Smart fishery supervision method and system based on multi-source data fusion

Abstract

The invention relates to the technical field of fishery management, in particular to an intelligent fishery supervision method and system based on multi-source data fusion, and the method comprises the steps: obtaining the multi-source trajectory data of a fishing boat, and carrying out the data fusion, and obtaining the precise trajectory data of the fishing boat; performing trajectory feature analysis, trajectory behavior pattern analysis and high-sensitivity point identification on the accurate trajectory data to obtain trajectory distribution of the fishing boat; obtaining multi-source fishery resource data and performing data fusion to obtain a fishery resource evaluation vector; the fishery resource prediction model processes the fishery resource evaluation vector to obtain fishery resource distribution; performing spatial coincidence analysis, time coincidence analysis and intensity coincidence analysis on the trajectory distribution of the fishing boats and the distribution of fishery resources to obtain analysis results; and an analysis result and management data are obtained and processed, and if abnormal behaviors exist, early warning is carried out.

Description

Technical Field

[0001] The present invention relates to the technical field of fishery management, and in particular to a smart fishery supervision method and system based on multi-source data fusion. Background Art

[0002] With the increasing shortage of fishery resources and the intensification of ecological and environmental problems, the supervision of fisheries continues to increase. Fisheries are not only related to food safety and ecological balance, but also an important economic pillar in many coastal and inland areas. The traditional fishery supervision and management model is mainly based on manual inspections, paper records and periodic reporting. It is not only inefficient and costly, but also difficult to detect violations and ecological risks in a timely manner. The response speed and handling capacity for problems such as illegal fishing, over-limit operations, and water pollution emissions are relatively weak.

[0003] In the context of rapid development of informatization, emerging technologies such as the Internet of Things, big data, cloud platforms and artificial intelligence have provided new means for fishery supervision and management. Smart fishery supervision and management aims to improve the comprehensiveness and timeliness of data collection by building a multi-level perception and decision-making system covering waters, fishing vessels, breeding sites and regulatory agencies, and to achieve visual, quantifiable and traceable management of the entire process of fishery production.

[0004] The application of big data technology in smart fishery supervision is manifested in the fusion processing of multi-source heterogeneous data, abnormal behavior identification, risk warning model construction, compliance assessment and analysis, etc. For example, by centrally processing and modeling data such as fishing vessel trajectories, water environment parameters, fishing quantity and operation time, it is possible to intelligently identify illegal fishing, cross-border operations, unlicensed breeding and other illegal behaviors; at the same time, combined with video surveillance and aquaculture water quality dynamic data, an ecological risk warning mechanism can be constructed to assist regulators in making real-time decisions.

[0005] Therefore, there is a need for a smart fishery supervision method and system based on multi-source data fusion that can connect the entire chain of data collection, processing, analysis and supervision decision-making, achieve comprehensive supervision of fisheries, thereby improving supervision efficiency, reducing management costs, and facilitating the sustainable utilization of fishery resources and the healthy development of the industry. Summary of the invention

[0006] The object of the present invention is to provide a smart fishery supervision method and system based on multi-source data fusion to improve the supervision efficiency of fishery resources. By obtaining the multi-source trajectory data of fishing vessels and performing data fusion, accurate trajectory data of fishing vessels is obtained; trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification are performed on the accurate trajectory data to obtain the trajectory distribution of fishing vessels; multi-source fishery resource data is obtained and data fusion is performed to obtain a fishery resource evaluation vector; the fishery resource prediction model processes the fishery resource evaluation vector to obtain the fishery resource distribution; spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis are performed on the trajectory distribution of fishing vessels and the fishery resource distribution to obtain the analysis results; the analysis results and management data are obtained and processed, and if there is abnormal behavior, a warning is issued.

[0007] To achieve the above object, the present invention provides the following technical solutions: A smart fishery supervision method based on multi-source data fusion, comprising: Obtaining the multi-source trajectory data of fishing vessels and performing data fusion to obtain accurate trajectory data of fishing vessels; Further, the multi-source trajectory data at least includes Automatic Identification System (AIS) data, fishing vessel monitoring system data, and fishing vessel trajectory information from satellite remote sensing; Segmenting the preprocessed multi-source trajectory data at preset time intervals, and generating several behavior pattern hypotheses for each trajectory segment based on local motion characteristics and data sources, and assigning an initial probability to each behavior pattern hypothesis; Iteratively processing the trajectory segments of fishing vessels in chronological order. If the data density of the trajectory segment is less than the threshold, use the data of the previous and subsequent trajectory segments, as well as the behavior pattern hypothesis and its probability of this trajectory segment, to predict the trajectory range of the fishing vessel within the time period, and perform fusion according to the data of the trajectory segment; If the data density of the trajectory segment is not less than the threshold, use Kalman filtering to smooth and refine the trajectory; After the preliminary reconstruction of all trajectory segments is completed, use a batch processing smoothing algorithm to perform global smoothing processing on the entire trajectory sequence to generate the accurate trajectory data of the final fishing vessel.

