of the archived potential illegal behavior mode, the credit score dynamic modulation module is configured to increase the score of the corresponding verification credit score account. 3.The island space resource normalizing monitoring and management system according to claim 1, characterized in that, The management and control system further comprises a key carrier database configured to store geographic location information of facilities or regions that have an impact on regional systematic security; and the management and control system is configured to set a disaster response mode, which is activated when the number of monitored suspected change plots exceeds a number threshold value stored in the system within a certain time period; in the state that the disaster response mode is activated, the dynamic modulation operation of the credit score dynamic modulation module is further configured to: assign the verification credit score of the suspected change plot that coincides with the geographic location stored in the key carrier database with a system preset highest priority value.

4. The management and control system for island space resource normalization monitoring according to claim 1, characterized in that, When the resource scheduling module schedules the verification resource for the UAV, the management and control system further comprises: an opportunity survey corridor generation module, configured to set an opportunity survey corridor with a determined width around the flight route when planning the flight route with the suspected change plot with a high ranking as the verification target; an on-the-way scanning instruction module, configured to instruct the UAV to scan the opportunity survey corridor when flying along the flight route, so as to obtain the real scene features of other suspected change plots in the corridor; and a dynamic task adjustment module, configured to update the verification credit score of other suspected change plots in the corridor in real time according to the obtained real scene features, and generate an adjustment instruction to suspend the current verification task or add a new verification task after the completion of the current verification task based on the updated verification credit score.

5. The management and control system for island space resource normalization monitoring according to claim 1, characterized in that, The management and control system further comprises: a decision stalemate identification module, configured to identify the suspected change plot as a stalemate target if the suspected change plot is associated with a communicable active entity when the modulated verification credit score is within a stalemate interval defined by a fourth threshold value and a fifth threshold value stored in the system; and an information probe module, configured to automatically send a challenge instruction containing a unique identification code and having a preset data format to the active entity through a communication channel for the stalemate target, and to re-modulate the verification credit score of the stalemate target. 6.The island space resource normalizing monitoring and management system according to claim 1, characterized in that, The specific operation rule of the closed-loop feedback adjustment module for adjusting the weight parameter of the scenario correlation rule triggered in the current verification is: when the data features of the verification result match the data features in the high-risk event feature library stored in the system, the weight parameter of the corresponding scenario correlation rule triggered in the current modulation is increased; when the data features of the verification result match the data features in the no-risk event feature library stored in the system, the weight parameter of the corresponding scenario correlation rule triggered in the current modulation is decreased.

Images