[0008] Performing trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of fishing vessels; Further, performing trajectory feature analysis on the accurate trajectory data, and the trajectory features at least include speed, course, position, and mooring / residence time; based on the trajectory feature analysis, identifying the behavior patterns of fishing vessels, and the behavior patterns at least include normal navigation, trawling, purse seining, fixed-point operation, stopping, drifting, turning, accelerating, or decelerating; Identify trajectory points or segments that deviate from normal or legal behavior as highly sensitive points of fishing vessels based on trajectory characteristics and behavior patterns; generate a trajectory distribution representing the activity range, activity density of fishing vessels, as well as the location and characteristics of highly sensitive points.

[0009] Obtain multi-source fishery resource data and perform data fusion to obtain a fishery resource assessment vector; Furthermore, the multi-source fishery resource data includes at least scientific survey data, fishery catch report data, environmental remote sensing data, and fishery biological data; Project the preprocessed multi-source fishery resource data onto a preset spatial grid uniformly; For point data, match and associate it with other data within the corresponding spatio-temporal grid; Construct a hierarchical fusion model, including bottom-layer fusion, middle-layer fusion, and top-layer fusion. Among them, the bottom-layer fusion includes proofreading, correcting, and fusing the real data of scientific survey data and fishery catch report data; the middle-layer fusion proofreads, corrects, and fuses environmental remote sensing data and fishery biological data; the top-layer fusion includes obtaining the results of the bottom-layer fusion and the middle-layer fusion, and performing fusion to construct a fishery resource assessment vector corresponding to the spatial grid.

[0010] The fishery resource prediction model processes the fishery resource assessment vector to obtain the fishery resource distribution; Furthermore, the fishery resource prediction model is a prediction model trained based on historical multi-source fishery resource data, and the model includes but is not limited to statistical models and machine learning models; The fishery resource prediction model receives the fishery resource assessment vector as input and outputs the fishery resource distribution within the regulatory area, including the abundance per unit area of different fishery resource types and the total resources within the regulatory area; The fishery resource distribution is expressed in the form of a spatial grid and contains time dimension information.

[0011] Conduct spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis on the trajectory distribution of fishing vessels and the fishery resource distribution to obtain the analysis results; Furthermore, the spatial coincidence analysis includes spatially overlaying the trajectory distribution of fishing vessels and the fishery resource distribution, and analyzing the spatial correlation of highly sensitive points falling into the hot spots of fishery resources, including calculating the spatial distance and the proportion of falling; The time coincidence analysis includes judging the synchrony between the occurrence time of highly sensitive points and the resource high-value time window of the fishery resource distribution; The intensity coincidence analysis includes correlating the sensitivity degree of highly sensitive points with the resource abundance or importance within the region, and judging whether the highly sensitive behavior targets high-value or protected resources.

[0012] Obtain the analysis results, manage the data and process it. If there are abnormal behaviors, give early warnings.

[0013] Furthermore, the management data at least includes: the operation permit of the fishing boat, historical activity records, fishery regulations, real-time meteorological information; comprehensively judge the analysis results of spatial coincidence analysis, time coincidence analysis and intensity coincidence analysis with the management data to construct a risk assessment rule set for illegal fishing; based on the risk assessment rule set, evaluate whether there are abnormal behaviors of the fishing boat within a time period and a region, and send early warning information to relevant regulatory departments. The early warning information includes the fishing boat suspected of abnormal behavior, time, location, behavior characteristics and risk assessment results.

[0014] The present invention also provides an intelligent fishery supervision system based on multi-source data fusion, including: The fishing boat data analysis module includes a trajectory processing unit and a trajectory analysis unit. The trajectory processing unit obtains multi-source trajectory data of the fishing boat and performs data fusion to obtain accurate trajectory data of the fishing boat; the trajectory analysis unit performs trajectory feature analysis, trajectory behavior pattern analysis and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of the fishing boat; The fishery data analysis module includes a fishery resource assessment unit and a fishery resource prediction unit. The fishery resource assessment unit obtains multi-source fishery resource data and performs data fusion to obtain a fishery resource assessment vector; the fishery resource prediction unit processes the fishery resource assessment vector according to the fishery resource prediction model to obtain the fishery resource distribution; The fishery supervision module includes a comprehensive analysis unit and an early warning unit. The comprehensive analysis unit performs spatial coincidence analysis, time coincidence analysis and intensity coincidence analysis on the trajectory distribution of the fishing boat and the fishery resource distribution to obtain analysis results; the early warning unit obtains the analysis results and management data and processes them. If there are abnormal behaviors, give early warnings.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. The method for generating accurate trajectories based on multi-source data fusion, through intelligent processing of the characteristics of different data sources and data missing patterns, and integrating behavior patterns, greatly improves the accuracy, continuity and integrity of the fishing boat trajectories; by fusing sparse but key observations such as satellite remote sensing data, reconstructs a more realistic fishing boat whereabouts, especially capable of revealing the activity trajectories of fishing boats that attempt to avoid supervision by turning off signals; the generated accurate trajectory data provides a solid and reliable basis for subsequent high-sensitivity point identification, behavior pattern analysis and superposition analysis with resource distribution.

[0016] 2. By defining and identifying hypersensitive points, potential suspicious activities can be quickly screened out from a large amount of normal trajectory data, greatly narrowing the scope of attention of supervisors and improving the efficiency of detecting abnormal behaviors. The generated trajectory distribution clearly shows the key activity areas, main behavior patterns of fishing boats, as well as the suspicious locations and time periods that need to be focused on for verification, providing a basis for deeper fusion analysis of trajectory information with other data such as resource distribution in the follow-up.

[0017] 3. The underlying fusion improves the reliability of large-scale data such as fishing reports; the middle-level fusion uses environmental and biological data to correct non-direct observations; the high-level skillfully fuses multi-source data to form a more comprehensive and robust resource assessment result. Especially in areas where scientific survey data is sparse or fishing reports are unreliable, the multi-level fusion data processing can use environmental data to make reasonable inferences about the resource status. Brief Description of the Drawings

[0018] Figure 1 It is a flowchart of a smart fishery supervision method based on multi-source data fusion provided by an embodiment of the present invention; Figure 2 It is a flowchart of generating accurate trajectory data of the final fishing boat provided by an embodiment of the present invention; Figure 3 It is a schematic structural diagram of a smart fishery supervision system based on multi-source data fusion provided by an embodiment of the present invention. Detailed Embodiments

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] Embodiment 1: In order to conduct more refined and efficient supervision and management of a certain fishery resource area, the present invention provides a smart fishery supervision method based on multi-source data fusion, and the method flow is as Figure 1 shown, and the specific implementation is as follows: Obtain multi-source trajectory data of fishing boats and perform data fusion to obtain accurate trajectory data of fishing boats; Furthermore, the multi-source trajectory data at least includes automatic identification system data, fishing boat monitoring system data, and fishing boat trajectory information from satellite remote sensing data; Segment the preprocessed multi-source trajectory data at preset time intervals, and generate several behavior pattern hypotheses for each trajectory segment based on local motion features and data sources, and assign an initial probability to each behavior pattern hypothesis. Further, the process of generating the precise trajectory data of the final fishing vessel is as Figure 2 shown. For each trajectory segment, based on the trajectory point features it contains (such as point density, speed change, heading stability) and data sources, analyze and generate several possible behavior pattern hypotheses. For example, this segment is most likely in states such as normal navigation, low-speed trawling, fixed-point anchoring, signal interruption and possible operation, etc.; according to the matching degree between the data features and known behavior patterns, assign an initial probability to each behavior pattern hypothesis.

[0021] Iteratively process the trajectory segments of the fishing vessel in chronological order. If the data density of the trajectory segment is less than the threshold, use the data of the previous and subsequent trajectory segments, as well as the behavior pattern hypothesis and its probability of this trajectory segment, to predict the trajectory range of the fishing vessel during the time period, and fuse according to the data of the trajectory segment. Further, for example, signal interruption or only a small number of satellite detection points indicate significant data loss or discontinuity. In this case, traditional filtering methods do not work well. At this time, use the behavior pattern hypothesis of this trajectory segment and its corresponding probability. According to the behavior pattern hypothesis with the highest probability, use the motion model of this behavior pattern, such as tending to move slowly in a specific area in the operation mode and tending to move in a straight line in the navigation mode, as well as the known trajectory points before and after the interruption, to predict the possible trajectory range of the fishing vessel during this time period. At the same time, regard all available sparse data points within this segment, especially the positions detected by satellite remote sensing, as observation evidence within this predicted range. According to these observation evidences, further correct the predicted trajectory range, and adopt methods such as particle filtering or optimization-based methods to fuse these sparse observations into the predicted trajectory range to reconstruct the most likely fishing vessel trajectory path during this time period. This process will dynamically adjust the probability of the behavior pattern hypothesis according to the position, quantity and confidence of the observation points, and converge to the trajectory that can best explain all observation points.

[0022] If the data density of the trajectory segment is not less than the threshold, use Kalman filtering to smooth and refine the trajectory. After completing the preliminary reconstruction of all trajectory segments, use a batch smoothing algorithm to globally smooth the entire trajectory sequence to generate the precise trajectory data of the final fishing vessel. Table 1 shows a sample of the precise trajectory data of some segments of a fishing vessel after processing.

[0023] Table 1. Sample of Precise Trajectory Data of Some Segments

[0024] A method for generating accurate trajectories based on multi-source data fusion, through intelligent processing of the characteristics of different data sources and data missing patterns, and integrating behavior patterns, greatly improves the accuracy, continuity, and integrity of fishing vessel trajectories; compared with single data sources or simple interpolation methods, it can effectively fill in the trajectory gaps during signal interruptions; by fusing sparse but critical observations such as satellite remote sensing data, it reconstructs more realistic fishing vessel whereabouts, especially capable of revealing the activity trajectories of fishing vessels that attempt to evade supervision by turning off signals. The generated accurate trajectory data provides a solid and reliable basis for subsequent high-sensitivity point identification, behavior pattern analysis, and overlay analysis with resource distribution.

[0025] Perform trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of fishing vessels; Furthermore, perform trajectory feature analysis on the accurate trajectory data, where the trajectory features at least include speed, heading, position, and mooring / residence time; based on the trajectory feature analysis, identify the behavior patterns of fishing vessels, and the behavior patterns at least include normal navigation, trawling, purse seining, fixed-point operation, stopping, drifting, turning, accelerating, or decelerating; Furthermore, the identification of behavior patterns can adopt rule-based methods. For example, if the speed changes continuously and uniformly and the heading change rate is small, it is identified as trawling, or more complex machine learning algorithms such as hidden Markov models, support vector machines, or deep learning models, which automatically classify behavior patterns by learning a large amount of labeled data.

[0026] Based on the trajectory features and behavior patterns, identify the trajectory points or segments that deviate from normal or legal behaviors as the high-sensitivity points of fishing vessels; generate a trajectory distribution representing the activity range, activity density of fishing vessels, and the position and characteristics of high-sensitivity points.

[0027] Furthermore, high-sensitivity points are potential indicators of suspicious activities, and their determination rules usually involve the combination of behavior patterns and context information. For example, the trajectory shows that a fishing vessel is moored or traveling in low speed with another vessel for a long time in a non-port area; near the edge of a known fishing ground or a sensitive area, there are frequent turning, accelerating, or decelerating behaviors, etc. These high-sensitivity points are objects that need to be focused on and further analyzed.

[0028] By defining and identifying high-sensitivity points, potential suspicious activities can be quickly screened out from a large amount of normal trajectory data, greatly narrowing the attention scope of supervisors and improving the efficiency of detecting abnormal behaviors. The generated trajectory distribution clearly shows the key activity areas of fishing vessels, the main behavior patterns, and the suspicious locations and time periods that need to be focused on for verification, providing a basis for further in-depth fusion analysis of trajectory information with other data such as resource distribution.

[0029] Obtain multi-source fishery resource data and perform data fusion to obtain a fishery resource assessment vector; Furthermore, the multi-source fishery resource data at least includes scientific survey data, fishery catch report data, environmental remote sensing data, and fishery biological data; Uniformly project the preprocessed multi-source fishery resource data onto a preset spatial grid; For point data, match and associate it with other data within the corresponding spatio-temporal grid; Construct a hierarchical fusion model, including bottom-layer fusion, middle-layer fusion, and top-layer fusion. Among them, bottom-layer fusion includes proofreading, correcting, and fusing the real data of scientific survey data and fishery catch report data; middle-layer fusion proofreads, corrects, and fuses environmental remote sensing data and fishery biological data; top-layer fusion includes obtaining the results of bottom-layer fusion and middle-layer fusion, and performing fusion to construct a fishery resource assessment vector corresponding to the spatial grid.

[0030] Furthermore, bottom-layer fusion mainly focuses on correcting and integrating the observational data directly reflecting resource information; middle-layer fusion obtains environmental remote sensing data and fishery biological data to realize the description of the constructed fishery resource ecology; top-layer fusion is to synthesize bottom-layer fusion and middle-layer fusion to generate the final fishery resource assessment vector.

[0031] Bottom-layer fusion improves the reliability of large-scale data such as catch reports; middle-layer fusion uses environmental and biological data to correct non-direct observations; the top layer cleverly fuses multi-source data to form a more comprehensive and robust resource assessment result. Especially in areas where scientific survey data is sparse or catch reports are unreliable, the middle and top layer fusions can use continuous environmental data to reasonably infer the resource status.

[0032] The fishery resource prediction model processes the fishery resource assessment vector to obtain the fishery resource distribution; Furthermore, the fishery resource prediction model is a prediction model trained based on historical multi-source fishery resource data, and the model includes but is not limited to statistical models and machine learning models; The fishery resource prediction model can be a spatio-temporal prediction model of fishery resources based on the long short-term memory network (LSTM), The fishery resource prediction model receives the fishery resource assessment vector as input and outputs the fishery resource distribution within the regulatory area, including the abundance per unit area of different fishery resource types and the total resources within the regulatory area; The fishery resource assessment vector includes the fishery resource abundance sequence and environmental factor sequence of this grid cell in the past period of time (for example, the past T time steps). The information of spatially adjacent grids is input as an additional feature to capture spatial correlation.

[0033] The distribution of fishery resources is expressed in the form of a spatial grid and includes information on the time dimension.

[0034] During the training process, historical multi-source fishery resource data is first collected and processed to construct time series samples for each spatial grid. The data is normalized to facilitate model training. The time series data is divided into a training set, a validation set, and a test set. The data from an earlier time period is used as the training set and the validation set, and the data from a subsequent time period is used as the test set. The mean squared error (MSE) is used as the loss function, and the Adam optimizer is selected as the optimizer. Finally, the trained model is used to predict the distribution of fishery resources.

[0035] Furthermore, after obtaining the evaluation vectors representing the fishery resource status in a specific region and time period through multi-source data hierarchical fusion, the next step is to use these evaluation results as inputs and generate an intuitive and comprehensive spatial distribution map of fishery resources through a pre-trained prediction model, clearly showing the abundance of different resource types and the total resources within the regulatory area.

[0036] Furthermore, the model will calculate and output the detailed distribution of fishery resources in each spatial grid cell within the regulatory area during the target time period, including: the abundance estimate of different fishery resource types within the unit area, i.e., the grid cell, such as the fish biomass per square kilometer; the abundance estimate of the total resources within the unit area, i.e., the total biomass of all fishery resource types within the cell. Furthermore, the model may also output the prediction uncertainty or confidence of these abundance estimates.

[0037] Furthermore, the output results are presented in the form of a raster map, and the value of each grid cell represents the abundance estimate of resources within that cell during a specific time period, such as a certain week, month, or quarter.

[0038] Through the processing of the fishery resource prediction model, comprehensive and continuous resource distribution information within the regulatory area can be provided; the generated spatial grid resource distribution map can be directly used as an input for overlay analysis and spatial correlation calculation with the high-sensitivity points of fishing vessel trajectories, greatly simplifying the subsequent illegal fishing risk assessment process.

[0039] Spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis are performed on the trajectory distribution of fishing vessels and the distribution of fishery resources to obtain the analysis results. Furthermore, spatial coincidence analysis includes spatially overlaying the trajectory distribution of fishing vessels and the distribution of fishery resources and analyzing the spatial correlation of high-sensitivity points falling into the hot spots of fishery resources, including calculating the spatial distance and the falling ratio. Time coincidence analysis includes judging the synchrony between the occurrence time of high-sensitivity points and the high-value time window of fishery resource distribution. The intensity coincidence analysis includes correlating the sensitivity of high-sensitivity points with the resource abundance or importance within the region to determine whether high-sensitivity behaviors target high-value or protected resources.

[0040] Furthermore, calculating the spatial distance includes calculating the minimum distance from high-sensitivity points to the boundary of the nearest resource hotspot area or high-abundance grid cells. Analyze the distribution characteristics of the distances from high-sensitivity points to hotspots, such as whether there are a large number of high-sensitivity points clustering near the boundary of the fishing moratorium area; the falling ratio analysis includes calculating the ratio of the number or duration of high-sensitivity points falling within a preset fishery resource hotspot area or grid cells with a resource abundance higher than a certain threshold to the total number or total duration of all high-sensitivity points.

[0041] Furthermore, the synchrony analysis includes calculating the ratio of high-sensitivity points occurring within the high-value time window of resource prediction. Analyze the potential time lag or lead relationship between the start, duration, or end time of high-sensitivity behaviors and the high-value time window of resources. For example, determine whether suspicious activities occur exactly during the season when a certain fish species is most likely to appear in the area or the period with the highest predicted abundance.

[0042] Furthermore, the intensity coincidence analysis aims to combine the "sensitivity" of high-sensitivity points with the "value or importance" of regional resources to evaluate the "risk intensity" of potential illegal activities. Specifically, it includes assigning a sensitivity score based on the type and characteristics of high-sensitivity points, such as the duration of AIS interruption, the severity of deviation from normal behavior, and occurrence within a specific control area; assigning a resource value or importance score to the region based on factors such as the predicted abundance of fishery resources in the region, the types of resources, and whether the region is an important breeding ground, migration route, or protected area; correlating the sensitivity score of high-sensitivity points with the resource value / importance score of the area where the point is located, such as multiplying or mapping through a risk matrix, to obtain a comprehensive "risk intensity" or "potential illegal fishing index", and analyze whether there are high-sensitivity behaviors.

[0043] This series of comprehensive analysis steps solves the problem that illegal fishing cannot be directly judged by analyzing trajectories or resources alone, creatively correlates the "behavior" of fishing vessels with the "goals" of fisheries, accurately identifies potential illegal fishing targets, provides quantitative evidence support, and focuses on key regulatory regions and time periods.

[0044] Obtain, analyze the results and management data, and if there are abnormal behaviors, issue early warnings. Table 2 shows the management data of some fishing vessels.

[0045] Table 2. Management Data of Some Fishing Vessels

[0046] Furthermore, the management data at least includes: the operation permit of the fishing boat, historical activity records, fishery regulations, and real-time meteorological information; the analysis results of spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis are comprehensively judged with the management data to construct an illegal fishing risk assessment rule set; based on the risk assessment rule set, it is evaluated whether there are abnormal behaviors of the fishing boat within a time period and a region, and a warning message is sent to the relevant supervision department. The warning message includes the fishing boat suspected of abnormal behavior, time, location, behavior characteristics, and risk assessment results.

[0047] This final step is the closed-loop link of the entire intelligent fishery supervision method and the key to transforming data analysis capabilities into actual supervision effectiveness. It combines complex analysis results with real-world supervision rules, vessel backgrounds, and environmental factors to automatically conduct risk assessments and achieve intelligent decision-making assistance.

[0048] Through an intelligent fishery supervision method based on multi-source data fusion provided by the present invention, it integrates multiple information flows of fishing boat trajectories and fishery resources, realizing higher-precision, more comprehensive, and more continuous supervision of fishery activities. Its core advantage lies in that it can not only generate accurate fishing boat trajectories that overcome the limitations of a single data source and fishery resource distribution maps predicted based on multi-source data, but more importantly, it can actively associate and verify potential abnormal behaviors of fishing boats with hot spots of fishery resources through intelligent spatio-temporal and intensity coincidence analysis, and conduct comprehensive judgments in combination with management data such as fishing boat qualifications and regulations, accurately identifying and filtering out high-risk events that are very likely to be suspected of illegal fishing, focusing the supervision focus on specific vessels, times, and locations where illegal acts are most likely to occur, greatly improving the efficiency and accuracy of fishery supervision.

[0049] Embodiment 2: The present invention also provides an intelligent fishery supervision system based on multi-source data fusion. The system structure is as Figure 3 shown, and the specific implementation method is as follows: The fishing boat data analysis module includes a trajectory processing unit and a trajectory analysis unit. The trajectory processing unit obtains multi-source trajectory data of the fishing boat and conducts data fusion to obtain accurate trajectory data of the fishing boat; the trajectory analysis unit conducts trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of the fishing boat; Furthermore, in the trajectory processing unit, the multi-source trajectory data at least includes automatic identification system data, fishing boat monitoring system data, and fishing boat trajectory information from satellite remote sensing data; The preprocessed multi-source trajectory data is segmented at a preset time interval, and several behavior pattern hypotheses are generated for each trajectory segment based on local motion characteristics and data sources, and an initial probability is assigned to each behavior pattern hypothesis; Iteratively process the trajectory segments of fishing vessels in chronological order. If the data density of a trajectory segment is less than the threshold, use the data of the previous and subsequent trajectory segments, as well as the behavior pattern hypothesis and its probability of this trajectory segment, to predict the trajectory range of the fishing vessel during the time period, and perform fusion based on the data of the trajectory segment; If the data density of the trajectory segment is not less than the threshold, use Kalman filtering to smooth and refine the trajectory; After completing the preliminary reconstruction of all trajectory segments, use a batch smoothing algorithm to globally smooth the entire trajectory sequence to generate the accurate trajectory data of the final fishing vessel.

[0050] Furthermore, in the trajectory analysis unit, perform trajectory feature analysis on the accurate trajectory data. The trajectory features at least include speed, course, position, and mooring / residence time; based on the trajectory feature analysis, identify the behavior patterns of the fishing vessel. The behavior patterns at least include normal navigation, trawling operation, purse seining operation, fixed-point operation, stop, drifting, turning, acceleration, or deceleration; According to the trajectory features and behavior patterns, identify the trajectory points or segments that deviate from normal or legal behaviors as the sensitive points of the fishing vessel; generate a trajectory distribution representing the activity range, activity density, and the position and characteristics of the sensitive points of the fishing vessel.

[0051] The fishery data analysis module includes a fishery resource assessment unit and a fishery resource prediction unit. The fishery resource assessment unit obtains multi-source fishery resource data and performs data fusion to obtain a fishery resource assessment vector; the fishery resource prediction unit processes the fishery resource assessment vector according to the fishery resource prediction model to obtain the fishery resource distribution; Furthermore, in the fishery resource assessment unit, the multi-source fishery resource data at least includes scientific survey data, fishery catch report data, environmental remote sensing data, and fishery biological data; Project the preprocessed multi-source fishery resource data onto a preset spatial grid uniformly; For point data, match and associate it with other data within the corresponding spatio-temporal grid; Construct a hierarchical fusion model, including bottom-layer fusion, middle-layer fusion, and top-layer fusion. The bottom-layer fusion includes proofreading, correcting, and fusing the real data of scientific survey data and fishery catch report data; the middle-layer fusion proofreads, corrects, and fuses environmental remote sensing data and fishery biological data; the top-layer fusion includes obtaining the results of the bottom-layer fusion and the middle-layer fusion, and performing fusion to construct a fishery resource assessment vector corresponding to the spatial grid.

[0052] Furthermore, in the fishery resource prediction unit, the fishery resource prediction model is a prediction model trained based on historical multi-source fishery resource data, and the model includes but is not limited to statistical models and machine learning models; The fishery resource prediction model receives the fishery resource assessment vector as input and outputs the fishery resource distribution within the regulatory area, including the abundance per unit area of different fishery resource types and the total resources within the regulatory area. The fishery resource distribution is expressed in the form of a spatial grid and includes time dimension information.

[0053] The fishery supervision module includes a comprehensive analysis unit and an early warning unit. The comprehensive analysis unit conducts spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis on the trajectory distribution of fishing vessels and the fishery resource distribution to obtain the analysis results. The early warning unit obtains the analysis results and management data and processes them. If there are abnormal behaviors, early warnings are issued.

[0054] Further, in the comprehensive analysis unit, the spatial coincidence analysis includes spatially overlaying the trajectory distribution of fishing vessels and the fishery resource distribution, and analyzing the spatial correlation of high-sensitivity points falling into the fishery resource hot spots, including calculating the spatial distance and the falling ratio. The time coincidence analysis includes judging the synchronization between the appearance time of high-sensitivity points and the resource high-value time window of the fishery resource distribution. The intensity coincidence analysis includes correlating the sensitivity degree of high-sensitivity points with the resource abundance or importance within the area to judge whether the high-sensitivity behavior targets high-value or protected resources.

[0055] Further, in the early warning unit, the management data at least includes: the operation permit of fishing vessels, historical activity records, fishery regulations, and real-time meteorological information. The analysis results of spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis are comprehensively judged with the management data to construct a risk assessment rule set for illegal fishing. Based on the risk assessment rule set, it is evaluated whether there are abnormal behaviors of fishing vessels within a time period and an area, and early warning information is sent to relevant regulatory departments. The early warning information includes the fishing vessels suspected of abnormal behaviors, time, location, behavior characteristics, and risk assessment results. Table 3 shows some of the monitored early warning data. Through the system provided by the present invention, efficient supervision of fisheries within the area can be achieved.

[0056] Table 3. Summary of Early Warning Information

[0057] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart fishery supervision method based on multi-source data fusion, characterized in that, Including: Obtain multi-source trajectory data of fishing vessels and perform data fusion to obtain accurate trajectory data of fishing vessels; Conduct trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of fishing vessels; Obtain multi-source fishery resource data and perform data fusion to obtain a fishery resource assessment vector; the fishery resource prediction model processes the fishery resource assessment vector to obtain the fishery resource distribution; Conduct spatial coincidence analysis, temporal coincidence analysis, and intensity coincidence analysis on the trajectory distribution of fishing vessels and the fishery resource distribution to obtain the analysis results; obtain the analysis results and management data and process them. If there are abnormal behaviors, give an early warning.

2. The intelligent fishery supervision method based on multi-source data fusion according to claim 1, wherein, The multi-source trajectory data at least includes Automatic Identification System (AIS) data, fishing vessel monitoring system data, and fishing vessel trajectory information from satellite remote sensing; Segment the preprocessed multi-source trajectory data at a preset time interval, and based on local motion characteristics and data sources, generate several behavior pattern hypotheses for each trajectory segment and assign an initial probability to each behavior pattern hypothesis; Iteratively process the trajectory segments of the fishing vessel in chronological order. If the data density of the trajectory segment is less than the threshold, use the data of the previous and subsequent trajectory segments, as well as the behavior pattern hypothesis and its probability of this trajectory segment, to predict the trajectory range of the fishing vessel during the time period and perform fusion according to the data of the trajectory segment; If the data density of the trajectory segment is not less than the threshold, use the Kalman filter to smooth and refine the trajectory; After completing the preliminary reconstruction of all trajectory segments, use a batch processing smoothing algorithm to globally smooth the entire trajectory sequence to generate the accurate trajectory data of the final fishing vessel.

3. The intelligent fishery supervision method based on multi-source data fusion according to claim 1, characterized in that: Conduct trajectory feature analysis on the accurate trajectory data, and the trajectory features at least include speed, course, position, berthing, and residence time; Based on the trajectory feature analysis, identify the behavior patterns of fishing vessels, and the behavior patterns at least include normal navigation, trawling, purse seining, fixed-point operation, stopping, drifting, turning, accelerating, and decelerating; According to the trajectory features and behavior patterns, identify the trajectory points and / or trajectory segments that deviate from normal and legal behaviors as the high-sensitivity points of fishing vessels; generate a trajectory distribution representing the activity range, activity density, and the position and characteristics of high-sensitivity points of fishing vessels.

4. A smart fishery supervision method based on multi-source data fusion according to claim 1, characterized in that, The multi-source fishery resource data at least includes scientific investigation data, fishery catch report data, environmental remote sensing data, and fishery biology data; Project the preprocessed multi-source fishery resource data onto a preset spatial grid uniformly; For point data, match and associate it with the data within the corresponding spatio-temporal grid; Construct a hierarchical fusion model, including bottom-layer fusion, middle-layer fusion, and high-layer fusion. Among them, the bottom-layer fusion includes proofreading, correcting, and fusing the real data of scientific investigation data and fishery catch report data; The middle-layer fusion proofreads, corrects, and fuses environmental remote sensing data and fishery biology data; The high-layer fusion includes obtaining the results of the bottom-layer fusion and the middle-layer fusion, and performing fusion to construct a fishery resource assessment vector corresponding to the spatial grid.

5. A smart fishery supervision method based on multi-source data fusion according to claim 1, characterized in that, The fishery resource prediction model is a prediction model trained based on historical multi-source fishery resource data, and the model includes but is not limited to statistical models and machine learning models; The fishery resource prediction model receives a fishery resource assessment vector and outputs the fishery resource distribution within the regulatory area, including the abundance per unit area of different fishery resource species and the total resources within the regulatory area; The fishery resource distribution is expressed in the form of a spatial grid and contains time dimension information.

6. A smart fishery supervision method based on multi-source data fusion according to claim 1, characterized in that The spatial coincidence analysis includes spatially overlaying the trajectory distribution of fishing vessels with the fishery resource distribution, and analyzing the spatial correlation of high-sensitivity points falling into the fishery resource hot spots, including calculating the spatial distance and the falling-in ratio; The time coincidence analysis includes judging the synchronization between the occurrence time of high-sensitivity points and the resource high-value time window of the fishery resource distribution; The intensity coincidence analysis includes correlating the sensitivity degree of high-sensitivity points with the resource abundance and importance within the area, and judging whether the high-sensitivity behavior targets high-value and protected resources.

7. A smart fishery supervision method based on multi-source data fusion according to claim 1, characterized in that The management data at least includes: the operation permit of fishing vessels, historical activity records, fishery regulations, and real-time meteorological information; comprehensively judging the analysis results of spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis with the management data to construct an illegal fishing risk assessment rule set; based on the risk assessment rule set, assessing whether there are abnormal behaviors of fishing vessels within a time period and an area, and sending a warning message to the relevant regulatory departments, where the warning message includes the fishing vessels suspected of abnormal behaviors, time, location, behavior characteristics, and risk assessment results.

8. A smart fishery supervision system based on multi-source data fusion, which executes a smart fishery supervision method based on multi-source data fusion as described in claim 1, characterized in that, including: The fishing vessel data analysis module includes a trajectory processing unit and a trajectory analysis unit. The trajectory processing unit obtains multi-source trajectory data of fishing vessels and performs data fusion to obtain accurate trajectory data of fishing vessels; the trajectory analysis unit performs trajectory feature analysis, trajectory behavior pattern analysis, and high-sensitivity point identification on the accurate trajectory data to obtain the trajectory distribution of fishing vessels; The fishery data analysis module includes a fishery resource assessment unit and a fishery resource prediction unit. The fishery resource assessment unit obtains multi-source fishery resource data and performs data fusion to obtain a fishery resource assessment vector; the fishery resource prediction unit processes the fishery resource assessment vector according to the fishery resource prediction model to obtain the fishery resource distribution; The fishery supervision module includes a comprehensive analysis unit and a warning unit. The comprehensive analysis unit performs spatial coincidence analysis, time coincidence analysis, and intensity coincidence analysis on the trajectory distribution of fishing vessels and the fishery resource distribution to obtain analysis results; the warning unit obtains the analysis results and management data and processes them. If there are abnormal behaviors, a warning is issued.

Images